Bidding against the machine: Emotions in electronic markets Marc Adam Karlsruhe Institute of Technology (KIT) ABSTRACT: This paper investigates the emotions of market participants and their impact on decision-­‐
making. We measure skin conductivity and heart rate during an experiment in an electronic market with human or automated counterparties. These measures serve as proxies for emotions and are combined with market results to provide insight into participant’s emotions during the auctions and at discrete events, such as submitting a bid and winning or losing an auction. We can show that arousal is stronger in human only markets and when the stakes of the experiment are higher. Interestingly, we find that arousal is negatively correlated with the bid price suggesting that humans are more willing to take risks when competing against other humans. The results have important implications for the design of electronic commerce platforms and markets where both humans and automated agents are active eBay is the most obvious example. Author(s): Marc Adam; Timm Teubner; Ryan Riordan What Makes Voters Turn Out: The Effects of Polls and Beliefs Marina Agranov Caltech ABSTRACT: We use laboratory experiments to test for one of the foundations of the rational voter paradigm -­‐ that voters respond to probabilities of being pivotal. We exploit a setup that entails stark theoretical effects of information concerning the preference distribution (as revealed through polls) on costly participation decisions. The data reveal several insights. First, voting propensity increases systematically with subjects’ predictions of their preferred alternative’s advantage. Consequently, pre-­‐election polls do not exhibit the detrimental welfare effects that extant theoretical work predicts. They lead to more participation by the expected majority and generate more landslide elections. Finally, we investigate subjects’ behavior in polls and identify when Bandwagon and Underdog Effects arise. Author(s): Marina Agranov; Jacob Goeree; Julian Romero; Leeat Yariv On the role of anonymity for peer punishment and cooperation Antonio Alonso The Univeristy of Nottingham ABSTRACT: Numerous laboratory experiments have shown that cooperation can be sustained with the introduction of peer punishment. However, recent critics suggest that the success of punishment might be overstated. They argue that sanctioners are protected by anonymity, a condition which might not hold outside the laboratory. Here we report experiments that systematically vary the extent of anonymity on three dimensions: interactions, punishment decisions and identities. Our preliminary results reveal that the benefits of peer punishment on cooperation are robust under conditions without anonymity. Author(s): Antonio Alonso; Simon Gaechter; Chris Starmer; Jeroen Nieboer Pricing Prototypical Products Wilfred Amaldoss Duke University ABSTRACT: When we think of colas, Coca-­‐Cola first comes to mind. Products such as Cola-­‐Cola, Tide laundry detergent and Chapstick lip balm are the prototypical products in their respective categories. Over three decades, research in consumer psychology has accumulated evidence on how prototypicality structures memory, shapes the composition of consideration set and influences purchase decision. Yet there is no research on how it affects the competitive behavior of firms. For example, some prototypical products are lower priced than other products in their category, whereas in certain other categories the prototypical product is higher priced. Using a novel model, we make an initial effort to examine theoretically the impact of prototypicality on the pricing decisions of firms competing in a horizontally differentiated market. Our analysis shows that when consumer valuations are low, the prototypical product charges a lower price and yet earns more profits compared to other products in the category. However, when consumer valuations are high, the rank order of the prices of the prototypical product and a nonprototypical product becomes reversed, but not the order of profits. We subject these predictions to an empirical test. The experimental results lend support for the qualitative predictions of the model. Author(s): wilfred Amaldoss; Chuan He Asymmetric Dominance Effect in the Financial Decision Making -­‐ An Experimental Study. Aparna Anand CUNY, GC ABSTRACT: The concept of asymmetric dominance effect (ADE) is in individual choices between alternatives x and y, the availability of a third alternative z, inferior to x but not to y, tends to increase the preferences for x. In ADE literature, this inferior alternative z is termed as 'decoy'. Our study explores the presence of asymmetric dominance effect in the context of financial decision making. In specific we want to study whether the preference to a particular choice of monetary stream of payments is increased in the presence of its decoy. Experimental Design. The design consists of five treatments. Each treatment had 50 questions; in each question the subjects are to choose an option from the given set of monetary stream of payments. The treatment 1 is the baseline treatment in which each subject was asked to make a choice out of two options of payments. An example of one such question is to choose between option 1:$3 today and $17 in 2 weeks and option 2: $15 today and $2 in 2 weeks. The first 25 questions had options as described above with different dollar amounts and the remaining 25 questions were designed to test the stationarity, in which the payment options (today and later) were equally delayed by 2 weeks. The options of payments are constructed such that each option is dominant in either the dimension of time or total amount. The stream of payments such as option 1 above are referred as total amount (quantity) dominant and payments such as option 2 as time dominant. The questions of treatments 2 and 4 included decoy options in addition to time dominant and quantity dominant choices. The decoys are constructed such that it is inferior to one of the options in the given choice set not to the other. The treatments 3 and 5 are the repeated versions of treatments 2 and 4 with the order of the questions and the label of the options (time and quantity dominant) reversed. A total number of 97 subjects participated in this experiment. Preliminary Results. The main findings are that we didn’t observe an increase in preference to an option in the presence of its decoy as suggested by the ADE theory. In the treatment that had decoy for the quantity dominant option, on average 20% of the subjects chose the decoy. There is no significant decoy effect observed for the time dominant decoys, but a large imprecise (10% level) effect of the quantity decoy on the probability of choosing the time dominant option. There is a significant delay effect for treatment 1, where there are no decoys. In treatment 1, there is a 7% reduction in time dominant choices in delayed options. This delay effect disappears in the decoy treatments. So we observe a small ‘hyperbolic’ sort of effect in treatment 1, but none in the other treatments. Also bigger the differences between the initial payments and the second payment, in the quantity and time dominant options, the less likely the subjects are to choose the time dominant option. Though the decoy effect was not observed in the context of ADE, the hyperbolic effects seem to disappear in the treatments that had decoy options. Author(s): Barry Sopher; Aparna Anand Risk-­‐taking on Behalf of Others Ola Andersson Lund University ABSTRACT: In this paper, we study risk-­‐taking on behalf of others experimentally using a large sample of the Danish population. Two web-­‐based and incentivized experiments are conducted using a variant of the well-­‐known multiple price list format. We find that people act responsibly with other people’s money if they do not have strong incentives to divert from this principle. To the contrary, when they are facing hedged reimbursement contracts or competition for reimbursement, we find that people take on excessive risks on behalf of others and that such risk-­‐taking is negatively affected by pro-­‐social preferences. Author(s): Ola Andersson; Jean-­‐Robert Tyran; Erik Wengstrom Rob Peter to Pay Paul. Are Charitable Contributions Substitutes? Olivier Armantier Federal Reserve Bank of NY ABSTRACT: We conduct a field experiment to establish the extent to which gifts to different charitable organizations are substitute goods. To do so, we solicit individuals to contribute to three charities A, B and C. The fundraising campaigns for A and B are separated by less than a week, while the fundraising campaign for C is conducted three months later. The subject pool consists of people approached to contribute to charity A. It is partitioned in three groups: regular donors, occasional donors and prospects. Unlike subjects in the control treatment, subjects in the ‘matching treatment’ are offered a 1:1 match for charity A. The question addressed in the paper is therefore: How does lowering the price of A, affect giving to charity A, B and C? Preliminary results suggest strong substitution effects in the short and, to a smaller extent, in the long run: more subjects give to charity A in the matching treatment, but they are less likely to give to B and C. The results also reveal that, although the matching grant influences the extensive margin of giving, it does not affect significantly the amount subjects who donate give to the three charities. Furthermore, we find that substitution affects primarily subjects who give frequently to charity A, as well as low income donors. More importantly, our results indicate that marketing schemes aimed at increasing donations to a charity (e.g. matching grants) not only have negative externalities on other charities, but they may also be socially inefficient. Indeed, we find that the total dollar amount given to the three charities is lower in the matching treatment. These results are important as they suggest that charitable organizations may have an incentive to compete in price to divert donations from other charitable institutions, thereby increasing the price of giving to charities and lowering the total amount given to charities. As a result, governments wishing to promote charitable giving may have an incentive to regulate the charity market and design specific tax policies (e.g. introduce deduction caps on amount given to a specific charity). Author(s): Olivier Armantier Emotion Regulation and the Disposition Effect Philipp Astor Forschungszentrum Informatik (FZI) ABSTRACT: There is evidence that the way individuals regulate their emotions has material consequences on behavior and financial decision performance. Psychological research differentiates broadly among subjects using predominantly antecedent-­‐focused (reappraisal) or response-­‐focused strategies (suppression) for emotion regulation. In order to gain more insight into how emotions and their regulation shape our financial decision making, we conducted a laboratory experiment with the goal to investigate how different emotion regulation strategies affect the disposition effect (N=100). By means of the Emotion Regulation Questionnaire (ERQ) and physiological measures (for skin conductance and heart rate) we identify subjects using predominantly reappraisal or suppression strategies. Our major results are (i) that using preferably suppression strategies does not influence decision-­‐making, and (ii) that using preferably reappraisal strategies results in a significant stronger susceptibility to the disposition effect. Based on these results, we conducted a second lab experiment (N=50), where we induce mindfulness, applying a 15-­‐minute meditation task prior to the session. Mindfulness is known for increasing awareness and attention. Interestingly, the subjects in the mindfulness treatment react more susceptible to the disposition effect than those without the mindfulness induction. More precisely, whereas subjects using dominantly reappraisal strategies behave similar to those without the mindfulness induction, those, who do not use dominantly reappraisal strategies show a significantly stronger exposedness to the disposition effect in the mindfulness induction treatment. Author(s): Philipp J. Astor; Marc T. P. Adam An experiment on resource and welfare division through a modified trust game Giuseppe Attanasi Toulouse School of Economics ABSTRACT: We design an experiment on sequential common-­‐pool resource extraction with side-­‐payments. Two players share a common resource sequentially. Each player is endowed with a production function transforming units of the resource into wealth. The production function is linear with diminishing return above a threshold. The first player decides how much to extract and, therefore, how much to leave to the other player. The latter might decide to transfer part of his production to the former. In this set-­‐up, which can be interpreted as a modification of the classical trust game (Berg et al., 1995), we define several natural solutions related to different concepts of fairness. Our experimental design is aimed at analyzing which of the natural solutions emerge in the strategic form and in the repeated version of the sequential common-­‐
pool resource extraction game. We find that efficiency can be achieved under specific compensation schemes, when the second player has a technological advantage and the game is repeatedly played within the same pair. Furthermore, we find that the non-­‐cooperative outcome is more likely to emerge for those pairs where players do not share the same idea of fairness. Author(s): Ambec; S.; Attanasi; G; Reynaud A Endogenous Entry across Auction Formats: Willingness to Pay and Threshold Entry Decisions Diego Aycinena Universidad Francisco Marroquin ABSTRACT: We experimentally examine threshold entry decisions in first-­‐price and English clock auctions with independent private values, when there is an explicit cost of participating in the auction. Bidders privately observe their value and report threshold entry decisions by revealing their maximum willingness to pay (WTP) to enter the auction via a Becker deGroot Marschak (BDM) mechanism. Every bidder whose revealed WTP exceeds a randomly drawn cost of participation enters the auction and incurs said cost of participation. Bidders are informed of the number of bidders in the auction prior to placing their bids. We find that although revealed WTP is higher than theoretical predictions in both auction formats it is increasing in the bidder’s value as predicted by theory. We also find a small revealed preference for first-­‐price auctions. In addition, we find that men have higher WTP for first-­‐price auctions than women, but with no corresponding gender difference in English clock auctions. Most surprisingly, bidders in both auction formats reveal a higher WTP when there are five potential bidders as opposed to three. Author(s): Diego Aycinena; Hernan Bejarano; Lucas Rentschler Cooperative Networks: Altruism, Group Solidarity, and Reciprocity in Ugandan Producer Organizations Delia Baldassarri Princeton University ABSTRACT: Repeated interaction and social networks are commonly considered a viable solution to collective action problems. However, scholars have rarely tested the various dispositional mechanisms that trigger cooperation among interconnected actors. In this paper I distinguish between four different mechanisms, i.e., generalized altruism, group solidarity, reciprocity, and the threat of sanctioning, and test which of them bring about cooperation in the context of 50 Ugandan producer organizations that face collective action problems on a regular basis. Using a methodological framework that combines lab-­‐in-­‐the-­‐field experiments with survey interviews and complete social networks data (N= 1,447), this paper goes beyond the assessment of a relationship between social networks and collective outcomes to study the dispositional mechanisms that are at the basis of cooperative behavior. First, I show that farmers with greater network centrality and non-­‐redundant social ties cooperate and participate more to the life of their producer organization. Second, relying on individual behavior in different variants of the dictator and public goods games, I show that group identification affects pro-­‐social behavior, and that repeated interaction and sanctioning both have a causal effect on cooperation. Third, I relate individual’s behavior in behavioral games to their level of cooperation in the life of the producer organization. Results show that, in the context of the farmer organizations object of this study, cooperation is not induced by other-­‐regarding preferences like altruism or group solidarity. Rather, repeated interaction favors the development of mechanisms of direct and direct reciprocity. Author(s): Delia Baldassarri Externalities, Fairness and Consumer Information in Markets Bjorn Bartling University of Zurich ABSTRACT: The efficiency of market exchange depends crucially on whether prices reflect all costs involved in production. We study a competitive market where buyers and sellers can exchange goods, but production of the goods may impose a negative externality on third parties. Imposing the externality reduces the cost of production for a seller, but the size of the externality is such that it renders trade inefficient. We analyze information acquisition decisions by consumers facing the possibility of externality-­‐producing products. We compare treatments in which the characteristics of a product are transparent, to different treatments in which consumers have to acquire product information about externalities. The treatments vary the cost of acquiring information and the presence of cheap-­‐talk product labels. We study the information acquisition of consumers, the incentives this creates for producers, and the composition and prices of goods in the market. We find persistent concerns for fairness: bad products trade at lower prices than good products and both types of products have a substantial market share. Thus, market exchange is not fully efficient but fairness towards the third party does not vanish in a competitive market. The willingness of consumers to acquire product information is generally robust to small costs to becoming informed. However, we also find that many buyers remain ignorant about the type of product and simply choose the cheapest one available, which increases the market share of the bad product. Author(s): Bjoern Bartling; Roberto Weber The Effects of Strategic Risk Aversion and Risk Perceptions on Participation in New Markets Hernan Bejarano Penn State ABSTRACT: This paper will present results of a study exploring how risk preferences and perceptions influence participation decisions in new markets. The research is conducted using student subjects in lab experiments and farmers in artifactual field experiments (Harrison and List 2004). Our interest in new and developing markets for ecosystem services that target farm landowners motivates our use of farmers in the experiments. Experimental tasks are a 2-­‐by-­‐2 game and a risk elicitation task. The 2-­‐by-­‐2 game called private value participation game (PVPG) is inspired by global games and mimics new environmental markets in which subjects choose between participating or not. Payoffs from participation depend on the private value and others’ choices. The payoff from no participation is certain. We find that participation premiums can induce experimental subjects to choose participation levels similar to those predicted by theory. Efficiency is affected by individual and group heterogeneity. Similar to previous global games experiments, a large proportion of the subjects play threshold strategies. But average thresholds differ from theoretical predictions. Subjects’ differences in participation decisions cannot be explained by their degrees of risk aversion. Additionally, when we compare student and farmer subjects, we find that the proportion of farmers choosing no participation with high frequency is significantly larger Analyzing the different choices of students and farmers reveals the risks of generalizing the outcomes of experimental findings to other populations of subjects. Author(s): Hernan Bejarano; James Shortle Measuring Ratchet Effects within a Firm: Evidence from a Field Experiment varying Contractual Commitment Charles Bellemare Laval University ABSTRACT: We present results from a field experiment designed to measure the importance of managerial commitment to a contract within a firm that pays its workers piece rates. In the tree planting industry the piece rate paid to workers is determined as a function of the difficulty of the terrain to be planted. During the experiment, firm managers told a crew of tree planters that the piece rate had yet to be determined on a particular block of land. The manager further told the workers that he would start them at a trial piece rate, but would revise the piece rate upwards if, after a few work days, average production levels were below those observed on similar land (on which he paid a piece-­‐rate equal to the trial piece-­‐rate). We compare worker productivity during the observation period (the two days without commitment) with productivity on the control land used by the manager, and planted by the same workers. Our results suggest that worker productivity decreased substantially during the observation period; in the order of 20\%, giving empirical support for the importance of commitment to a piece rate. Moreover, the reduction in productivity was less pronounced when workers had less time to benefit from any subsequent increase in the piece rate. The later provides support for models of worker turnover as a means to overcome ratchet effects. Author(s): Charles Bellemare; Bruce Shearer Team composition/diversity and performance: A (soccer) field study Avner Ben-­‐Ner University of Minnesota ABSTRACT: The relationship between team composition and diversity on the one hand and performance on the other hand have been studied in the lab by economists, psychologists and other behavioral scientists (see Kugler, Kausel & Kocher, 2012 and Joshi & Roh, 2009 for surveys). A few field studies, many focused on sports teams, have also been conducted (e.g., Timmerman, 2000, Franck & Nuesch, 2010). The present paper takes a novel approach to the investigation of the relationship between team composition and performance. (1) We consider a wide range of player-­‐level variables, including ethnicity, nationality, language, age, experience and talent composition, as well manager characteristics, fan base, and so on, that may affect individual and team performance. (2) We recognize explicitly the strategic nature of the game and that the choice of a team’s composition for a particular game depends on the composition of the rival team. (3) We implement the strategic aspect of the game in our empirical specifications by modeling a) interactions between teams in individual games and the choice of the composition of the two opposing teams, b) longer-­‐term choices of team composition and manager for a season and c) individual player performance in individual games and seasons. (4) We use multiple measures of team performance season standing, goals scored and conceded per game, wins/losses and draws as well as individual performance game rating, goals scored or blocked, assists, and so on. (5) Our empirical strategy is based on FE and two-­‐stage IV GMM estimation that recognizes the possible endogeneity of team composition/diversity and other variables. Our dataset consists of 10 seasons (2000/2001-­‐2009/2010) of the 18-­‐team top German soccer league (Bundesliga). We study all league 3,060 games for the 28 teams that played in the league for at least one season during the sample period, and the 1,723 players who were on the roster of these teams for at least one season. Author(s): Avner Ben-­‐Ner; John-­‐Gabriel Licht; Jin Park Privacy Concerns and Unraveling in a Lemons Market: An Experiment Volker Benndorf University of DÃ_sseldorf ABSTRACT: In a lemons-­‐market experiment with six participants, we analyze whether subjects choose to reveal their type at a cost. By revealing her type, a player can improve her payoff if her productivity minus the revelation cost is above the market average, in which case revelation imposes a negative externality on the other players. We find that subjects generally choose to reveal too little compared to the Bayesian Nash equilibrium benchmark. We explore the reasons for this result and find that is due to limited depth of reasoning as our median subject fails to reason beyond k=2 levels. Strikingly, we find a statistically and economically significant framing effect: a loaded frame where subjects may reveal their productivity by purchasing a health certificate leads to even lower revelation rates, suggesting that subjects have privacy concerns in such contexts. Author(s): Volker Benndorf; Dorothea Kuebler; Hans-­‐Theo Normann Fair value accounting and investment behavior Are nonprofessional investors naïˉve? Ralf Bergheim Department of Macroeconomics, Univ. Bochum ABSTRACT: This paper examines the investment decisions of nonprofessional investors under Fair Value Accounting (FVA). In particular, we experimentally investigate whether nonprofessional investors incorporate note disclosures mandated by SFAS No. 157 on fair values into their investment decisions or if they invest in a more naive manner. Our study focuses on two important issues. First, we investigate whether and to what extent the investment decisions of nonprofessional investors are affected by valuations based on different input levels according to the fair value hierarchy. Second, we investigate how nonprofessional investors respond to different variations in net income caused by fair value adjustments in trading assets. Our results document that nonprofessional investors significantly distinguish between different input levels and that the willingness to invest is decreasing across the three levels of the fair value hierarchy. In addition, we observe lower willingness to invest in any case of more variation in net income caused by fair value adjustments -­‐ even if net income is increased. In further analyses we find that the attributed risk to investments and the judgment about future performance are the main drivers for the observed investment decisions. We conclude that non-­‐professional investors are able to incorporate note disclosures about the input levels of fair values into their decisions. Furthermore, they do not naively invest more when net income is increased by a fair value change in trading assets. Our findings have implications for standard setters and regulators. Moreover, our results provide new insight to the debate about FVA. Author(s): Bergheim; Ralf. Ernstberger; Juergen. Roos; Michael W. M. Gender differences in the reference group neglect: An experimental test Noemi Berlin Paris 1, PSE ABSTRACT: Men are known to have a higher taste for competition than women, but the determinants of choice to enter a competition are still not fully understood. Beliefs and the competition level one evolves in are often suspected to play an important role in this choice. Our experimental design aims at understanding how participants update their beliefs after receiving a feedback informing them of whether their performance is below or above the median performance and how they choose whether to enter a competition against a participant belonging to the same ability group, who thus received the same feedback as they did. Our first result is that, participants react too strongly to the feedback they received, implying a deviation from Bayesian beliefs. This effect is larger for women than for men. As far as entry in the competition is concerned, low-­‐performing participants adjust their entry decision to the level of the competition while high-­‐performing participants do not. However, the behaviors leading to these results are quite different for men and women: women mainly react to the information on their own performance while men seem to respond more to their beliefs concerning the level of the competition they will be evolving in. Author(s): Noemi Berlin; Marie-­‐Pierre Dargnies Designing a Sequential Choice Architecture to Reduce Choice Overload Tibor Besedes Georgia Institute of Technology ABSTRACT: Previous studies have demonstrated that a multitude of options can lead to choice overload, reducing decision quality. Through controlled experiments, we examine sequential choice architectures that enable the choice set to remain large while potentially reducing the effect of choice overload. A specific tournament-­‐style architecture achieves this goal. An alternate architecture in which subjects compare each subset of options to the most preferred option encountered thus far fails to improve performance due to the status quo bias. Subject preferences over different choice architectures are negatively correlated with performance, suggesting that providing choice over architectures might reduce the quality of decisions. Author(s): Tibor Besedes, Cary Deck, Sudipta Sarangi, Mikhael Shor The Impact of Group Membership on Preferences for Equality and Efficiency Sudeep Bhatia Carnegie Mellon University ABSTRACT: I explore altruism towards in-­‐group, out-­‐group and anonymous recipients in a modified dictator game. I estimate both individual level and group level Constant Elasticity of Substitution (CES) utility functions, and find that higher order motives guiding choice change systematically as group membership is varied. Altruism towards in-­‐group recipients is best represented by elasticity of substitution greater than -­‐1, indicating a relative preference for equal outcomes. In contrast, altruism towards out-­‐group recipients is less sensitive to inequality, and is best represented by social welfare maximizing preferences with elasticity of substitution less than -­‐1. Altruism towards anonymous recipients is between these two extremes. These results are accompanied by strong gender effects. As with previous work, I find that men are driven by efficiency concerns whereas women are driven by equality concerns. This difference, however, applies only to the anonymous treatment. Both men and women display a preference for equal outcomes when giving to in-­‐group recipients, and a preference for social welfare maximizing outcomes when giving to out-­‐group recipients. These findings suggest that group membership and other contextual variables affect not only the weights that the giver places on the recipient’s payoffs, but also the types of tradeoffs that the giver is willing to make. Subsequently non-­‐linear functional forms, such as the CES utility function, are better suited to modeling the interplay of these variables with altruistic choice, compared to their linear counterparts. Author(s): Sudeep Bhatia Social Status and the Moral Hazard Problem: Experimental Evidence Haimanti Bhattacharya University of Utah ABSTRACT: We analyze how real social status of parties (a principal and an agent) facing a moral hazard problem may affect the problem. To test the above question, we conduct a field experiment in India where the extant caste system provides an exogenous status hierarchy. We find that when high status principals are aware of the low status of agents, contracts are rarely offered than when high status principals are unaware, leading to high incidence of sub-­‐optimal outcomes. In contrast, when low status principals are aware of the high status of agents, contracts are almost always offered than when low status principals are unaware, thus again leading to serious market failure. When principals and agents share the same social status and this information is available to both parties, the frequency of the socially optimal outcome is significantly higher than when this information is unavailable to one of the two parties. Our results have important implications for the design of non-­‐pecuniary contract enforcement devices. Author(s): Haimanti Bhattacharya; Subhasish Dugar Splitting an Uncertain Natural Resource: Evidence on Risk-­‐Taking Behavior Olivier Bochet University of Bern ABSTRACT: We conduct an experiment on a stochastic version of the celebrated Nash demand game (Nash, 1953) coupled with a risk preferences elicitation (Holt and Laury, 2002). A stock of resource available can take two different values -­‐-­‐high and low-­‐-­‐ with probabilities $p$ and $(1-­‐p)$. Agents face uncertainty on the realized value of the stock of resource. Laurent-­‐Lucchetti, Leroux and Sinclair-­‐Desgagne (2011) show in this set-­‐up that there exist three types of Nash equilibria: cautious equilibria-­‐-­‐coordinating on collectively demanding the low value-­‐-­‐, dangerous equilibria -­‐-­‐coordinating on collectively demanding the high value-­‐-­‐ and dreadful equilibria -­‐-­‐overshooting. We first show that when p=1 -­‐-­‐i.e. no uncertainty on the realized value of the resource-­‐-­‐ there is no uncertainty in coordination: agents focus on symmetric equilibrium in which the high value is collectively demanded. Next we run three treatments with, respectively, p=0.7, p=0.5 and p=0.3. While the rate of coordination on either (i) collectively demanding the high value or (ii) individually asking for the dangerous symmetric demand is negatively correlated to p, we find that regardless of risk preferences agents often attempt at coordinating on dangerous equilibria. But as p decreases, agents gradually shift away from dangerous equilibria but never manage to coordinate on playing cautious equilibria. Thus despite the payoff discontinuity, agents are willing to collectively take-­‐risk, and this regardless of their risk preferences. In addition, lowering the probability that the high value is realized generates noise and coordination failure: agents get stuck at non-­‐equilibrium demand profiles. Author(s): Olivier Bochet; Jeremy Laurent-­‐LUcchetti; Justin Leroux; Bernard Sinclair-­‐
Desgagne That's How We Roll (under high volatility) Ciril Bosch-­‐Rosa UCSC ABSTRACT: Short term debt has been at the center of the recent financial crisis. He and Xiong (2011) show that commonly used measure to control credit panics in the recent crisis might have exacerbated runs rather than mitigating them. Based on their theoretical paper, I construct a continuous time experiment in which subjects have to decide whether to rollover a credit or ‘run’ on it. If the credit is rolled, then the subject will be ‘locked in’ for a random amount of time in which there could be a credit run (by other subjects). He and Xiong show that this problem has a closed form unique threshold solution for the value of the firm, under which subjects should always run on the credit. My experiment will be testing policy related predictions of the model. The first one is that, under high volatility, subjects have a higher propensity to run if government bailouts are more reliable (but not certain). The second prediction is that under high volatility, if the policy maker decides to ‘artificially’ make credit contracts longer, then the probability of credit runs increases. Both these predictions run against our initial intuition, which is in fact correct in a low volatility environment. Note: pilots will start to be run in the following weeks, no results at this point. Author(s): Ciril Bosch-­‐Rosa The role of higher-­‐order beliefs in trust and coordination Steven Bosworth University of Pittsburgh -­‐ Department of Economics ABSTRACT: Coordination games with multiple Pareto-­‐ranked equilibria feature player beliefs as crucial in determining outcomes. If players are to coordinate on an efficient outcome, rational expected utility maximization requires a player to place high subjective probability on other players taking actions which support that equilibrium. This experiment induces second-­‐order beliefs; and tests the hypothesis that, just as common knowledge of rationality plays a role over and above the existence of rationality in Nash equilibrium, common knowledge of trust can drive selection of efficient but risky equilibria. Author(s): Steven Bosworth Do immaterial transfers induce reciprocity? An experiment on information provision and effort choice Leif Brandes University of Zurich ABSTRACT: In this paper, we test whether immaterial transfers induce reciprocity in a controlled laboratory experiment using an extended version of the dictator game. Immaterial transfers are implemented as provision of information: The receiver can decide whether or not to inform the dictator about her endowment level. The information is not action-­‐relevant in the sense of standard economic theory, but inter-­‐personal information provision is known in social-­‐
psychology to be a form of interactionally fair behavior. We design four treatments that differ in the dictator’s knowledge about her endowment level and the receiver’s ability to inform the dictator. We use the strategy method and elicit believes of both parties’ choices. Our results show that 67-­‐75% of all subjects in the role of a dictator respond strongly to their endowment level, the receiver’s information decision, or both. Comparing dictators’ choices between treatments, we find that the offer is significantly lower when the receiver decided not to inform the dictator, independent of the endowment level. Non-­‐information is also punished if dictators already know the level of endowment. The overwhelming majority of receivers anticipates the dictators’ behavior and makes use of their opportunity to inform. Even if the information is no news to the dictator 74% of receivers decide to incur costs of information, and 90% do so if the information is news to the dictator. Author(s): Leif Brandes; Donja Darai How effective are pay-­‐for-­‐performance incentives? An experimental comparison of incentive schemes for physicians Jeannette Brosig-­‐Koch University of Duisburg-­‐Essen ABSTRACT: Recent reforms in European and North American healthcare have introduced a variety of pay-­‐
for-­‐performance programs either combined with fee-­‐for-­‐service or capitation incentives. But how effective are these forms of payment for physicians? In this study we investigate the effects of introducing pay-­‐for-­‐performance with the help of controlled laboratory experiments. Treatment variations include four types of incentive schemes (fee-­‐for-­‐service, capitation, fee-­‐
for-­‐service combined with pay-­‐for-­‐performance, and capitation combined with pay-­‐for-­‐
performance) and nine types of patients (differing with regard to their illness and the severity of this illness). Our results reveal that patients significantly benefit from introducing pay-­‐for-­‐
performance, independent of whether it is combined with capitation or fee-­‐for-­‐service incentives. The magnitude of this effect is significantly influenced by the patient type, though. Moreover, comparing the decisions made by non-­‐medical students with that made by medical students reveals almost no significant differences. Current experiments investigate whether the latter result also holds for experienced physicians. Author(s): Jeannette Brosig-­‐Koch; Heike Hennig-­‐Schmidt; Nadja Kairies; Daniel Wiesen An Experimental Investigation of the Political Economy of Mixed Systems of Finance Neil Buckley York University / Economics ABSTRACT: Discrete versions of formal political economy models of mixed finance (Epple and Romano, 1996; Myers and Lülfesmann, 2011) are implemented in a controlled laboratory environment to test the theoretical predictions of how voting equilibria with respect to public financing differ depending on the configuration of public and private finance sectors. Three configurations or treatments are considered: public finance only, mixed financing with top-­‐up and mixed financing with opt-­‐out. In each decision round, individuals are assigned an exogenous endowment and make two types of decision: choice of proportional tax rate and the amount of the private good to purchase privately. The proportional tax rate used to finance the public provision of the private good is determined by the median tax rate of those selected by the voter’s. Preliminary results suggest some interesting implications for mixed systems of finance. First, higher income subjects continue to support the public system under mixed financing despite not having the financial incentive to do so. This runs counter to the theoretical predictions. Interestingly, lower income subjects’ support for the public system is lower than predicted in systems of mixed financing. Finally, there is no evidence of a treatment effect given income. Author(s): Neil Buckley; David Cameron; Katherine Cuff; Jeremiah Hurley; Stuart Mestelman Multiplayer Coordination and Competition in a Vertically Differentiated Market Klaus Burgmeier WHU -­‐ Otto Beisheim School of Management ABSTRACT: In a vertically differentiated market, the firms can offer products with different qualities. A producer of an established brand can leverage its reputational advantage by offering an inferior product with lower quality and price beside its premium brand. The so-­‐called fighter brand competes directly with the newcomers in the market. Based on the vertical differentiation model in Jost (2010), we use laboratory experiments to examine the effect of increasing the number of players in a market. In a market with only one incumbent and one entrant, the subjects tend to coordinate so that the profits of both firms are balanced. A post-­‐experimental questionnaire shows that inequity aversion and fairness concerns significantly influenced the subjects’ decision making. In the treatment with one incumbent and two entrants, the incumbents deviate from the equilibrium to drive the entrant’s profits to the minimum, even though the deviation results in less profit for the incumbent, too. The fairness concerns and inequity aversion completely vanish and cut-­‐throat competition emerges, possibly because of the difficulty to coordinate and reciprocate when more than two players are involved. Author(s): Klaus Burgmeier Social Learning or Social Preferences? Opening the Black Box of Peer Effects in Financial Decisions Leonardo Bursztyn UCLA Anderson ABSTRACT: Using a high-­‐stakes field experiment conducted in partnership with a large financial brokerage in Brazil, we disentangle the two channels through which one person’s financial decisions affect his peers’ decisions. When someone purchases an asset, his peers may want to purchase it as well, both because they learn from his choice (``social learning’) and because his possession of the asset affects others’ utility of owning the same asset (``social utility’). We randomize whether one member of a peer pair is allowed to possess an asset that he chose to purchase. Then, we randomize whether the second member of the pair 1) receives no information about the first member, or 2) is informed of the first member’s desire to own the asset and the result of the randomization that determined possession. This allows us to estimate the effects of (a) learning plus possession, and (b) learning alone, relative to a control group, allowing us to separately identify the effects of the ``social learning’ and ``social utility’ channels. We find that both channels are quantitatively important, and have statistically significant effects on the decision to purchase the asset. In the control (no information) group, 42.3% of individuals purchase the asset; this increases to 70.8% in the ``social learning only’ group; and, it increases to nearly 92.9% in the ``social learning and social utility’ group. These results provide evidence regarding two important classes of models of peers’ influence on individual and household choices in financial markets: both models emphasizing social learning and those emphasizing ``keeping up with the Joneses’ receive support. Author(s): Leonardo Bursztyn; Florian Ederer; Bruno Ferman; Noam Yuchtman Competitive bastards? -­‐ On the link between competition and cooperation Thomas Buser Amsterdam School of Economics ABSTRACT: Many selection and promotion mechanisms are based on tournament-­‐like competition. But often, the success of a project hinges not only on individual performance but also on the willingness of team-­‐members to cooperate. We aim to answer two questions. First, does being in a competitive environment make people less cooperative? And second, do competitive environments attract uncooperative people? Subjects are divided into groups of four and perform a simple task under piece-­‐rate and under tournament incentives, in randomised order. In each round, the task is interrupted and the subjects play a public good game with three randomly selected others. In a third round, players are given the choice between piece-­‐rate and tournament incentives. Preliminary results indicate that, contrary to our expectations, making subjects compete makes them more cooperative and that those who choose for competition are more cooperative. Author(s): Thomas Buser; Anna Dreber The role of intuition and reasoning in driving aversion to risk and ambiguity Jeffrey Butler Einaudi Institute for Economics and Finance ABSTRACT: Using a large sample of retail investors as well as experimental data we find that risk and ambiguity aversion are positively correlated. We show the common link is thinking mode: intuitive thinkers tolerate more risk and ambiguity than effortful reasoners. One interpretation is that intuitive thinking confers an advantage in risky or ambiguous situations. We present supporting lab and field evidence that intuitive thinkers outperform others in uncertain environments. Finally, we find that risk and ambiguity aversion vary with individual characteristics and wealth. The wealthy are less risk averse but more ambiguity averse, which has implications for financial puzzles. Author(s): Jeffrey V. Butler; Luig Guiso; Tullio Japelli Trust but Verify: An Indefinitely Repeated Trust Game with Stochastic Costs John Cadigan Gettysburg College ABSTRACT: We examine the impact of hidden information and verification on the evolution of subject decision making in an indefinitely repeated trust game. Our design features two innovations to the methodology presented in Engle-­‐Warnick and Slonim (2004). Trustees have an exogenously drawn return cost that is private information. Without (ex post) verification, the trustor cannot distinguish between a low cost trustee who could return and chose not to and a high cost trustee who could not return. We utilize a 2x2 design that varies the probability associated with the trustee being high cost and the ability of the trustor to verify costs. We restrict our attention to parameter configurations in which the efficient outcome can be sustained in equilibrium using a grim trigger strategy. Our preliminary findings indicate that the efficient outcome is more likely when the probability of a low cost trustee is high and costs can be verified. Author(s): John Cadigan; Brian Roberson; Malcolm Kass; Kusum Mundra; Kevin Siqueira Bidding Behavior in Pay-­‐to-­‐Bid Auctions Michael Caldara University of California, Irvine ABSTRACT: This paper experimentally studies the pay-­‐to-­‐bid auction format and compares average revenues in the discrete time simultaneous decision model to average revenues in the continuous time setting experienced in pay-­‐to-­‐bid auctions on the internet. For both of the group sizes studied, 3 and 5, there is no difference in the average revenues between the two environments. However, there is significant over-­‐bidding, as has been observed in pay-­‐to-­‐bid auctions on the internet, for both group sizes and this over-­‐bidding depends on the number of participants. Over-­‐bidding decreases with experience, and strategic sophistication plays a large role in the outcomes of individuals. Some of the least successful subjects cease auction participation all together, suggesting that the pay-­‐to-­‐bid auction mechanism can only sustain revenues above the value of the prizes as long as new inexperienced participants can be attracted. Author(s): Michael Caldara Bucket Auctions for Charity: A field trial Jeffrey Carpenter Middlebury College ABSTRACT: Recently there has been considerable interest in the use of auction mechanisms to fund public goods. Economists have developed theories that predict which of the standard mechanisms should do well and they have run a variety of lab (and a few field) experiments to test these mechanisms head-­‐to-­‐head. One aspect that has been largely overlooked, however, is whether new mechanisms can do even better. We run field charity auctions at the meetings of a well-­‐
known service organization in twelve states to examine the properties of five mechanisms: one that is common in the literature (first-­‐price all-­‐pay), two that are common in the field (the English and the raffle), and two that are new (the ‘bucket’ auction and our auction/lottery hybrid). Our results confirm previous lab results and demonstrate the potential of mechanisms design to transform auction-­‐based fundraising. Author(s): Jeffrey Carpenter; Jessica Holmes; Peter Hans Matthews I won't start my diet on Monday: The Effects of Increasing Awareness about Self-­‐Control Problems on Savings Leandro Carvalho RAND ABSTRACT: This paper investigates whether awareness about self-­‐control problems can affect consumption-­‐
saving decisions by analyzing the results of a randomized experiment designed to reduce individual’s naiveté©. Respondents of an Internet longitudinal study, the RAND American Life Panel, were randomly assigned to treatment and control groups. The treatment group was provided easy to understand information about self-­‐control problems, illustrated by the results of a study conducted by Read and van Leeuwen (1998). The control group was not provided such information. Survey participants completed an incentivized intertemporal choice task, the Convex-­‐Time Budget (CTB), proposed by Andreoni and Sprenger (2010). Respondents were asked to divide a $500 reward into two payments and told that the amount saved for the later payment would be paid with interest. They were asked to make six of these choices, in which we varied the interest rate and the mailing date of the first payment (the time delay between the mailing dates of the first and second payments was always one month). In order to ensure that participants made these decisions thoughtfully, a fraction of participants were selected at random to be paid one of their choices. We find that increasing awareness about self-­‐control problems has a large effect on consumption-­‐savings decisions. Respondents assigned to the treatment group choose to save more for later than respondents assigned to the control group. Author(s): Leandro Carvalho; Silvia Barcellos Mistakes and Game Form Recognition: Challenges to Theories of Revealed Preference and Framing Tim Cason Purdue University ABSTRACT: The Becker, DeGroot, Marschak (BDM) method of eliciting and measuring preferences (1964 Behavioral Science) is widely accepted as an incentive compatible mechanism that produces an accurate measure of preferences. Major theories of individual preferences, however, do not include the possibility of mistakes. This study reports a simple experiment eliciting preferences for induced-­‐value items, and finds that the BDM mechanism can be empirically unreliable due to susceptibility to subject misconceptions about the game form. Misconceptions are distinct from framing effects, as they decline through direct exposure to mistakes that reduce monetary earnings. The data reveal different types of misconceptions through specific patterns of behavior, such as sensitivity to the random price range consistent with a misconception that the BDM is a first-­‐price auction mechanism. We argue that unless misconceptions are accounted for, the theory of framing can lead to incorrect interpretations of data produced by the BDM. Author(s): Timothy Cason; Charles Plott Risk Taking and Risk Sharing -­‐ Does Accountability Matter? Elena Cettolin Maastricht University ABSTRACT: Risk sharing arrangements diminish individual’s financial vulnerability associated to undesirable events. We experimentally investigate how the risk taking behavior of others and own risk preferences affect an individual’s willingness to engage in risk sharing. First, each subject is asked to choose between two lotteries that vary in their riskiness. Pairs are then formed and one individual in each pair is asked to choose a risk sharing level before the actual outcomes of the lotteries are determined. The chosen level determines which percentage of the lottery outcomes is deposited in a common account, which is then equally divided. We compare risk sharing behavior in this setting to a treatment where lotteries are randomly assigned to subjects, so that any responsibility for risk taking is excluded. We find that, irrespective of their own risk attitude, individuals are less willing to share risks with someone facing the high risk lottery as compared to a low risk taker. Further, the more undesirable the low outcome in case of high risk taking the more individuals differentiate their sharing behavior. However, this is only observed when high risk exposure is deliberate, that is in situations where subjects can be held accountable for taking risks. Our results suggest that the support for risk sharing arrangements by a large majority may demand measures that increase the contributions for deliberate high risk takers. Author(s): Elena Cettolin; Franziska Tausch Reference Point and Risk Attitude: An Experimental Study Hong Chao Shanghai Jiao Tong University ABSTRACT: Previous researches about the relationship between reference point and risk attitude have produced mixed results. Some studies find that prior gains induce decision makers to engage in greater risk averse while prior losses induce greater subsequent risk seeking, others report an opposite result. Instead of studying the role of prior gain or loss as a solo reference point, this paper investigates the importance of relative economic status toward one’s risk attitude as we allow it to mingle with the effect of prior gain or loss in an experiment. We find that when subjects suffer a larger prior loss than their peers, they tend to engage in significantly greater subsequent risk seeking than when they suffer a smaller prior loss compared to their peers. However, the relative economic status does not generate an effect when subjects experience a prior gain. We estimate the reference points based on which subjects change risk attitude under various conditions of prior outcome. These results may complement foregoing studies on prospect theory and have the potential to improve our understanding about people’s varying risk attitude under different situations. Author(s): Hong Chao; Xiangdong Qin; Zexi Lu Incentives, Information and Psychological Competition in a Real Effort Task Ananish Chaudhuri University of Auckland ABSTRACT: There is a large literature looking at the impact of various incentive schemes in general and tournaments in particular on performance. The basic insight coming out of this literature is that tournaments in general improve performance. We explore aspects of these questions using a multiple cue probabilistic learning task requiring real effort. We compare the impact of extrinsic incentive schemes, where pay is explicitly tied to performance such as piece rates and tournaments, with intrinsic schemes, where players earn a flat salary regardless of performance. In order to separate out the role of psychological competition as opposed to the incentive effects, within our tournament treatments we vary the information that our subjects receive where in some treatment they are paid using a winner-­‐take-­‐all tournament scheme and learn about their relative standing while in others they are unaware of how they are performing vis-­‐a -­‐
vis a pair member. We also use psychological questionnaires to control for trait anxiety and intrinsic motivation. We find that after controlling for levels of anxiety during the task, overall we get better performance (as measured by lower forecasting errors) in the piece-­‐rate and salary treatments, where participants face less competitive pressure. We also find that women perform worse than men in all treatments except for salary. We vary the degree of task difficulty and find interesting differences in performance under various schemes. Our results have implications for the design of incentive schemes and suggest that we need to take better account of intrinsic motivation. Author(s): Paul Brown; Linda Cameron; Ananish Chaudhuri; Tony So The Wages of Pay Cuts Daniel Chen Duke Law School ABSTRACT: To understand why firms rarely cut nominal wages, we hired workers for a data entry task, paid them a high wage and then offered some of the workers the opportunity to keep working, albeit for a lower wage. We framed the new wage offer in different ways across treatment groups. Workers were more likely to reject lower offers, but ‘reasonable’ justifications largely eliminated this effect. Not all justifications were effective-­‐-­‐justifying the cut on the ground that it would increase our profits actually increased quits. We also measured whether the treatments affected quality, trust and cooperation. The ‘profits’ treatment reduced cooperation and possibly reduced quality; the other treatments had generally weak or nonexistent effects. Author(s): Daniel L. Chen; John J. Horton Broken Promises and Hidden Partnerships Jingnan Chen ICES, GMU ABSTRACT: Previous research shows unenforceable promises can promote both trust and reciprocity. While this can benefit existing exchange, in dynamic environments promises might hinder one’s willingness to form new trust relationships, even when doing so would be efficient. This arises in both business and social contexts, including industry non-­‐compete agreements and personal relationship commitment decisions. Despite its practical importance, we are aware of no previous experimental research that informs behavior when facing tension between keeping a previous promise and taking advantage of a new trust opportunity. Here we take a step in that direction. We report data from an experiment using three person trust games with a hidden trust action. In our game, the second mover reciprocates the first mover’s trust by not trusting the third mover, and players can make non-­‐binding promises to each other prior to making decisions. We find people are willing to break promises to pursue new trust relationships, but only when the party involved in the new trust makes a promise to cooperate. When promises encouraging new trust are not made, we find that people are overwhelmingly likely to keep old promises and avoid entering new and potentially beneficial hidden trust relationships. Our data provide novel evidence on the role of promises in affecting economic outcomes, and shed new light on the contexts under which promises are likely to be broken. Author(s): Jingnan (Cecilia) Chen; Daniel Houser Discounting the risky future Stephen Cheung The University of Sydney ABSTRACT: The relationship between risk and time preferences has been a lively topic of recent research in experimental economics. Andersen et al (2008) demonstrate procedures for joint estimation of risk and time preferences under the assumption that the same utility function applies to both. However, Andreoni and Sprenger (in press) argue that two distinct utility functions govern choices under risk and certainty. If this claim is correct, the Andersen procedure will result in incorrect inferences. On the other hand, the inter-­‐temporal portfolio allocation instrument of Andreoni and Sprenger may itself be confounded in risky settings if subjects perceive an opportunity for diversification. I will present the findings of two studies intended specifically to evaluate the validity of these two procedures, and to disentangle their implications for the estimation of time preferences. Author(s): Stephen L. Cheung Gift exchange with migrant workers and students in China Thorsten Chmura Shanghai Jiao Tong University ABSTRACT: In this experiment we investigate the determinants of wage discrimination. Therefore, we conducted two gift-­‐exchange experiments in Shanghai and Ningbo (both China) involving students and migrant workers. Students act in the role of employers and decide about the wage of the migrant workers, while the migrant workers choose their effort level given the actual wage. The employers can condition the wages only on the workers’ origin. The workers’ home province is the only criteria to differentiate between the migrant workers. We observe a high variance in wages across provinces. Combining our lab results with real world data, we can investigate the determinants of wage discrimination. In our setting wages are positively influenced by the economic power of the workers’ home provinces, and negatively influenced by the geographical distance between the employers’ and workers’ home provinces and the ethnic heterogeneity in workers’ provinces. The same pattern is observed within a pure student sample. Author(s): Thorsten Chmura; Sebastian Goerg Financing and Signaling Under Asymmetric Information Angelina Christie The Catholic University of America ABSTRACT: This paper presents an experimental investigation of a financing-­‐investment environment under asymmetric information in which the finance-­‐seeking firms know the value of the project with certainty while the investors do not. In particular, it examines experimentally the underpricing-­‐
signaling hypothesis. Importantly, the paper tests and compares this hypothesis under the two institutions for financing offers that are commonly observed in corporate financial markets: take-­‐it-­‐or-­‐leave-­‐it (fixed-­‐price) offer and the competitive bidding offer. The results of this experimental study suggest that underpricing can serve as a credible signal of quality under certain parameters. Not surprisingly the underpricing is lower under the competitive bidding institution than under the fixed-­‐price offer. Using experimental methods, this paper empirically contributes to the role of underpricing and the choice of institution in equity offering and similar-­‐type environments in financial markets under informational asymmetry. Author(s): Angelina Christie; Daniel Houser Can Financial Incentives Induce Unequal Parental Investment in Children? Evidence from a Field Experiment Amanda Chuan University of Chicago ABSTRACT: The economic literature documents a variety of financial reward interventions that aim to improve the academic outcomes of targeted children. However, few studies have examined the potential spillover effects on non-­‐targeted siblings. We report on survey results from a field experiment in a low-­‐income Chicago-­‐area community where 206 families were randomly selected to receive financial rewards conditional on their preschool-­‐aged child’s academic performance; in one treatment, financial rewards were given in cash, while in another deposits were made into an account accessible in the future. We find that parental investment in targeted children was not significantly different between the treatment groups, but parental investment in non-­‐targeted siblings was substantially lower in the treatment that rewarded parents with cash. We develop a theoretical framework based on Becker’s quantity-­‐quality fertility model that explains this re-­‐allocation of household resources away from non-­‐targeted siblings. These results suggest the importance of considering spillover effects of interventions targeting only one child. Author(s): Amanda Chuan; John List; Anya Savikhin Dishonesty and Charitable Behavior Doru Cojoc University of Iowa ABSTRACT: We examine in the laboratory how having the opportunity to behave prosocially in the future affects the likelihood of currently engaging in antisocial behavior. Subjects first complete a task which determines their payoff. They then self-­‐report their performance, which provides them with an opportunity for undetected cheating. In the second stage they are given the opportunity to donate some of the money earned in the first stage to a charity. Only subjects in the treatment group know about the opportunity to donate in the second stage. We find that subjects cheat more if they know they can donate some of the money to charity. Surprisingly, they also end up donating less to charity. The first result is consistent with charitable giving alleviating guilt. The second result can be explained by models of identity, time-­‐inconsistent preferences, or self-­‐signaling. Author(s): Doru Cojoc; Adrian Stoian Bubbles in Experimental Markets with Purely Speculative Assets Sean Collins Fordham University ABSTRACT: In a seminal study on the subject, Smith, Suchanek, and Williams (1988) find that that prices exceeding intrinsic value are a common and reproducible phenomena in laboratory experiments. Recent work by Lei and Vesely (2009) indicates that bubble-­‐alleviating common expectations can be formed without experience in the market when subjects develop a full understanding of the structure of the dividend. Giusti, Jiang, and Xu (2012) report that bubbles can be eliminated by paying a sufficient return on cash, increasing the opportunity cost of buying shares and reducing the ‘active participation’ problem raised in Lei, Noussair and Plott (2001) This study furthers these lines of investigation. Results from a laboratory asset market experiment are considered in which securities have no intrinsic value (i.e. no dividends and zero terminal value). In this framework, dividends provide neither liquidity nor a challenge to backward induction. A return on cash is then introduced to study the ‘active participation’ problem in isolation. Implications for real-­‐world securities markets without internal rates of return (i.e. precious metals) are also considered. Author(s): Sean M. Collins Centralized vs. Decentralized Management: An Experimental Study David Cooper Florida State University ABSTRACT: We study a central question of firm organization: should the divisions of a firm be centrally managed or should divisions be allowed to make decisions independently. The divisions can gain by coordinating their activities but each division has private information about the state of the world and, critically, the interests of the two divisions are not perfectly aligned. Central management reduces the coordination problem facing the divisions, but gives divisions an incentive to distort the information they report to the central manager. Theory predicts that these informational problems are sufficiently severe that decentralization will outperform centralization, but this relies on strong assumptions about the ability of divisions to solve their coordination problem. We present experimental evidence showing that the coordination problems under decentralization are more severe than the informational problems under centralization, leading to better performance with centralization. Author(s): Jordi Brandts; David Cooper Favorable Selection and Excludable Public Goods with Interdependent Values Caleb Cox The Ohio State University ABSTRACT: In this paper I study a simple model of binary excludable public goods with interdependent values. Previous research on excludable and non-­‐excludable public goods has largely focused on private values or complete information. Allowing for interdependent values (including the special case of pure common value) leads to a potentially important selection effect. Under a simple threshold or provision point mechanism, provision of the public good in this environment is good news about its value in Bayesian Nash equilibrium. Because of this favorable selection effect, naive agents who condition their expectations only on their own private information will contribute too little. I experimentally examine whether such a ‘curse’ exists in this environment by testing Bayesian Nash equilibrium against naive strategies. I also consider a similar ‘anti-­‐
threshold’ treatment exhibiting adverse selection and an uncertain private values treatment with no selection effect to examine how subjects respond to and learn to account for favorable and adverse selection. Preliminary results suggest that neglect of the favorable selection effect does lead to under-­‐contribution. However, there is also some evidence that subjects learn to at least partially account for the favorable selection effect over multiple rounds of play with full feedback. Author(s): Caleb A. Cox What to do without Cournot or Bertrand? An experimental study of mixed strategy equilibria in simultaneous price-­‐quantity games Daniel Cracau University of Magdeburg ABSTRACT: We experimentally study a Bertrand duopoly market where each competitor is modeled as a three-­‐player team. Each player states a bid and the team with the lower total bid wins the competition. There are two incentive structures by which profits are divided among members of the winning team: each member receives (1) her own asking price or (2) the average asking price in the team. In different experimental conditions the market is either homogeneous (both teams have the same incentive structure) or heterogeneous, and information about the incentive structure of the opposing team is either public or private. Our results show that heterogeneity notably reduces the ability of market participants to tacitly collude, contradicting predictions derived from game theory and reinforcement learning, but in line with the legal intuition of the current EU horizontal merger guidelines. Additionally, we find that when information about the incentive structure of the opposing team is public (rather than private) collusion is more likely in homogeneous markets, but less likely in heterogeneous markets. Author(s): Daniel Cracau; Benjamin Franz Intermittent Sanctions and Cooperation in a Public Good Game Zhixin Dai GATE-­‐CNRS and University of Lyon 2 ABSTRACT: That punishment plays a crucial role in sustaining cooperation in linear public good games is experimentally well documented as well as its detrimental effect on efficiency. We show that using an exogenous intermittent sanctioning schedule can improve cooperation and efficiency in the long run. Precisely, we manipulate both the frequency and randomness of punishment in a 50-­‐period public good game. We compare a continuous reinforcement schedule in which audits and sanctions of deviations from the average contribution are implemented every period with a random reinforcement schedule in which higher sanctions are implemented on average every three periods. A regime shift occurs after 22 periods such that no audit and sanction are ever implemented. We recreate a highly ambiguous environment in that subjects are not informed on the exact probability of an audit of their contribution and they have therefore to form beliefs (that we elicit) on the risk of being audited and possibly sanctioned from their past experience. While we find no difference in cooperation across treatments before the regime shift, after the regime shift intermittent sanctions leads to a much longer persistence of cooperation than the continuous reinforcement schedule. Additional manipulations of the frequency of audits and sanctions allows us to identify the minimum frequency for which the highest level of cooperation is sustained in the long run with a minimum cost of sanctions. Author(s): Zhixin Dai; Robin Hogarth; Marie Claire Testing the CNBC Effect: Behavioral Impact of Public Information Donald Dale Muhlenberg College ABSTRACT: We report on the results of laboratory experiments designed to study the ‘CNBC Effect’-­‐-­‐-­‐
situations where the addition of public information leads to worse decision making. Payoffs for decisions depend on two factors: being right (i.e. matching an underlying, but unknown state variable) and coordinating with others. In our baseline treatment, each individual privately receives an informative signal about the state. We then add a second, lower quality, private signal and, not surprisingly, decisions improve. Finally, we make this lower quality signal public. In that treatment, (a) subjects place greater weight on the low quality signal and (b) losses increase by 12% compared to the baseline. In addition to the strategic effect, losses are also higher than expected owing to random variation in choices. JEL #s: C72, C92 Keywords: public information, CNBC effect, coordination Author(s): Donald Dale; John Morgan Does the Unemployment Benefit Program affect Workers Productivity? Evidence from a Field Experiment in Colombia Patricio Dalton Tilburg University ABSTRACT: Unemployment benefit programs are key policy instruments to cushion the negative impact of unemployment. The most implemented schemes are welfare and workfare programs. While the former provides unconditional monetary transfers to the unemployed, the latter makes such transfers conditional on the unemployed spending some time on mandatory activities (e.g. community or administrative work). In spite of their broad use, the effects of these programs on labor supply and workers’ productivity are yet unknown. Economic theory predicts that workers productivity under workfare is higher than under welfare. We designed a field experiment to test this prediction. We hired more than 200 research assistants (RAs) for a one-­‐
month period to code news on politicians from newspapers archives. At the beginning of the job, workers were informed that the demand for RAs was not fixed, and that some days there wouldn’t be enough vacancies for everyone. Those with lower productivity would become temporarily unemployed. Unemployed RAs were randomly assigned to either a welfare treatment or workfare treatment. Those under welfare treatment would receive 30% of their daily salary for each day they were unemployed, and those under workfare treatment would receive the same amount of money conditional on spending some time doing administrative tasks. Contrary to standard economics predictions, we found that workers under the welfare treatment performed significantly better than those under workfare. One possible explanation being that people derive high enough intrinsic benefits from working while they are unemployed making the outside option under workfare higher than under welfare. Author(s): Mariana Blanco; Patricio S. Dalton The dark side of team identity: Experimental evidence from financial service professionals Anastasia Danilov University of Cologne ABSTRACT: In an experiment with professionals from the financial services sector, we investigate the impact of team bonus payments and team identity on deception in the recommendation of investment products. Advisors (bankers) had to recommend one of six products to customers (students). The products were ranked according to their expected returns and risk such that the best products were associated with the lowest commission rates. The experimental results reveal that under team incentives advisors recommend bad products significantly more often than under individual incentives but only, when team identity is strong. Author(s): Anastasia Danilov; Torsten Biemann; Thorn Kring; Dirk Sliwka Social networks in a labor market with adverse selection: An experiment Aurelie Dariel maastricht university ABSTRACT: Social networks are widely used by firms to obtain information about potential employees. We experimentally study how social networks affect outcomes in labor markets with adverse selection. In the experiment, workers are either of high or low ability. Workers who belong in the same network have a higher probability of being of the same type. The experimental results indicate that firms utilize social networks to overcome the problem of adverse selection. Consequently, a higher fraction of high-­‐ability workers is hired in a treatment with social networks relative to a baseline treatment where there are no social networks. High-­‐ability workers receive higher wages than low-­‐ability workers. Author(s): Dariel Aurelie; Arno Riedl Letting the Briber Go Free: An Experiment on Mitigating Harassment Bribes Utteeyo Dasgupta Franklin and Marshall ABSTRACT: We investigate the proposition laid out in Basu(2011) that bribe-­‐giving behavior can be brought down if bribe-­‐givers are given legal immunity, and all punishments are lumped on the bribe-­‐
taker. We first build a simple model to incorporate Basu’s proposition, as well as follow-­‐up comments by other social scientists. Second we test them using experiments. Our results suggest that Basu’s policy moves behavior in the right direction. However, some of the pitfalls identified by the social scientists do have an effect of dousing some of the positive effects of the policy. Author(s): Klaus Abbink; UtteeyoDasgupta; LataGangadharan; Tarun Jain Social Identity, Coordination, and Discrimination: An Experiment Vessela Daskalova Queen Mary, University of London ABSTRACT: An important question is whether the social identities of decision-­‐makers bias their decisions. We present an experimental study examining the effect of social identity on individual and joint hiring decisions about job candidates. In control sessions we investigate whether people make different hiring decisions when they decide individually compared to when they try to coordinate their decision with a second decision-­‐maker. In treatment sessions we add a pre-­‐
decision-­‐making stage, in which we artificially induce group identity for both decision-­‐makers and candidates. We are interested in: 1) is there any favouritism towards own group and discrimination against other group candidates when people make individual decisions; 2) do people under the presence of group identities decide differently when trying to coordinate with a second person than when making decisions alone; 3) how the group identities of the decision-­‐
makers and the project candidate affect decisions in the joint decision-­‐making case. We find that in some cases when decisions are individual, decision-­‐makers favour candidates of own group compared to candidates of the other group. There is a substantial and significant increase in favouritism towards own group candidates when decisions are made with a second person. There are substantial and significant differences in decisions about own and other group candidates depending on the identity of the second decision-­‐maker. These results suggest that who makes a decision with whom about whom might matter. Author(s): Vessela Daskalova Sequential versus Simultaneous Election Contests: An Experimental Study Shakun Datta Mago University of Richmond ABSTRACT: This experimental study compares sequential and simultaneous multi-­‐battle election contests. Consistent with the theory, we find substantial evidence of a ‘New Hampshire effect’ in the sequential contests, i.e. the winner of the first battle wins the overall contest with much higher probability than the loser of the first battle. However, contrary to the theory, sequential contest generate substantially higher expenditure than simultaneous contest. This is mainly because in the sequential contest losers of the first battle do not decrease their expenditure in the second battle; and winners of the first battle substantially increase their expenditure in the second battle, instead of decreasing their expenditure as predicted. Finally, we find that although subjects learn to behave more in line with equilibrium predictions, even in the last periods of the experiment, their behavior is substantially different from predictions. Author(s): Zeynep B. Irfanoglu; Shakun D. Mago; Roman M. Sheremeta Willpower Depletion and Framing effects Thomas de Haan Friedrich Schiller UniversitÀt Jena ABSTRACT: We investigate whether depleting people’s cognitive resources (or `willpower’) affects the degree to which they are susceptible to framing effects. Recent research in social psychology and economics has suggested that willpower is a resource that can be temporarily depleted and that a depleted level of willpower is associated with self-­‐control problems in a variety of contexts. In this study, we extend the willpower depletion paradigm to framing effects and argue that willpower depletion should increase the power of framing effects. For this purpose we run an experiment in which we deplete participants’ willpower and subsequently have them take part in a series of framing tasks, including a framed prisoner’s dilemma, an attraction effect task, a compromise effect task and an anchoring task. However, we found no evidence that framing effects are indeed more prevalent in willpower depleted participants than in controls. Abstract Keywords: willpower, ego depletion, framing, willpower depletion, experiment, behavioral economics. Abstract JEL codes: D81, C91 Author(s): Thomas de Haan; Roel van Veldhuizen Social Behavior Toward Risk and Inequality Gregory DeAngelo Rensselaer Polytechnic Institute ABSTRACT: Subjects are placed in social and individual decision situations with an obvious unchanging reference point. We observe whether subjects risk behavior changes on either side of the reference point as prospect theory suggests. We also observe if social and private decisions are similar allowing us to evaluate if there is inequality aversion. Finally we observe redistribution behavior when positions are known and fixed allowing an assessment if the observed redistribution reflects risk preferences, inequality preferences or simple self interest. Our primary result is that subjects are heterogeneous in their response to risk with a majority exhibiting risk preference either above or below the reference point. Men are more prone to take risks at the low end of the distribution but are unwilling to force this preference on others in social situations. We do not observe a separate preference for equality. With positions known, self interest predicts well at the low end of the distribution but not at the high end. Author(s): Steven Beckman; Gregory DeAngelo; W. Jim Smith; Buhong Zheng Why We Can't Pay for Express Checkout: An Experiment on Queue Price Discrimination Cary Deck University of Arkansas ABSTRACT: We model competition between two firms selling identical goods to consumers that arrive in the market over time and choose which firm to patronize based on both price and the costs of waiting to be served. One firm controls two queues and the other controls only one. We explore the welfare impact of allowing the two-­‐queue firm to charge a different price at each queue it controls, thereby permitting price discrimination over individuals with different waiting costs. We search for equilibrium prices computationally, and then we test the model with a laboratory experiment to observe whether the predictions hold. Consistent with our model, we find that price discrimination is harmful to firms and beneficial to consumers. When the two-­‐queue firm charges two different prices, the one-­‐queue firm competes only with the lower price, and this drives down prices and profits while increasing consumer surplus. Author(s): Cary Deck; Erik Kimbrough Willingness to Wait under Risk and Ambiguity: Theory and Experiment Marco Della Seta Radboud University Nijmegen ABSTRACT: Waiting is an important feature of economic decisions. In numerous situations, individuals and firms can choose between acting immediately and waiting for more favorable conditions. In a canonical example, a firm holds an opportunity to invest and waits, observing market development, to decide on investment timing. Recognition that an opportunity to delay an irreversible decision in an uncertain environment is akin to an option led to development of real options theory. The original application of real options is capital investment but the approach has been used to model a variety of economic and non-­‐economic problems. This paper studies theoretically and experimentally the distinctive effects of risk and ambiguity on willingness to wait. We analyze a simple real options model in which a decision maker chooses the timing of investment and can observe the environment to learn about it. The model predicts that higher risk delays investment. Higher ambiguity accelerates investment if decision makers are ambiguity averse, while a delay of investment indicates an ambiguity seeking attitude. Experimental data reveal that willingness to wait increases in both risk and ambiguity. This confirms the predicted effect of risk and implies that subjects are ambiguity seeking. The last finding is in contrast with aversion towards ambiguity often found in static willingness-­‐to-­‐pay experiments. Author(s): Marco Della Seta; Sebastian Gryglewicz; Peter Kort Networks in School Choice: An Experiment Study Tingting Ding NYU ABSTRACT: School matching problem has been one of the most important applications of mechanism design and matching theory in the last decades. In this paper we experimentally investigate the role of network in school matching problem, one aspect that has been ignored in the previous literature but one, we think, is empirically important. In the laboratory the subjects are connected in various network architectures and allowed to communicate with others in the same sub-­‐network. We examine the Boston and the Gale-­‐Shapley mechanism, both of which are used in the real world. Our main results show that (a) Communications do change the behavior of subjects. In the Gale-­‐Shapley mechanism, communications lead to more truth-­‐telling. (b) When schools have weak orderings the efficiency of the Gale-­‐Shapley mechanism may be inferior to that of the Boston mechanism, and hence the increase in truth-­‐telling after communications in the Gale-­‐Shapley mechanism makes the welfare of subjects decrease. (c) After communication the subjects without priority rights are better off in the Boston mechanism while worse off in the Gale-­‐Shapley mechanism. These results have important implication for the recent school choice reforms. Author(s): Tingting Ding; Andy Schotter Simultaneous Ultimata: Responder Induced Competition Daniel Dittmer European University Viadrina Frankfurt (Oder) ABSTRACT: The standard ultimatum game is expanded by doubling the number of proposers with responders playing two ultimata simultaneously. Full answering behavior for the standard and expanded game is measured with the strategy method. Moreover, the dynamics in this sort of contribution game is investigated for twenty repetitions. We find that proposers as well as responders choices strongly depend on the other available offer. More precisely, responders reject offers more frequently when they have received a higher offer in parallel (even when it is above 50 percent of the stake). This effect is tremendous when there is repeated interaction. Moreover, proposers’ offers, when they act sequentially and offers are perfectly observable, strongly depend on the other proposer’s offer. Moreover, proposer often overbid the observed offer in anticipation of the responder’s strategy. Together these results highlight the importance of other ultimata as reference points for proposals as well as responses. Author(s): Daniel Dittmer; Philipp E. Otto Group Identity and Leading-­‐by-­‐Example Michalis Drouvelis University of Birmingham ABSTRACT: We study the interplay between leading-­‐by-­‐example and group identity in a three-­‐person sequential public goods game experiment. A common identity between the leader and her two followers is beneficial for cooperation: average contributions are more than 30% higher than in a benchmark treatment where no identity was induced. In two further treatments we study the effects of heterogeneous identities. We find no effect on cooperation when only one of the followers shares the leader’s identity, or when followers share a common identity that differs from that of the leader. We conclude that group identity is an effective but fragile instrument to promote cooperation. Author(s): Michalis Drouvelis; Daniele Nosenzo Learning, Forecasting and Optimizing: An Experimental Study John Duffy University of Pittsburgh ABSTRACT: Rational Expectations (RE) models have two crucial dimensions: 1) agents correctly forecast future prices given all available information, and 2) given expectations, agents solve optimization problems and these solutions in turn determine actual price realizations. Experimental testing of such models typically focuses on only one of these two dimensions. In this paper we consider both forecasting and optimization decisions in an experimental cobweb economy. We report results from four experimental treatments: 1) subjects form forecasts only, 2) subjects determine quantity only (solve an optimization problem), 3) they do both and 4) they are paired in teams and one member is assigned the forecasting role while the other is assigned the optimization task. All treatments converge to Rational Expectation Equilibrium (REE), but at very different speeds. We observe that performance is the best in treatment 1) and worst in the treatment 3). We further find that most subjects use adaptive rules to forecast prices. Given a price forecast, subjects are less likely to make conditionally optimal production decisions in treatment 3) where the forecast is made by themselves, than in treatment 4) where the forecast is made by the other member of their team, which suggests that ‘two heads are better than one’ in learning a REE. Author(s): Te Bao; John Duffy; Cars Hommes When Does Cheap-­‐Talk (Fail to) Increase Efficient Coordination? Subhasish Dugar THE UNIVERSITY OF CALGARY ABSTRACT: This paper experimentally analyzes the effects of cheap-­‐talk on the levels of efficiency in Stag-­‐
Hunt coordination games. The novelty of our paper is that we are the first to systematically control for the key equilibrium-­‐selection principles (payoff-­‐ and risk-­‐dominance) while analyzing the effects of cheap-­‐talk. Our major conclusion is that cheap-­‐talk may not necessarily increase efficiency, as shown previously. For a specific combination of values of the equilibrium-­‐selection measures, cheap-­‐talk is largely untruthful and is detrimental to efficiency. A level-­‐k model of beliefs gracefully explains the behavior and econometric estimations confirm the superiority of our model over an alternative model. Author(s): Subhasish Dugar; Quazi Shahriar Decisions and beliefs on the insurance market in presence of risk, imprecision and conflict Theodora Dupont-­‐Courtade Paris School of Economics ABSTRACT: This experiment investigates behaviors under risk, imprecision and conflict of insurance demand and insurance supply. Subjects reveal their preferences for low-­‐probability-­‐high-­‐consequence and large-­‐probability-­‐small-­‐consequence events. They face four information types about the occurrence probability according to expert’s positions. Under risk (R), the experts agree that the probability is well-­‐known and precise. Under imprecision (I), they agree that it takes a range of possible values, but they cannot narrow down the estimation. Under conflict (C), they disagree and each has its own precise estimate. Finally, under imprecision within conflict (IC), they disagree and each has its own but imprecise estimate. We vary the probabilities, holding the expected value constant, which generates six events by information type. The subjects come twice and play two experiments relative to both market sides: a demand experiment in which they choose their coverage against possible losses and a supply one in which they accept or not some coverage contracts. BDM mechanisms are used to elicit WTP/WTA, quantities of contracts accepted for different prices and subjective probabilities of events realization. This design enables to obtain complete information about insurance choices (prices, quantities and probabilities). Contrary to standard insurance experiments, choices are not hypothetical; subjects are paid according to their overall performance by contract type relative to a benchmark based on expected-­‐value minimizer’s behaviors. Preliminary results show that demand and supply do not meet for low-­‐probability events. Furthermore, a rank of preferences is obtained over the contract types (R>I>C>IC), that appears to be stronger for buyers than for sellers. Author(s): Theodora Dupont-­‐Courtade; Sebastien Massoni Believing and Achieving: the effect of confidence on performance David Eil UCSD ABSTRACT: Many recent studies have documented the existence of overconfidence and sought to understand what kind of belief updating behavior produces it. Theoretical models have taken as given some kind of benefit to more optimistic beliefs and derived an optimal level of overconfidence. Showing this assumption to be true empirically using field data is difficult as beliefs are correlated with ability, which affects performance directly. We exogenously vary beliefs in a laboratory setting and show that these beliefs do impact performance on a math test. This suggests that confidence does in fact have a positive impact on performance. Author(s): David Eil; Justin Rao The Value of Voice Yohanes Eko Riyanto Nanyang Technological University ABSTRACT: This paper presents an experimental analysis on the role of voice and its impacts on behavioral attitudes in an ultimatum game setting. We elicited responder’s willingness to pay (WTP) for the opportunity to send a message before the proposer makes decision using the Becker-­‐DeGroot-­‐
Marschak (BDM) mechanism. We ran two experimental treatments. In one treatment, responders were given the opportunity to pay to send a message to their matched proposer, and in another treatment, responders were given the opportunity to pay to send a message to an inconsequential third party. We find that 82% of responders in the former treatment were willing to pay for the opportunity to voice to their matched proposer, and 62% of the responders in the latter treatment were willing to pay for the opportunity to voice to an inconsequential third party. This result shows that not only voice has inherent value, it also has strategic value. Further, we show that voicing to the matched proposer increases responder’s expectations of offers regardless of their valuation for voice. On the contrary, voicing to an inconsequential experimenter may or may not increase responder’s expectations of offers depending on their valuation for voice. Finally, we also find that sending a message to the proposers make responders believe that proposers would take their opinions into consideration, resulting in lower responder’s willingness to accept low offers. In contrast, sending a message to an inconsequential experimenter makes responders more willing to accept low offers instead. Author(s): Qiyan Ong; Yohanes E Riyanto; Steve Sheffrin; Walter E. Theseira Consumer noncompliance with intellectual property rights Anouar El Haji Amsterdam Business School ABSTRACT: A possible explanation of why intellectual property rights are often violated is because intellectual property is nonrivalrous. To explore to what extent humans differentiate between rivalrous and nonrivalrous goods, two experiments with U.S. citizens and Indians are conducted for this study. In the first experiment participants could choose how much to steal or copy depending on the treatment. In the second experiment a novel exchange game is employed, in which economic efficiency in the nonrivalrous treatment does not differ from that in the rivalrous treatment. Participants behaviorally distinguished between rivalrous and nonrivalrous goods, where the property rights of the latter were more likely to be violated. However, a higher economic efficiency in favor of the violator, as in the first experiment and in the field, may make the effect more pronounced. In the second experiment noncompliance with intellectual property rights among U.S. citizens decreased with age while among Indians it was significantly more likely to happen than theft after controlling for age. These findings indicate that noncompliance with intellectual property rights can be internalized but it is unclear what cultural forces, which are present in the U.S. but apparently not in India, are responsible for this effect. Author(s): Anouar El Haji; Mark A.A.M. Leenders Strategic commitment and cooperation in experimental games of strategic complements and substitutes Matthew Embrey Maastricht University ABSTRACT: We study the impact of strategic commitment on cooperation in repeated games with strategic substitutes (Cournot) and complements (Bertrand) using laboratory experiments. The Bertrand and Cournot stage-­‐games, in which duopolists have four available actions, have the same diagonal payoffs, including the Nash and collusion outcomes, and the same deviation and sucker payoffs. The games primarily differ in the location of the deviation action: for Cournot it is sub-­‐
Nash, for Bertrand it is between the collusion and Nash actions. At the beginning of a super-­‐
game, subjects program a dynamic response that specifies a recommended action in response to their rival’s previous choice. There are four strategic commitment settings: strong, bilateral, unilateral and hot. In all variations subjects can deviate from their recommendation for a modest cost, except the hot where it is for free. In the unilateral, dynamic responses can be modified for a small cost; in the bilateral such modifications also require your rival’s consent. For all treatments collusion tends to increase over super-­‐games, except the bilateral-­‐Cournot treatment. Overall, strategic commitment has no effect on cooperation in Cournot and a negative one in Bertrand. We find more cooperation in Bertrand than Cournot in the absence of strong commitment. However, when subjects are more committed to initial plans, a higher level of cooperation is achieved in Cournot. We relate our findings to risk-­‐dominance and renegotiation considerations. JEL classification codes: C73, C92, D43 Keywords: strategic commitment, cooperation, repeated games, complements vs. substitutes Author(s): Matthew Embrey; Friederike Mengel; Ronald Peeters Choosing how to choose: efficiency concerns and constitutional choice Dirk Engelmann University of Mannheim, Department of Economics ABSTRACT: We study group decision making in a two-­‐step process. In the first step, group members decide by a random dictator mechanism upon the rule they will use in the second step of their decision process. In the second step, all group members then vote between two alternatives and the decision is implemented according to the rule chosen in the first step. One alternative implies zero payoffs for all group members, the other alternative can have positive and negative valuations for each different group member, where valuations are drawn independently. Selfish players should choose a rule in the first stage that implements their preferred choice for sure in the second stage. Inequality averse players should choose even for small positive valuations a rule that implements the alternative that yields zero payoffs for all. Subjects that are concerned with maximizing total payoffs should for small positive or negative valuations choose majority voting as the decision rule. We find in a treatment with a symmetric distribution of valuations that in the second stage group members almost always vote in favor of the alternative that maximizes their own payoff, whereas the rule choice is often inconsistent with selfish payoff maximization. Furthermore, the rule choice shows no evidence of inequality aversion, but is consistent with efficiency concerns. In a further treatment with an asymmetric distribution of valuations, choices of subjects with a small negative valuation often deviate from selfishness. Author(s): Dirk Engelmann; Hans-­‐Peter Gruner Reference Dependence or Asset Integration? A Test of Competing Hypotheses for Choice under Risk Thomas Epper ETH Zurich ABSTRACT: We analyze a rich data set on 153 individuals’ risky choices over a large number of gain and loss prospects. Prospects covered a wide range of outcomes, up to an average subject’s monthly income. Subjects were paid in an incentive-­‐compatible manner. Each loss prospect was accompanied by an initial endowment leading to the same terminal outcome as a gain prospect. Therefore, prospects in the loss domain differed from prospects in the gain domain only in presentation format. This framing manipulation permits us to test three main hypotheses: First, asset integration which predicts identical behavior over gains and losses; second, reference dependence which predicts domain-­‐specific behavior; third, an even stronger hypothesis, reflection (Kahneman and Tversky 1979), which predicts risk aversion over gains and risk seeking over losses. Our analysis yields surprising results: We find that the large majority of subjects are risk averse in both domains or risk seeking in both domains. However, they exhibit comparatively higher risk tolerance over losses, but the effect is extremely small. While not perfectly consistent with asset integration, the weakness of the effect clearly contradicts an S-­‐
shaped utility function as well as the reflection hypothesis. A minority of subjects exhibit systematic domain-­‐specific behavior that can be classified as follows: One group displays the classical reflection effect while an even smaller group exhibits reverse reflection. We discuss potential explanations. Author(s): Thomas Epper; Helga Fehr-­‐Duda Bounded Rationality in Principal Agent Relationships Mathias Erlei Clausthal University of Technology ABSTRACT: We conducted six treatments of a standard moral hazard experiment with hidden action. All treatments had identical Nash equilibria. However, the behavior in all treatments and periods was inconsistent with established agency theory (Nash equilibrium). In the early periods of the experiment, behavior differed significantly between treatments. This difference largely vanished in the final periods. We used logit equilibrium (LE) as a device to grasp boundedly rational behavior and found the following: (1) LE predictions are much closer to subjects’ behavior in the laboratory; (2) LE probabilities of choosing between strategies and experimental behavior show remarkably similar patterns; and (3) profit maximizing contract offers according to the LE are close to those derived from regressions. Author(s): Mathias Erlei; Heike Schenk-­‐Mathes A Choice Prediction Competition for Social Preferences in Simple Distribution Games Eyal Ert Hebrew University ABSTRACT: Experimental studies of simple social interactions reveal robust deviations from rationality that can be described as the product of a small set of psychological factors (e.g., fairness, reciprocity). Yet there is a little agreement about the best abstraction and the relative importance of these factors. To address the quantitative ‘best abstractions and relative importance’ questions we organized choice prediction competitions for predicting the actions of the first and second movers in such interactions (http://sites.google.com/site/extformpredcomp/). Each competition was based on two experimental datasets: An estimation dataset, and a competition dataset. The studies that generated the datasets used the same methods and subject pool, and examined 120 games randomly selected from the same distribution. After collecting the experimental data to be used for estimation, we posted them on the Web, together with their fit with several baseline models, and challenged researchers to compete on predicting the results of the second (competition) set of experimental sessions. Thirty-­‐seven teams participated in the challenge. The experimental results show that behavior is highly correlated with the predictions of the subgame perfect equilibrium, and that when behavior is ‘irrational’ it seems to correspond to six well-­‐known behavioral tendencies. The competition results show that ‘strategy based models’, which assume reliance on several simple strategies, outperformed social utility models in predicting the behavioral data. Author(s): Eyal Ert; Ido Erev; Alvin E. Roth Gender, Information and Self-­‐Selection into Leadership in Group Decision-­‐Making Seda Ertac UCLA ABSTRACT: This paper studies (1) the willingness of individuals to make decisions on behalf of a group, i.e. become a ‘leader’, in a context that involves risk, (2) the amount of risk taken on behalf of the group. We manipulate whether leadership is voluntary or assigned, as well as whether leaders have access to information about group members’ risk preferences, in the form of advice given to the leader. In terms of the choice of being a leader, we find that a significantly lower proportion of women than men are willing to make decisions on behalf of the group, irrespective of information availability. Information about the group’s preferences/advice influences leaders’ decisions more strongly when leadership is assigned than voluntary. Moreover, male leaders who have self-­‐selected into the group decision-­‐maker role tend to follow the group’s preferences less than both female leaders and males who have been assigned to the leader role. Self-­‐selected male leaders also take more risk on behalf of the group than female leaders, whereas a gender difference in risk-­‐taking is absent when leadership is assigned. Author(s): Seda Ertac; Mehmet Y. Gurdal Information and Over-­‐dissipation in Rent-­‐Seeking Contests Francesco Falucchi University of Nottingham ABSTRACT: We investigate deterministic and stochastic repeated rent-­‐seeking contests. In the deterministic contest a contestant receives a share of the rent equal to her share of rent-­‐seeking expenditures. When subjects get feedback about the choices and earnings of all contestants we find that changes in behavior can be explained partially by adjustments in the direction of best response to rival’s choices in the previous round and partially by adjustments in the direction of imitating the most successful contestant. The implication of imitating the most successful contestant is over-­‐dissipation relative to Nash equilibrium expenditures levels. When subjects are only informed of own earnings imitating the most successful player is no longer possible and average behavior converges to Nash equilibrium levels. In the stochastic contest a contestant wins the entire rent with probability equal to her share of rent-­‐seeking expenditures. Here we also see over-­‐dissipation relative to Nash equilibrium in a full information feedback treatment. However, when we restrict information feedback to own earnings we observe the highest over-­‐
dissipation rates of all. Author(s): Francesco Fallucchi; Elke Renner; Martin Sefton Sacrifice Enrique Fatas University of East Anglia ABSTRACT: In this paper we experimentally analyze a variant of a multi-­‐person simultaneous Volunteer’s Dilemma (VD) with two novelties. First, players decide how to distribute their resources within a group under a veil of ignorance, dramatically increasing the potential team outcome. The VD becomes a complex sequential game with a trivial unique SPE in which resources and concentrated in the hands of one player because volunteering becomes a dominant strategy. Second, we study the effectiveness of sanctions to prevent individuals from volunteering, exploiting the benefits of redistribution. Sanctions are vertically implemented, as one group member is randomly selected and severely punished if actively contributing to the team. This random punishment mechanism closely resembles some security policies implemented in conflict areas. We test redistribution and sanctions in a neutrally framed experiment. Our results suggest that redistribution makes a big difference only when imposed exogenously. More surprisingly, we find that sanctions generate a significant increase on volunteering. Author(s): Marco Castillo; Enrique Fatas; Ragan Petrie Power of Sunspots: An experimental analysis Dietmar Fehr Social Science Research Center (WZB) Berlin ABSTRACT: We present an experiment in which extrinsic information (signals) may generate sunspot equilibria. The underlying coordination game has a unique symmetric non-­‐sunspot equilibrium, which is also risk-­‐dominant. Other equilibria can be ordered according to risk dominance. We compare treatments with different salient, but extrinsic signals. By increasing the precision of private signals, we manipulate the available public information, which allows us to measure the force of extrinsic signals. We also vary the number of signals and combine public and private signals, allowing us to see how subjects aggregate available (and possibly irrelevant) information. Results indicate that sunspot equilibria emerge naturally if there are salient (but extrinsic) public signals. However, salient private signals of high precision may also cause sunspot-­‐driven behavior, even though this is no equilibrium. The higher the precision of signals and the easier they can be aggregated, the more powerful they are in dragging behavior away from the risk-­‐dominant to risk-­‐dominated strategies. Sunspot-­‐driven behavior may lead to welfare losses and exert negative externalities on agents, who do not receive extrinsic signals. Author(s): Dietmar Fehr; Frank Heinemann; Aniol Llorente-­‐Saguer Temporal Stability of Discounting Behavior Helga Fehr-­‐Duda ETH Zurich ABSTRACT: We use an exceptionally rich panel data set on time discounting behavior of a broad sample of the Swiss population. To date there are only a few studies investigating the temporal stability of discount rates, and the evidence so far is inconclusive: on average, discount rates remained stable or increased over the study period. In this paper we study the discounting behavior of two treatment groups: One group of subjects responded to hypothetical tasks, in the other group every single subject was paid in an incentive-­‐compatible manner. Subjects were presented with identical intertemporal choice tasks 8 months after the first wave of experiments. The following results emerged: In the first wave, average discount rates in the hypothetical treatment exceeded discount rates in the real treatment by 17 percentage points p.a. This difference is highly significant. This substantial incentive effect even increased in the second wave: Whereas average discount rates in the hypothetical treatment remained stable, incentivized subjects’ rates declined, resulting in a treatment effect of 28 percentage points p.a. Treatment groups differed neither in observable characteristics nor in measured risk preferences. Author(s): Thomas Epper; Helga Fehr Pro-­‐Social Missions and Worker Motivation: An Experimental Study Sebastian Fehrler University of Zurich ABSTRACT: Do employees work harder if their job has the right mission? In a laboratory labor market experiment, we test whether subjects provide higher effort if they can choose the mission of their job. We observe that subjects do not provide higher effort than in a control treatment. Surprised by this finding, we run a second experiment in which subjects can choose whether they want to work on a job with their preferred mission or not. A subgroup of agents (roughly one third) is willing to do so even if this option is more costly than choosing the alternative job. Moreover, we find that these subjects provide substantially higher effort. These results suggest that relatively few workers can be motivated by missions and that selection into mission-­‐
oriented organizations is important to explain empirical findings of lower wages and high motivation in the latter. Author(s): Sebastian Fehrler; Michael Kosfeld The Bomb Risk Elicitation Task Antonio Filippin University of Milan ABSTRACT: This paper presents the Bomb Risk Elicitation Task (BRET), an original procedure aimed at measuring risk attitudes. The lab version is a visual continuous-­‐time task asking each subject to decide at which point to stop collecting a series of 100 boxes, one of which containing a bomb. Earnings increase linearly with the number of boxes collected but are equal to zero if they contain the bomb. The task is designed in order to avoid potential truncation of the data, so that subjects are free to choose any number between 0 and 100. This task allows to estimate both risk aversion and risk seeking very precisely. Moreover, it is not affected either by the degree of loss aversion or by violations of the reduction axiom. We present the results of a large scale experiment aimed at validating the task. First, we test its robustness along several dimensions (speed, high-­‐low stakes, number of boxes, real-­‐time explosion, paper-­‐and-­‐pencil version). Second, we compare the results to those obtained replicating other well-­‐known tasks and surveys. The main results show that a) the BRET is quite robust to changes in its protocol; b) it is perceived as either significantly simpler or not more difficult than any other task; c) it generates a distribution of risk preferences with a relatively larger fraction of risk lovers; d) it correlates quite well with other tasks. Author(s): Paolo Crosetto; Antonio Filippin Decentralized vs. Centralized Punishment Sven Fischer Max Planck Institute ABSTRACT: Research on the efficiency of punishment in public good games tends to focus on decentralised punishment. Although this allows for counter punishment, overall efficiency increases if punishment is possible. This, however, does not hold under noisy signals. In a set of experiments we compare the standard decentralised punishment institution employed by most experimental studies, with a centralised institution where one additional player accepts the role of the sole punisher. While he can not contribute to the public good, he, nevertheless, benefits equally from it. We, furthermore, compare noisy with reliable signals about the contributions of other group members. Clearly, with one centralised punisher there are no conflicting norms and counter punishment is not possible. However, we only observe little effects of the punishment institution on both, contributions and distributed punishment points. This holds irrespective of the signal quality. Author(s): Sven Fischer; Kristoffel Grechenig; Nicolas Meier When to pay more: status and culture in Principal-­‐Agent interactions Josepa Miquel Florensa Toulouse School of Economics ABSTRACT: We study the effects of status and culture in an experimental Principal-­‐Agent game. In the model we allow for inequity-­‐averse principals and ‘envious’ agents: those who receive an endowment and use it to make piece rate offers may care about fairness(or at least the appearance of fairness), while those who receive the offers and choose their effort may resent being the ones that have to work, receiving only part of the resulting benefit. Moreover, behaviour may be affected by status and cultural norms. In the experiments, status is awarded to subjects either randomly or based on the answer to a question from an IQ test. In each randomly matched principal-­‐agent pair, the principal chooses the agent’s status-­‐contingent piece rate for each correct answer given to a set of 12 questions from a (different) IQ test. We perform the experiment in Cambridge (UK) and in HCMV (Vietnam). We find that in Vietnam piece rate offers are significantly different for status and non-­‐status agents under both mechanisms of status allocation, while in Cambridge they differ (only) when status is related to intelligence. On the other hand, the responsiveness of agents’ performance to piece rates is significant in Cambridge, and insignificant in Vietnam. Author(s): R. Dessi; J. Miquel-­‐Florensa Biased Effects of Taxes and Subsidies on Portfolio Choices Martin Fochmann University of Magdeburg ABSTRACT: We study how the perception of taxes and subsidies affect portfolio choices in a laboratory experiment. We find highly significant differences after intervention, even though the net income is identical in all our treatments and thus the decision pattern of investors should be constant. In particular, we observe that the willingness to invest in the risky asset decreases rapidly when an income tax has to be paid or when a subsidy is paid. If we combine both a tax and a subsidy, this effect intensifies. These findings give interesting insights into the decision behavior of investors under government interventions. Author(s): Hagen Ackermann; Martin Fochmann; Benedikt Mihm; Abdolkarim Sadrieh An Experiment in Hiring Discrimination via Online Social Networks Christina Fong Carnegie Mellon University ABSTRACT: Many Internet users reveal online personal information that employers are not supposed to use during the hiring process. Self-­‐reported surveys of employers suggest that US firms have, in fact, started using popular online social networks to seek information about prospective hires.. However, no randomized experiment has yet measured the extent to which firms use online social networks to screen job applicants, and how their hiring activities are influenced by the information they find. We use two IRB-­‐approved randomized experiments to investigate the effects of job candidates’ personal information, posted on an online social network, on the search activities of employers. We design resumes and online social network profiles in which we vary personal traits of prospective job candidates. We manipulate information about religious affiliation, family status, sexual orientation and professionalism. In a field experiment we submit applications on behalf of our candidates to real job openings, and measure employer responses. Our main dependent variable is the number of callbacks each candidate receives. We use an additional survey experiment as a manipulation check. In a pilot version of the survey experiment we have found that our manipulations are successful and that subjects believe that our candidates are real. We have also found statistically significant discrimination against our candidate with unprofessional online social network information. We expect to present preliminary data from the field experiment at the June 2012 ESA meetings. Author(s): Alessandro Acquisti; Christina M. Fong Social Preferences and Effort Provision Joshua Foster University of Arkansas ABSTRACT: Prior studies on social preferences have found some individuals exhibit stable and consistent pro-­‐social preferences. Other studies have shown that pro-­‐social preferences are conditional on specific psychological factors, such as the (lack of) ambiguity regarding one’s ability to influence an outcome. Through a binary dictator game with an effort provision task, I use the principles of the general axiom of revealed preference (GARP) to check for consistency between an individual’s stated social preferences and their behavior. In general, dictators who prefer to maximize their own payouts are consistent in their stated preferences and their effort provision to keep those preferences. However, dictators who exhibit pro-­‐social preferences are much less consistent in their effort provision to keep their stated preferences. Author(s): Joshua Foster Identifying Predictable Players Daniel Fragiadakis Stanford University ABSTRACT: Non-­‐equilibrium behavioral game theory models have been useful in organizing experimental data in strategic decision-­‐making studies. An appropriate next step is whether these models can identify strategic players and make out of sample predictions of their behavior. In our experiment, we control subjects’ beliefs by having subjects play against random participants, and then against their past selves as in Ivanov et al., who found it quite difficult to use these models to make out of sample predictions in games of incomplete information. We adapt their paradigm to simple two-­‐player guessing games of complete information of the form in Costa-­‐
Gomes and Crawford. We argue that players that best respond to their past selves should be predictable, and seek to describe all such subjects’ behavior using existing (and possibly novel) behavioral game theory models. Additionally, for subjects with behavior we can purely describe from initial play against other subjects, we investigate whether or not we can correctly predict their behavior when they play against their past selves. We find that most subjects that best respond to their past behavior are those that we ‘classify’ (using the methods in Costa-­‐Gomes and Crawford) in phase I, and that the classified subjects in phase I are correctly predicted in phase II. In addition, we investigate various methods of classifying subjects and find that as we classify more subjects, the fraction of classified subjects that are correctly predicted in phase II gets worse. Author(s): Daniel E. Fragiadakis; Asen Ivanov; Daniel T. Knoepfle; Muriel Niederle Rock-­‐Paper-­‐Scissors as a Population Game Daniel Friedman UC Santa Cruz ABSTRACT: We report laboratory experiments using new, visually oriented software in order to explore the dynamics of 3x3 games with best response cycles. Each moment, each player is matched against the entire population, usually 8 human subjects. A ‘heat map’ offers instantaneous feedback on current profit opportunities. In the continuous slow adjustment treatment, we see distinct cycles in the population mix. Their amplitude, frequency and direction seem consistent with standard learning models. Cycles are more erratic and higher frequency in the instantaneous adjustment treatment. Control treatments (using discrete time simultaneous matching) replicate previous results that do not exhibit cycles. Average play is approximated fairly well by Nash equilibrium, and an alternative point prediction, ‘TASP’ (Time Average of the Shapley Polygon), captures some regularities that NE misses. Author(s): Timothy N. Cason; Daniel Friedman; Ed Hopkins Who cares for social image? Interactions between intrinsic motivation and social image concerns Jana Friedrichsen University of Mannheim ABSTRACT: Empirical research suggests that choices for products considered to be ethical are driven both by intrinsic motivation and concerns for social approval. How these forces interact is important for supply and pricing policy of these products. However, little is known about this interaction from empirical work, which we hence assess in an experiment. In particular we try to find out whether intrinsically motivated individuals react more or less strongly to opportunities for image building than other individuals. We recruited new participants, offering a choice between Fairtrade and non-­‐Fairtrade chocolate as a reward for signing up to the subject pool. We use their choice as a proxy for intrinsic preference for Fairtrade products. The participants are later on recruited for laboratory experiments where they take part in a market game. At the end of the experiment we elicit their willingness-­‐to-­‐pay for their preferred type of chocolate (Fairtrade or non-­‐Fairtrade). Treatments differ in whether these choices are made privately or publicly. The difference in willingness-­‐to-­‐pay for Fairtrade between these treatments serves as our measure for image concerns. We compare this measure for the two groups of individuals who chose Fairtrade or non-­‐Fairtrade chocolate at the recruiting stage, respectively. The difference-­‐
in-­‐differences in willingness-­‐to-­‐pay across the groups of participants shows whether image concerns interact with our measure of intrinsic motivation. Author(s): Dirk Engelmann; Jana Friedrichsen An experimental consideration of strong ambiguity in call markets and double auction markets Sascha Fullbrunn Radboud University Nijmegen ABSTRACT: Several individual choice experiments show that decision makers prefer taking gambles with known-­‐risk probabilities over equivalent gambles with ambiguous probabilities. Thus, subject’s willingness to pay for risky assets tends to be higher than for equivalent ambiguous assets. When it comes to markets, however, subject’s decisions are no longer separated from others and may be reevaluated by market feedback. But is market feedback sufficient to overcome the ambiguity effect? Or do prices reflect ambiguity aversion? To evaluate this question, we conduct laboratory experiments where subjects simultaneously trade risky and ambiguous assets, using an offline version of the source method for Ellsberg-­‐like uncertainties from Abdellaoui et al. (AER, 2011). While recent experimental results from Sarin and Weber (MS, 1993) and Bossaerts et al. (RFS, 2010) suggest that aversion to ambiguity does not vanish in markets, we find no systematic difference between risky and ambiguous assets in terms of prices and bids, volume or portfolio decision. In contrast to other market studies, we consider strong ambiguity (rather than weak) in both double auction markets and call markets. The same source method in a Becker-­‐DeGroot-­‐Marschak treatment, however, confirms ambiguity aversion on the individual level. Author(s): Sascha Fullbrunn; Holger Rau; Utz Weitzel Labor Market Discrimination in Lima, Peru: Evidence from a Field Experiment Francisco Galarza Universidad del Pacifico, Peru ABSTRACT: Latin America is seen as a highly discriminatory society. However, such a widespread belief appears not to be based on strong empirical evidence (Chong and Ñopo, 2007). This paper utilizes experimental data to identify the existence of discrimination in the labor market of Lima, Peru, a fast-­‐growing country where much anecdotal evidence suggests the presence of discriminatory practices at many instances of daily life. Focusing on two dimensions, sex (female/male) and surnames (indigenous/white), we sent 4,820 fictitious and equivalent CVs in response to 1,205 real job vacancies advertised in an important Peruvian newspaper. We randomly allocated indigenous and white surnames across CVs sent in application to professional, technical, and unskilled jobs. Overall, we find that males receive 20 percent more callbacks than females, and whites receive 80 percent more calls than indigenous applicants. Within job categories, we find sexual discrimination only in unskilled jobs, while discrimination against indigenous is verified across all job categories. There are no statistically significant differences in the time to receive a phone call among male/female, and white/indigenous applicants. Author(s): Francisco B. Galarza; Gustavo Yamada “Beijing’s Little Emperors -­‐ Behavioral Impacts of China’s One-­‐
Child Policy” Lata Gangadharan Monash University ABSTRACT: We examine the behavioral impact of exposure to China’s One Child Policy. Individuals born just before and just after the One-­‐Child Policy, who are otherwise similar, participate in experiments designed to elicit preferences with regard to altruism, trust, risk-­‐taking and competitiveness. Those born under the policy and those who grow up as single-­‐children because of the policy are found to be significantly less trusting, less trustworthy, less likely to take risks, and less competitive. Survey data shows they are also less optimistic, less conscientious, and more prone to neuroticism. Thus, the One Child Policy has significant ramifications for Chinese society and for the world with which China increasingly deals. Author(s): L. Cameron; N. Erkal; L. Gangadharan; X. Meng An experimental study of allocation procedures and their effect on trust Diana Gaspari Department of Economics and Management -­‐ University of Padua ABSTRACT: A relationship frequently begins with the act of splitting a common endowment. The fairness of this division may influence the success of the relationship. This paper investigates the effects of two versions of a Divide-­‐and-­‐Choose (D&C) mechanism in a first phase on a trusting relationship between the same subjects in a second phase. In the first version of the D&C mechanism, the division is informative about pro-­‐social intentions, whereas in the second version an equal division is consistent both with self-­‐interest and other-­‐regarding concerns. The experimental data find that, irrespective of the mechanism used, couples who reach an equal division in the first phase show higher levels of trust in the second phase. Hence, subjects do not rationally extract and use the information about the intentions of their partner from the division phase. The results suggest that reaching an equal division creates a positive feeling that enhances trust inside the couple even if it is not possible to infer about the intentions of the partner. Author(s): Diana Gaspari A Structural Estimation of French Farmers' Risk and Ambiguity Preferences in domain of gain and loss. Xavier Gassmann INRA ABSTRACT: Risk and ambiguity are pervasive in farming activities. Although agricultural economists have a long tradition of analysing risk, there is still a lack of understanding of farmers’ risk and ambiguity preferences. We designed an artefactual field experiment involving real payments to elicit French farmers’ risk and ambiguity preferences. The 197 participants were face-­‐to-­‐face interviewed. Preferences were elicited for both domains: gains and losses. We use multiple price lists. We aim at structurally estimating these preferences. We use a model that combines a second order model for ambiguity and a model that allows for differences in utility in the gain and loss domains and probability distortion. Moreover, we allow for an endogenous reference point that we estimate. We find (i) farmers are slightly risk averse in the gain and loss domains and have an inverse s-­‐shaped probability weighing function for risk; (ii) farmers are slightly ambiguity averse in the gain and loss domains and have an inverse s-­‐shaped probability weighing function on subjective beliefs in the gain domain but do not distort probabilities in the loss domain; (iii) farmers have a positive reference point that differs according to the order in which loss and gain tasks are presented to subjects. Author(s): Douadia Bougherara. Xavier Gassmann. Laurent Piet. Arnaud Reynaud Emergence of Risk Sharing in a Stochastic Environment Steven Gazzillo Rutgers ABSTRACT: Controversy exists about the act of giving as altruistic instead of self-­‐interested behavior. Each side of this argument interprets similar results from similar experiments in different ways. One side argues the results show that the appearance of altruistic behavior can be explained by self-­‐
interested motives. The other side argues these results are evidence of something called group selection, where a group member takes an action that is harmful to itself but beneficial to the group. We consider this question using a novel approach. We create a rich experimental environment in which subjects have the ability to cooperate to improve the group’s outcome by sharing their wealth in non-­‐compulsory, non-­‐enforceable risk-­‐sharing arrangements. We find that average subject behavior appears to be motivated by self-­‐interest rather than group survival. Author(s): Athena Aktipis; Lee Cronk; Steven Gazzillo; Barry Sopher Disclosure of personal information under risk of privacy shocks Caterina Giannetti GSBC Jena, University of Siena ABSTRACT: Individuals disclose an increasing amount of personal information when buying goods or services. Personal details are revealed despite the rising threat of privacy breaches at firms that collect the information. Yet, we know surprisingly little about the trade-­‐offs individuals consider when disclosing personal information in economic transactions. We present this problem as a dynamic lottery with personal information under the risk of privacy shocks. We explore how the notification about a privacy breach changes an individual’s behavior with regard to disclosing personal information. Author(s): Francesco Feri; Caterina Giannetti; Nicola Jentzsch Efficiency Wages Survive Multiple Sources of Income Inequality Karina Gose University of Magdeburg ABSTRACT: Employee’s effort choices in gift-­‐exchange games do not change once employees face disadvantageous payoff inequality. However, when this payoff inequality results from an additional lump sum payment to the employer, effort per wage unit is significantly smaller than in a game in which payoff equality can be established. Thus, instead of inequality aversion, the perceived fairness of wages seems to drive employee’s effort choices under disadvantageous payoff inequality. Author(s): Karina Gose Facing a Dilemma: Cooperation and Beauty Silvia Graetz University of Zurich ABSTRACT: We explore the influence of physical attractiveness on cooperative behavior. A large body of literature has a shown the existence of a ‘beauty-­‐is-­‐good’-­‐stereotype inducing a beauty premium in various economic interactions. We combine data from 211 episodes of a television game show in which contestants play a face-­‐to-­‐-­‐face prisoner’s dilemma game, and from independent facial appearance ratings of the contestants. Our results suggest that attractiveness is an important factor for cooperative behavior even in an environment of extraordinary high stakes, communication, and past behavior. Although there is no difference between facially attractive and unattractive players regarding the decision to cooperate, facing a facially attractive opponent increases cooperation significantly. The finding is robust to gender, age, stake size, communication, and pre-­‐play; and is especially prevalent in mixed-­‐gender interactions. The marginal beauty premium for a one standard deviation increase in facial attractiveness amounts to an increase of contestant’s expected earnings by up to £ 2153. Also, the probability to obtain positive earnings rises by 5.9 percentage points for facially attractive contestants. Addressing mutual outcomes, the presence of at least one facially attractive contestant fosters mutual cooperation, and pairs of facially unattractive contestants are even significantly more likely to defect than pairs of attractive and of unequally attractive contestants. Author(s): Silvia Graetz; Donja Darai An Experimental Study Of Payment Mechanisms In A Multiple Principal Agent Setting Ellen Green University of Delaware ABSTRACT: Trends in healthcare contract design encourage organizations to micromanage their physicians using payment as an incentive for the completion of various tasks (e.g. pay-­‐for-­‐performance). However, as with other interventions, these contracts have not demonstrated improvements in the quality or the cost of care, and in some cases have been detrimental to the patient’s quality of care (Rosenthal et al. 2006). These failures emphasize the need for a more fundamental understanding of how these contracts incentivize doctors. To aid this understanding, we treat the established physical-­‐client-­‐employer relationship as a multiple principal agent problem, a situation where an agent has responsibilities to two different principals (e.g. a physician is an agent of their patients (their clients) and an agent of their health care employer.). We introduce, the ‘Dual Principle Agent Game’ which provides a novel experimental design to explore multiple principle agent problems. We apply this game design, which uses a real-­‐effort task to test the relative performance of common payment mechanisms employed in this dual-­‐principal agent relationship (Piece Rate, Flat Rate, Salary, Bonus, and Socialization) and examines the effect that different incentive structures have on agent behavior. This study suggests, contrary to standard contract theory, that a direct link between an agent’s task and their income may not be necessary to achieve a high quality of service in the dual-­‐principal agent problem. And therefore, the Salary and Flat Rate payment structures should be heavy contenders in the current physician payment structure debates. Author(s): Ellen P. Green The Importance of Knowing Your Own Reputation Matthias Greiff JLU Giessen ABSTRACT: The effect of reputation on prosocial behavior has been investigated in several experimental studies (e.g. Bolton, Katok and Ockenfels, 2005; Seinen and Schram, 2006). In these studies, reputation is objective information about past behavior, i.e., how behavior affects reputation is commmon knowledge. However, there are almost no experimental studies in which reputations are generated endogenously. If reputations are generated endogeneously, a subject’s reputation is a noisy signal about past behavior because the process by which reputations are generated is not common knowledge. In this paper we study how subjects evaluate each other, and how the resulting reputations affect prosocial behavior. Each period, two subjects are randomly matched and play a public good game. Subjects contribute and evaluate each others contributions by assigning 0 to 10 stars. A subject’s reputation is her most recent evaluation. The treatment effect is the information available: each subject receives no information, information about her own reputation, information about the other subject’s reputation, or information about her own and the other subject’s reputation. Throughout all treatments subjects who contribute more receive higher evaluations, so that a subject’s reputation is an informative though noisy signal of past behavior. Information about own and other’s reputation significantly increases contributions compared to the control treatment in which no information is provided. Information about the other’s but not the own reputation is not sufficient for increasing contributions, and information about the own but not the other’s reputation decreases contributions. Our results indicate that subjects are forward-­‐looking: they condition their behavior on the expected behavior of others (conditional cooperation) rather than rewarding good and punishing bad reputations (trigger-­‐strategy). Author(s): Matthias Greiff; Fabian Paetzel Patents versus subsidies Jens Grosser Depts of Political Science and Economics ABSTRACT: We study the effects of patents and subsidies on R&D investment decisions. Firms first invest in cost reductions and, thereafter, compete in two periods of a Bertrand market. In the social welfare optimum, patents and subsidies are linear substitutes. The choice from the set of all optimal R&D policies (i.e., mixes of various degrees of patent-­‐protection and subsidies) has strong effects on rent shifting between consumers, firms, and the government. Firms prefer patents and consumers prefer subsidies. We use laboratory experimentation to test a simplified model where we concentrate on the firms R&D investment and pricing decisions (optimal R&D policies and consumer decisions are automated). Most of our comparative statics predictions are supported. But there is overinvestment in R&D in all conditions, yielding perverse results where social welfare is about the same with or without patents and subsidies. Our study suggests that R&D policies should be mainly interpreted as the outcome of political demands instead of mere social welfare optimization. Author(s): Donja Darai; Jens Grosser; Nadja Trhal Reputation and Advice: Political Correctness in the Laboratory Brit Grosskopf Texas A&M University ABSTRACT: We experimentally test the bad reputation predictions of Morris (2001). An informed advisor can give advice to an uninformed decision maker in a repeated cheap talk game. The actions of the decision maker determine the payoffs to both agents. We vary the incentive of the advisor to truthfully reveal valuable information by varying the importance of future payoffs to both agents. We find that advisors are concerned about their reputation. However, they are not as "politically correct" as suggested by theory. Author(s): Brit Grosskopf; Rajiv Sarin Conflict Networks: An Experiment Zeynep Gurguc Imperial College Business School ABSTRACT: We examine a model of multiple bilateral conflicts embedded in a network structure where opponents invest in specific conflict technology in order to increase their probability of wining available resources. By analyzing the relationship between network structures and total conflict intensity along with the amount of investment realized by each party, it is possible to provide useful insights for conflict resolution. Hence, this paper tests if the individual conflict spending depending on the underlying network characteristics is significantly different than suggested by the theory. Moreover, it also compares the conflict investment of each party in various network structures differing in degree and/or size. Five separate treatments are implemented in order to check for the cases of circle, star and complete networks in three or five people structures. The results show that on average individuals tend to over-­‐invest in the conflict technology occasionally even to the point of being irrational. However, for regular class networks, i.e. circle and complete, the level of over-­‐investment is lower than for the irregular class, i.e. star. Another striking result is that when subjects have more than one link available they choose to invest significantly differently for each link even though the theory suggests that they should have a unique and equal investment strategy per link. Author(s): Zeynep Gurguc & S. Basak Gunes Learning to be Probabilistically Sophisticated Yoram Halevy University of British Columbia ABSTRACT: Risk equivalents to uncertain environments are elicited in an experiment that controls the instruction level subjects receive. We find much higher frequency of probabilistically sophisticated behavior than in previous experiments that elicited certainty equivalents, and establish a causal relation between reduction of compound objective lotteries and ambiguity attitude. Author(s): Yoram Halevy Don't Take What Isn't yours Simon Halliday University of Siena ABSTRACT: Many economists have argued whether people treat positive and zero allocations similarly. Recent research has gone beyond positive and zero allocations to assess whether people will make negative allocations, that is, to take or steal. We examine whether people treat giving differently to how they treat taking. We use a Dictator Game with third-­‐party punishment in which we introduce a treatment allowing the dictator to take from the receiver, we contrast this with a baseline in which no dictator could take, and another treatment without third parties. The largest proportion of dictators in all three games chooses the most self-­‐interested allocation and the proportion appears to be unaffected by the presence of third parties. But when dictators can take, more third parties punish the most self-­‐interested option and they punish it much more severely than in the baseline. Third parties, it appears, view zero as the boundary between good behavior and bad behavior as though saying, ``Don’t take what isn’t yours.’ Author(s): Simon Halliday Managing Credibility John Hamman Florida State University ABSTRACT: While individual decision making regularly demonstrates concerns for fairness and equality, recent experimental evidence reveals that delegation and hierarchy can dramatically reduce fair and egalitarian behavior. For example, a recent study by Hamman, Loewenstein and Weber (2010, HLW) shows that delegation of decisions almost entirely extinguishes fairness. In this paper, we study the effectiveness of alternative interventions aimed at restoring fairness in contexts in which delegation by a principal ton an agent may otherwise result in unfair behavior. We specifically study two distinct categories of approaches. The first category aims to increase the perceived responsibility of those farther up the hierarchy (‘principals’) for unfair outcomes produced on their behalf by those to whom they delegate decisions (‘agents’). The second category of interventions aims to increase the ability of agents to collectively resist competitive pressure to act unfairly on principal’s behalf. None of our interventions limit the choices available to any individual principal or agent. We find that most interventions are ineffective, highlighting the power of delegation, as studied by HLW, for reducing fairness. However, we also find that when agents are given the ability to collectively set standards, and sanction those who violate such standards, fairness is entirely restored to the baseline levels observed without delegation. Our findings support the role of labor unions and professional associations in mitigating unfairness produced through hierarchy and delegation. Author(s): John Hamman; Peter Kriss; Roberto Weber Preferences of Migrants: A Field Experiment in China Li Hao University of Arkansas, Fayetteville ABSTRACT: In this paper we study whether migration decisions can be predicted by one’s preferences regarding risk, ambiguity, inequality aversion, and competitiveness. We conducted field experiments in six locations in China. Three on the east coast (East China), and three inland (West China). Subjects are either (i) migrant workers, (ii) non-­‐migrant workers in an area with migrants, or (ii) non-­‐migrant workers in an area without migrants. Our main findings are twofold. First, migrants are significantly more likely to enter in a market entry game than non-­‐
migrants. Second, workers in East China, whether migrants or not, demonstrate a higher tolerance for risk, ambiguity and inequality than their counterparts in West China. Our results suggest that migration may be driven more by competitive impulses than attitudes towards risk or inequality. If holding a job is considered ‘winning’, then this competitive orientation may help to explain both the nature of positions offered to migrants, as well as their willingness to devote substantial work effort to those positions. Author(s): Li Hao; Daniel Houser; Lei Mao; Marie Claire Villeval Performance Appraisals in Germany and China A Lab Experiment Christine Harbring RWTH Aachen ABSTRACT: There are many anecdotes on the problems of ‘Western’ companies trying to implement their organizational practices in foreign countries, e.g. on the occasion of a multi-­‐national joint venture. However, the reasons and underlying mechanics behind the lack of acceptance and effectiveness of certain management practices are widely unexplored. One instrument that is often mentioned to be problematic is the process of evaluating employee’s performance and rewarding them. We approach this discussion by providing an experimental test of behavior in Germany and China in a situation in which supervisors have to assign ratings (resulting in monetary bonuses) to employees for performing a real-­‐effort task in the lab. Moreover, it has often been stressed that in Western companies supervisors are known to be too ‘lenient’ and reluctant to use the lower spectrum of possible performance ratings and/or do not sufficiently differentiate among workers. These biases may put the incentive effect of the performance appraisals at risk. One solution typically proposed is to force managers to rank employees. In our experiment, we analyze two appraisal systems in each country: one in which supervisors are free to assign ratings and another in which they are forced to differentiate among workers. Our results indicate interesting behavioral differences between the two countries in both treatments. For example, we find that supervisors in Germany assign good grades to the majority of participants whenever they can (often labeled as ‘leniency bias’) while Chinese supervisors rather tend to give very similar ratings to employees (‘centrality bias’). Author(s): Christine Harbring; Heike Hennig-­‐Schmidt; Gari Walkowitz Session size and its effect on identity building: Evidence from a public good experiment Haoran He Beijing Normal University ABSTRACT: The effect of session size has been ignored in the experimental studies, although it may play a role through changing people’s perceptions of the interactions with others. The present paper tries to take the potential influence of session size into consideration, and investigates how induced identity affects cooperative behavior and how this effect differs depending on the session size in a repeated public good experiment with constant group size and partner matching. We find that induced identity significantly enhances cooperation only when session size is small and in the initial period. The similar effects of identity in small and large sessions from the second period onwards suggest that session size is not a confounding factor of identity in repeated interaction settings. Author(s): Qian Weng; Haoran He Incentives in Experiments: A Theoretical Analysis Paul Healy Ohio State University ABSTRACT: Experiments elicit choices from a given set of decision problems. When multiple decision problems are offered in the same experiment, behavior in one decision problem may be distorted by the choices made in others; incentive compatibility of the experiment is violated. Assuming all preferences that respect stochastic dominance are admissible, we show that the Random Decision Selection (RDS) mechanism-­‐-­‐-­‐which pays for one randomly-­‐selected decision-­‐-­‐-­‐
is essentially the only incentive compatible payment mechanism. If an experiment pays for multiple decisions, then a `no complementarities’ condition must be assumed. Practical implications for experimentalists are discussed. [Note: An early version of this was presented at ESA Tucson 2011. The paper has been substantially updated since then.] Author(s): Yaron Azrieli; Christopher Chambers; Paul J. Healy We Should Totally Open a Restaurant: A Taxonomy of Optimism and Overconfidence Stephanie Heger Washington University in Saint Louis ABSTRACT: We define and experimentally investigate two concepts that are inconsistently treated in previous literature: optimism and overconfidence. These types of ``incorrect’ beliefs have potentially major decision-­‐making implications. We are particularly interested in the role optimism and overconfidence may play in an individual’s choice to become self-­‐employed despite the high probability of failure in entrepreneurship or to remain in self-­‐employment despite unexpectedly low earnings. Our definition of optimism amounts to a formalization of ``wishful thinking,’’ whereby agents’ incorrect beliefs are driven by their preferences for certain outcomes. We characterize overconfidence as agent certainty that is greater than what can be warranted by available data. We depart from previous literature in three important ways. (1) We propose definitions that are general, i.e., not context-­‐dependent, and therefore applicable to diverse decision-­‐making scenarios. (2) Our definitions explicitly distinguish optimism from overconfidence, which allows us to separately identify the presence of these beliefs and their interaction. (3) Our experimental approach is designed to eliminate potentially confounding factors in eliciting errors in beliefs. In particular, we can pinpoint optimism by maintaining explicit control over which outcomes are ``favorable.’’ Moreover, we elicit agent beliefs about an objective distribution through a non-­‐performance-­‐based task, which helps to ensure that our results do not reflect agent beliefs about their own ability. Author(s): Stephanie A. Heger; Nick W. Papageorge The Ongoing Quest for QWERTY Leif Helland BI Norwegian Business School ABSTRACT: First, we replicate the remarkable result of Hossain & Morgan (AER 2009), in which subjects in an experimental market tip almost perfectly to the superior platform even if an inferior platform enjoys initial monopoly. Next, we show that this result disappears when seemingly innocent increases in out-­‐of-­‐equilibrium payoffs are introduced. The inflated payoffs do not alter payoff-­‐ or risk-­‐dominance relations, and do not impact on players’ security levels. Lastly, we show that coordination is mostly driven by a simple, non-­‐rational, learning rule. We conclude that the need for a theory of equilibrium selection cannot be bypassed by appealing to the realities of the (experimental) market place. Author(s): Tom-­‐Reiel Heggedal; Leif Helland Leaders Are Made: The Role of Circumstance in Taking the Initiative Pablo Hernandez UC Berkeley-­‐Haas ABSTRACT: In practically every historical context we find visionary individuals who take the initiative to guide groups towards misery or success. One view is that leaders have distinctive attributes that make them so exceptional that they will lead regardless of circumstance. An opposite view is that personal traits matter little, and circumstance a lot. Investigating the role of personal characteristics and circumstance in the emergence of leaders is difficult using empirical data. The natural confound is that the features of leaders and circumstances may be jointly determined. To make progress on this matter, we study leadership emergence through a lab experiment relying on simple tests and strategic interactions. We base our experimental design on a parsimonious theoretical setting in which we study the effect of the payoff structure on initiative. In a nutshell we vary the degree to which the games subjects play pose a strategic conflict. We find that higher strategic conflict (think for example, on a higher payoff from defection conditional on cooperation in a 2-­‐by-­‐2 PD) reduces the set and the variety of individuals who take the initiative. In addition, we observe that those willing to lead (and also the ones willing to ‘not-­‐to-­‐lead’) possess different attributes depending on the payoff structure -­‐
circumstance does matter. We argue that the mechanism through which circumstance affects initiative is the differential activation of a form of social preference similar to Andreoni and Miller’s (1993) reciprocal-­‐altruism. Author(s): Pablo Hernandez On the Communication Efficiency: An Experiment Penelope Hernandez ERI-­‐CES University of Valencia ABSTRACT: This work analyzes communication efficiency in experimental laboratory. It is well known how communication may reduce inefficiencies among players. Nevertheless, to measure how important the communication is, remains still being an open problem. Our experiment is based on the theoretical model of communication from Gossner, Herna¡ndez and Neyman, 2006 (GHN henceforth). In that paper, optimal communication strategies are designed for the 3–player games under asymmetric information. In particular, player 1 represents the random nature, an i.i.d. procedure, player 2 is a fully informed player (the wiser) and player 3 is the less informed player, called the agent. It is characterized how the wiser may efficiently transmit online information regarding the play of the player 1 to the player 3. We apply the above model for the particular instance, online matching pennies, but in a finite environment. First at all, we characterize given a finite length of the game what is the optimal structure of the equilibrium strategies for the wiser and the agent. Second we establish a length of the game such that the optimal strategies are the majority rule for the minimal length equals to three. Given the above insights we adjust the experimental design to the case of sequences of length 55 and random binary sequences with probability (1/2,1/2) as the nature. The experiment was run at Lineex Laboratory in Valencia University. There was two sessions with 60 participants each one. Our subject pool is drawn from volunteering students of the social sciences campus of the University of Valencia. The recruiting procedure followed the protocol conducted at Lineex. The task did not last more than 60 minutes and the subjects earned 19 euros on average. The subjects did not receive a show-­‐up fee. The experiment had two phases: a chat phase and a play phase. During the chat phase, both players exchanged information to coordinate their strategies for the second phase. After the chat-­‐phase, only the wiser knows the realized (random) sequence of the nature. Both players play during 55 rounds a binary game. If both actions coincide with the nature action atroundt, then both players get a positive pay off. Otherwise, they get zero. Our data show that communication is possible and that may be efficiently transmitted from the wiser to the agent through a 3 minutes online chat. Actually, we categorize three levels of communication. The first level corresponds to no communication, the second level the use of one informative signal and the third one the use of more than one informative signal. We obtain 57%, 41% and 2% couples for each level. Moreover there is a positive correlation between the level of communication and the payoff achieved for each agent. Actually the average payoff for each group is 14.8, 18.32 and 21, 25 respectively. Therefore, more communication implies more payoffs. Finally, the measure of communication obtained in GHN is a good approximation of the grade of communication in the laboratory for this game. Author(s): Aurora Garcia Gallego; Penelope Hernandez; Amalia Rodrigo Determinants of participation, effort and success in group contests: Evidence from combined field and lab data Florian Hett University of Mainz ABSTRACT: We exploit a field setting in which students can voluntarily participate in a group contest competing over a non-­‐monetary award. We combine this information with laboratory data of experimental measures of cooperation and competitiveness, as well as information about socio-­‐
economic and personality characteristics. We find that more pronounced cooperative behavior in the lab is correlated with more effort in the field. Furthermore, we find a negative effect of competitiveness on effort provision. Finally, large groups perform significantly better than small ones. Author(s): Yann Girard; Florian Hett Relative Culpability in Contingent Valuation and Laboratory Experiments of Willingness to Pay to Reduce Negative Externalities Benjamin Ho Cornell ABSTRACT: Recent field experiments show that peer information can induce people to reduce their production of negative externalities. Related work in psychology demonstrates that inducing feelings of relative culpability can induce pro-­‐social behavior. We use a contingent valuation and parallel lab experiment to explore patterns of responses to norm-­‐based interventions. Asymmetric responses between those whose impacts are above or below the norm is found to be robust across decision settings. Substantial heterogeneity in responses is observed across a number dimensions not explored in large field experiments, raising questions about the universality of peer-­‐information effects and the design of such programs. Author(s): Benjamin Ho; Gregory Poe; Antonio Bento Stock Market Investors: Who Is More Rational, and Who Relies on Intuition? Shlomit Hon-­‐Snir The Max Stern Academic College of Emek Yezreel ABSTRACT: Contemporary research documents various psychological aspects of economic and financial thought and decision-­‐making. The main goal of our study is to analyze the effects of five well-­‐
documented behavioral biases, namely, disposition effect, herd behavior, availability heuristic, gambler’s fallacy and hot hand fallacy, on the mechanism of stock market decision-­‐making, and, in particular, the individual differences in the degrees of these effects. Employing an extensive online survey, we document that on average, active stock market investors exhibit moderate degrees of behavioral biases. Furthermore, we find that, on the one hand, more experienced investors are less affected by behavioral patterns, yet, on the other hand, professional portfolio managers do not behave, in this respect, differently (more rationally) from non-­‐professional investors. We, therefore, infer that investor’s experience in stock market matters, but not the ‘status’ itself of being a professional, may decrease the effect of behavioral biases on her. In addition, we detect that the major ‘rationalizing’ effect of experience is already accumulated in the first years of investors’ stock market activity. Finally, we document that female investors are more strongly affected by all the five behavioral biases. Author(s): Shlomit Hon-­‐Snir; Andrey Kudryavtsev; Gil Cohen Public Good Contributions Among Coffee Farmers In Costa Rica: Cooperativists And Private Dealers Astrid Hopfensitz Toulouse School of Economics ABSTRACT: In this paper we present an experimental study on coffee farmers in Costa Rica that aims at investigating how experience with real world institutions (members of a cooperative vs. free market traders) shapes decisions in an unrelated anonymous decision situation. Coffee farmers participated in experimental public good games with partners drawn either from their own or a different background (members of a cooperative or free market participants). We observe that while free market participants are largely uninfluenced in their decisions by the identity of their partners, cooperativists strongly react to it. We further predict behavior of cooperativists in the experimental game through real world behavior, experience and motivations for cooperative membership. We specifically observe that labels that create specific institutional constraints (‘fair trade’, rainforest alliance) have strong effects on decisions. Author(s): Astrid Hopfensitz; Pepita Miquel-­‐Florensa Price adjustment processes in experimental asset markets with distinct fundamental value regimes. Juergen Huber University of Innsbruck ABSTRACT: The enormous costs imposed on society by the recent financial crisis highlight the importance of understanding conditions under which markets are efficient and when they may not be efficient. Before potential major changes in regulatory frameworks of real markets are implemented (e.g., financial transaction taxes, permitting/prohibiting short selling), it is prudent to explore possible consequences. Multi-­‐period laboratory asset markets are a suitable tool for this exploration, as they expose real humans to the market setting or regulatory regime under consideration. But inferring likely consequences of various regulatory instruments on markets also requires a profound understanding of the dynamics of the experimental asset market model employed. Here we analyze price adjustment processes in multi-­‐period laboratory asset markets with five distinct fundamental value (FV) regimes. We find that (i) underreaction of price changes to changes in the FV is a common characteristic of all regimes with uctuating FVs. Consequently, (ii) markets with increasing FVs exhibit undervaluation (negative bubbles), while markets with decreasing FVs exhibit overvaluation (positive bubbles). We test whether anchoring on past prices and orders distracts subjects’ attention away from FV information preventing complete price-­‐adjustment. In fact, (iii) integrating FV information on the trading screen significantly reduces mispricing in settings that previously exhibited strong mispricing. Author(s): Juergen Huber; Michael Kirchler; Thomas Stoeckl Implicit Contracts, Unemployment, and Labor Market Segmentation David Huffman Swarthmore College ABSTRACT: This paper provides evidence that two important features of labor markets the existence of involuntary unemployment, and the segmentation of markets into rms o ering \good’ and \bad’ jobs may have a common underlying cause. In particular, in the prevalent case that work e ort is not veri able, the implicit contracting strategies adopted by rms may simultaneously generate involuntary unemployment, and labor market segmentation. We develop a simple model that illustrates these mechanisms, and we empirically test for a causal impact of contractual incompleteness on unemployment and market segmentation, using experimental labor markets that differ in the veri ability of work e ort. Our data demonstrate that involuntary unemployment is much higher when explicit contract enforcement is not feasible. Moreover, we show that the necessity to provide implicit performance incentives can lead to a segmentation of the labor market. Firms in both segments earn similar pro ts, but workers in the secondary sector face much less favorable conditions than their counterparts in primary-­‐sector jobs. Author(s): Steffen Altmann; Armin Falk; Andreas Grunewald; David Huffman Reciprocity towards Groups in a Public Goods Game David Hugh-­‐Jones University of Warwick ABSTRACT: Experiments have shown that humans are reciprocators, who wish to reward others’ good behaviour and punish their bad behaviour. Field studies of conflict report episodes of mutual revenge between groups. In two-­‐round public goods games, our subjects reciprocated to low contributions by their first round partner by contributing less to another subject in that partner’s group. This group reciprocity was strongest when subjects felt that the other group had been unfairly advantaged. Group reciprocity was motivated by preferences, as well as by expectations about outgroup member’s future behaviour: subjects reciprocated towards outgroup members who were unable to make choices affecting them. Our design identifies group reciprocity cleanly, controlling for generalized reciprocity, copying, and in-­‐group favouritism. We test for third-­‐party group reciprocity (responding to harm done to other members of one’s own group), but find no evidence for it. We also examine interesting comparative statics. Subjects with high group identity group-­‐reciprocated more. Men did not group-­‐reciprocate more than women. Lastly, in contrast to the psychology literature on superordinate group identity, allowing subjects to work cooperatively with members of other groups did not appear to reduce group-­‐reciprocity. Our results demonstrate a psychological mechanism leading from injustice to intergroup conflict. They will be of interest to experimental and behavioural economists, psychologists and social scientists studying intergroup behaviour. Author(s): David Hugh-­‐Jones; Martin Leroch Beliefs and (In)Stability in Normal-­‐Form Games Kyle Hyndman Maastricht University ABSTRACT: In this paper, we use experimental data to study players’ stability in a variety of normal-­‐form games where subjects have to report beliefs and to choose actions without feed-­‐back. We implement several sets of similar games presented under equivalent strategic form and spread over two days. We find a fair degree of stability in players’ stated beliefs within sets of identical games, although time and changes in the presentation of the game did lead to less stability. Using the framework of the level-­‐k theory, beliefs tend to be less comparable across different games. We then estimate a structural model in which players can stick to a specific model of mind of their opponent, or switch between different appreciations of their opponents’ depth of reasoning in different games. Results show that a fraction of players consistently assign a low level of reasoning to their opponent, while others alternate between different assessments of their opponent’s cognitive skills. Author(s): Kyle Hyndman; Antoine Terracol; Jonathan Vaksmann Personal values and moral motivation: disentangling moral integrity and moral hypocrisy Bernd Irlenbusch University of Cologne ABSTRACT: Positioning moral motivations within the framework provided by Schwartz's (1992) values theory, we ran three dictator game studies (total N = 256) investigating moral integrity and moral hypocrisy. We adapted Baton’s (et al., 1997; et al., 1999; et al., 2002) landmark research design into the experimental economics laboratory (Study 1), and showed that the behavioral inconsistency out of 64 dictators, all 26 who chose to flip a coin to determine the allocation of money ended up with the self-­‐favoring outcome revealed in such a design is indeed indicative of dishonest claims to morality (arguably the core of moral hypocrisy), and not overpowered moral integrity. Supporting this interpretation, dictators who masked their selfishness behind the coin flip were motivated by high Conformity values (Study 1), and thereby similar to participants who made more obviously disingenuous claims to morality (Study 2). Further, dictators did generally not select the coin flip in case the result could not be rigged (only four out of 32 dictators did this; Study 3). Universalism and Benevolence values were predictive of moral integrity (Studies 1 and 3). Morality ratings of behavior generally showed both self-­‐serving and outcome bias. Author(s): Jan-­‐Erik Lonnqvist; Bernd Irlenbusch; Gari Walkowitz Yeast Mark Isaac Florida State University ABSTRACT: In a previous paper, an intriguing result arose when we created a class of participants (‘agents of grace’)-­‐-­‐-­‐ in a combined VCM and endogenous taxation (median voting) environment-­‐-­‐-­‐ who had a dominant strategy to provide the public good through voluntary provision. The results were suggestive of, but not conclusive, that the agents of grace affected the behavior of the other participants in terms of their willingness to donate through the VCM. The purpose of the experimental design reported here is to streamline the previous design to focus upon the agents of grace and thus to add some clarity to the previous results. We report results from both a pure VCM experiment and one, as in the previous paper, in which the VCM is combined with an endogenous voting institution. Author(s): R. Mark Isaac; Douglas A. Norton Solicited and Voluntary Promises Huseyn Ismayilov Tilburg University ABSTRACT: We implement a trust game with structured two way pre-­‐play messages. First, the trustor writes a free-­‐form message to the trustee. After receiving the message from the trustor, the trustee replies to the trustor. More than half of the trustors solicit a promise from the trustee. When solicited the vast majority of trustees make a promise. We compare promise keeping rates after solicited promises and after non solicited (voluntary) promises. We, also, run a baseline treatment with one way preplay messages from trustees to the trustor. Author(s): Huseyn Ismayilov; Jan Potters Preference and Belief Imprecision in Games Andrea Isoni University of Warwick ABSTRACT: Many experimental studies have found that behaviour in simple one-­‐shot games is inconsistent with the assumption that strategy choices are best responses to equilibrium beliefs. These findings have been explained either as best response to non-­‐equilibrium beliefs as in level-­‐k and Cognitive Hierarchy models or as equilibria that reflect noisy preferences as in Quantal Response Equilibria. We investigate to what extent failure to best respond to stated beliefs is the result of preference and/or belief imprecision. We elicit belief ranges and confidence in strategy choices in four classes of one-­‐shot 2x2 two-­‐person games. Our measures of imprecision show a substantial degree of sensitivity to parameter changes, both within and between game structures. Best response rates are higher when players are more confident about their strategy choices, and for games in which belief ranges are relatively narrow. Author(s): Andrea Isoni; Anders Poulsen; Robert Sugden; Kei Tsutsui Local leadership and the provision of public goods: A field-­‐
laboratory experiment in Bolivia B.Kelsey Jack Tufts University ABSTRACT: This study examines the effect of authority contributions on public good provision in a field-­‐lab experiment in rural Bolivia, where community members pool resources to provide environmental education material for local schools. We compare average contributions to the public good when an elected authority makes an initial public contribution to two types of controls: (1) a randomly selected community member makes an initial public contribution and (2) all contributions are private. Results show that total contributions are higher when elected authorities lead by example. This is due both to the fact that authorities give more when called upon to contribute publicly and because followers are more likely to make a high contribution when an authority contributes a high amount first. Similar influence is not seen among random individuals called upon to lead. To test whether these results are driven by a differential signaling value from authority contributions, we induce exogenous variation in follower’s information about the quality of the public good. Uninformed followers make high contributions more often following an informed leader’s high contribution. This effect disappears when followers of a randomly selected leader are themselves informed about the quality of the public good. A rich set of household and individual characteristics from a census of all households in the study communities allow us to explore the role of social status, social capital and self-­‐
selection in to the experimental setting. Author(s): B. Kelsey Jack; Maria P. Recalde A Case for the Use of Perception Measures Salar Jahedi University of Arkansas ABSTRACT: This paper utilizes corruption data from a laboratory experiment in order to examine the e effectiveness of perceptions as valid empirical measures. Our results indicate that perception-­‐
based measures correlate highly with the true levels they are designed to capture. More importantly, the results show that perception-­‐based measures convey additional information which is complementary to that obtained from facts alone and that the use of these measures may actually be preferable to the use of fact-­‐based measures even when facts are readily available and can be measured perfectly. Our results also indicate that self-­‐reported data deviates away from actual facts in systematic ways similar to those observed for the case of perceptions. Author(s): Salar Jahedi; Fabio Mendez Endogenous Public Goods Institutions: Who Likes to Punish? Julian Jamison Boston Fed ABSTRACT: We extend the standard public goods game in a variety of ways, in particular allowing for endogenous preference over institutions and studying the relationship between individual types, their preferences, and later behavior within the various environments. We collect individual data on a variety of demographic factors, in addition to measuring levels of risk aversion and ambiguity aversion (over both gains and losses). We then elicit preferences in an incentive compatible manner over voluntary contribution mechanisms both with and without reward and punishment options. Finally we randomly assign subjects to one of the four institutions and observe repeated play. We find that payoffs are significantly greater when punishment is allowed, but only a small minority of participants prefers such an environment. There is at most a weak link between individual characteristics and elicited preferences over environments. On the other hand, institutional preferences, as well as individual characteristics, are more strongly predictive of behavior in the public goods game. For instance, loss averse individuals pre-­‐emptively reward more often when that option is available. This suggests that when studying social interactions, especially if people can choose whether to participate, it is important to consider individual attitudes toward risk and uncertainty. Author(s): Michalis Drouvelis; Julian Jamison Eliminating Laboratory Asset Bubbles by Paying Interest on Cash Janet Jiang Bank of Canada ABSTRACT: The seminal work of Smith Suchanek and Williams (1988) finds price bubbles are frequently observed in an experimental asset market where a single asset with a finite lifetime is traded. Ever since, many studies have been carried out to understand the reason why bubbles occur in such an environment and to find mechanisms to eliminate bubbles. In this paper, we introduce an interest-­‐bearing savings account to the experimental asset market. We find bubbles disappear with high interest rates. The effect of the interest rate potentially works in two ways. First, the savings account increases the opportunity cost of buying shares, which in turn, reduces the incentive to speculate and alleviates the ‘active participation’ problem as raised in Lei, Noussair and Plott (2001). Second, fixing the dividend process and terminal value of the asset, the time trend of the fundamental value of the asset becomes positive with a high interest rate. An increasing fundamental value is more compatible with subjects’ perception that asset prices tend to be flat or increasing in the long run. Therefore, subjects are more likely to follow the fundamental value when they trade and over-­‐pricing is lessened. To disentangle the effects through the two channels, we run a second set of experiments with high interest rate but a lower terminal value to induce the fundamental value of the asset to decrease over time. Bubbles reappear in these sessions, which suggests the time path of the fundamental value is more important for reducing bubbles. Author(s): Giovanni Giusti; Janet Hua Jiang; Yiping Xu Fairness, Cooperation and Productivity Shuguang Jiang Shandong University ABSTRACT: Distribution fairness and its relationship with economic performance is a lasting hot topic in Economics. I design a model and experimentally investigate the effect of fairness perceptions on cooperative production using a two-­‐phase typing game. I suppose people have fairness preference and in some circumstances they can share the same view of what is a fair distribution. An unfair treatment can cause negative emotion reactions to people and consequently do harm to cooperation and productivity, while behaviors between those who get advantageous and disadvantageous unfair treatment should be different. In the first phase game, participants were asked to type Chinese characters into computers in order to earn money for their group. Then we gave one group a fair distribution and the other an unfair distribution following our definition of fairness. In the second phase, people in both groups were facing identical situation that the total group outcomes would be distributed equally among all the members. We found significant difference of productivity rate of change from Phase 1 to Phase 2 between the two groups. Particularly, being treated fairly induced good performance while being treated unfairly induced terrible performance in second phase production. We also found in the unfairly treated group the disadvantaged ones who typed more correct words but received fewer incomes displayed worse in the second phase. While in the fairly treated group those who produced more and received their just earnings showed better performance than their counterparts in the second phase. Author(s): Shuguang Jiang; Roland K. Cheo Increasing employees's work-­‐related stability in the lab Natalia Jimenez Jimenez University of Granada ABSTRACT: This paper compares the performance of three different types of contracts. The novel aspect of our experimental design is that workers have the chance of investing the money obtained from the labor market in order to increase their profits. The two first types are the standard period-­‐
by-­‐period contract and a permanent one. We also propose a new type of contract (automatic renewal) in which workers are re-­‐hired if their effort level is equal to the desired effort given by the firm. We find that efficiency (measured as the sum of both workers and firms’ profits) is maximized with the ‘automatic renewal’ contract. Results also show that the presence of dismissal barriers institutions in the labor market may help workers to make investment decisions more accurately, because subjects wait for a safer job position. Author(s): Ramon Cobo-­‐Reyes; Natalia Jimenez; Juan A. Lacomba; Francisco Lagos Media Effects in Persuasive Communication Paan Jindapon University of Alabama ABSTRACT: Based on Crawford and Sobel'ss (1982) model of strategic communication, Cai and Wang (2006) find that experimental subjects consistently overcommunicate, i.e., the sender’s messages and the receiver’s actions are more correlated than predicted by theory. Groseclose (2011) reexamines Cai and Wang’s data and identifies media effects on the receiver’s actions. In particular, when sender bias is positive, the average action of the receivers is statistically higher than the average state of the world. Since we are interested in media persuasion, unlike Cai and Wang’s experiment, the receiver’s action in our model is either accept or reject a policy recommended by the sender and the sender’s incentive is private information. Unlike standard cheap-­‐talk games, our persuasive communication game has a unique perfect Bayesian equilibrium and it allows us to separate communication and persuasion. The results from our experiment suggest (i) overcommunication, biased sender’s messages are not as exaggerated as predicted, and (ii) underpersuasion, the receivers do not accept the sender’s policies as often as predicted by theory. We estimate a two-­‐parameter version of McKelvey and Palfrey'ss (1998) Agent-­‐Quantal Response Equilibrium (AQRE) and find that the receivers overweight the probability that a sender is biased. As a result, the receivers overdiscount the sender’s messages and underpersuasion is present. Since the senders believe that they are more likely to receive distorted messages than they actually do, the biased senders will include less exaggeration in equilibrium and that explains why overcommunication occurs in the experiment. Author(s): Paan Jindapon; Carlos Oyazun How to improve patient care? An analysis of pure and mixed incentive schemes for physicians Nadja Kairies University Duisburg-­‐Essen, Chair of Quantitative Economic Policy (Prof. Brosig) ABSTRACT: In recent years, several countries have replaced pure incentive schemes for physicians (fee-­‐for-­‐
service, capitation) by so-­‐called mixed incentive schemes. Until now it is an open question whether patients are really better off after these healthcare reforms. In this study we compare the effects resulting from pure and mixed incentives for physicians under controlled laboratory conditions. Subjects in the role of physicians choose the quantity of medical services for different patient types. Real patients gain a monetary benefit from subjects’ decisions. Our results reveal that overprovision observed in fee-­‐for-­‐service schemes and underprovision observed in capitation schemes can, in fact, be reduced by mixed incentives. Though, the magnitude of reduction is significantly affected by the patient type. Moreover -­‐ and in contrast to the theoretical prediction -­‐framing pure incentives as mixed incentives already significantly affects physician’s behavior. Our findings give some valuable hints for designing healthcare reforms. Author(s): Jeannette Brosig-­‐Koch; Heike Hennig-­‐Schmidt; Nadja Kairies; Daniel Wiesen Alternation Bias and Reduction in St. Petersburg Gambles: An Experimental Investigation Kim Kaivanto Department of Economics, Lancaster University Management School ABSTRACT: The Reduction of compound lotteries is an implicit assumption both in the statement of the St. Petersburg Paradox as well as in its resolution by Expected Utility (EU). Yet despite the pivotal role of this assumption, to date there has been no empirical substantiation of its validity. Here we report three real-­‐money experiments in which the standard compound-­‐lottery form of the (truncated) St. Petersburg Gamble is explicitly juxtaposed with its reduced form. In the first experiment, we elicit Subjects’ Certainty Equivalents for each form of the gamble. In the second experiment, Subjects choose between reduced and compound forms within a multiple price list format, where a different sure payment (in 1 Euro increments), is added either to the reduced or the compound form. With this instrument, we can test for both `weak-­‐form’ and `strong-­‐
form’ violations Reduction. The third experiment replicates the second and then checks for robustness against range and increment manipulation. In the first experiment we find that the Certainty Equivalent of the compound form is stochastically dominated by, and significantly smaller than, the objectively equivalent reduced form. This bias toward the reduced form is borne out in the second and third experiments, where 90%-­‐-­‐100% display weak-­‐form violation and 48%-­‐-­‐87.5% display strong-­‐form violation. These results are consistent with the operation of alternation bias, which may be understood as a subjective distortion of conditional probability. Together these experiments offer evidence that the Reduction assumption may have limited descriptive validity in St. Petersburg Gambles. Author(s): Kim Kaivanto; Eike B. Kroll An Experiment on the Causes of Bank Run Contagions Todd Kaplan University of Exeter ABSTRACT: We conduct a bank run experiment in a modified Diamond-­‐Dybvig setup with two banks (Left and Right). To understand the mechanisms behind bank run contagion, we consider two treatments where liquidity levels are either linked or independent. Left Bank depositors see their bank’s liquidity and make withdrawal decisions first. Right bank depositors only see Left Bank withdrawals before deciding. Left Bank depositors run more often with low liquidity. Right Bank depositors run more often the more Left Bank withdrawals in both treatments, especially with linked liquidity. This effect is driven by increases (as opposed to decreases) in Left Bank withdrawals. Author(s): Surajeet Chakravarty; Miguel Fonseca; Todd Kaplan Strategic Behavior in an Open and Sequential Vote: An Experimental Test of a Bounded Rationality Model Ryan Kendall U.C. Irvine ABSTRACT: Kendall (2012) uses the Quantal Response Equilibrium (QRE hereafter) to identify a middle-­‐
mover advantage in an open and sequential vote. This paper presents results from a laboratory experiment designed to test this result. The Experimental data reveal a middle-­‐mover advantage. The data also unearth a behavioral richness that is not captured by the QRE. Specifically, subjects’ behavior shows evidence of sophistication, learning, and revenge. Adding these concepts, along with the QRE, allows for an equilibrium model that is reflected in the data. Author(s): Ryan Kendall Finding the Hidden Costs of Control Judd Kessler Harvard University ABSTRACT: An increasing literature has demonstrated that psychological factors (such as fairness, trust, norms, and intrinsic motivation) interact with explicit incentives (including fines and monitoring regimes) in complicated ways. Empirical and experimental evidence has demonstrated that these explicit incentives can undermine intrinsic motivation and lead to less prosocial behavior. For example, Falk and Kosfeld (2006) find that average agent effort in a principal-­‐agent setting is lower when the principal controls the agent by imposing a minimum action than if she leaves the agent free to choose any action. However, Kessler and Leider (2012) show that when prosociality is supported by informal agreements, having a mutually agreed upon minimum action often does not diminish effort. In this paper, we vary the contractual relationship between principal and agent to determine when Falk and Kosfeld's ‘hidden cost of control’ result undermines prosocial behavior and when it does not. We replicate their result and then demonstrate that the hidden cost is eliminated when the binding minimum affects both parties and reversed when both parties must agree upon the minimum. Author(s): Judd B. Kessler; Stephen Leider The Impact of Burden Sharing Rules on the Voluntary Provision of Public Goods Martin Kesternich ZEW GmbH ABSTRACT: We investigate how burden sharing rules may impact the voluntary provision of a public good which generates heterogeneous benefits to agents. We compare different rule-­‐based contribution schemes that are based on the principle of the smallest common denominator: all agents can suggest a minimum provision level of the public good that is allocated across agents according to some predetermined rule. We find that rule-­‐based contribution schemes significantly increase payoff levels relative to the VCM. Importantly, the equal-­‐payoff rule Pareto-­‐dominates all other rules. This holds in particular relative to a scheme where different types of players separately can determine their minimum contribution levels. Our results lend insights not only into efficient institutional design for voluntary private provision of public goods, but may also inform the recent climate policy debate on whether to have small agreements among more homogeneous players instead of having one comprehensive Kyoto style agreement that creates complicated burden sharing issues. Author(s): Martin KESTERNICH; Andreas LANGE; Bodo STURM The Endowment Effect as a Trading Strategy: Evidence from Experimental Markets across Cultures Joyce Wu Keyu Columbia University ABSTRACT: People exhibit asymmetries when valuing items for exchange: owners appraise their possession much higher than prospective buyers do. The origins of these asymmetries, or endowment effect, have been fiercely debated: some attribute them to fear of loss that pushes owners to demand high prices for their property (Kahneman, Knetsch, & Thaler, 1990), while others view them as outcomes of experimental design (Plott & Zeiler, 2007) or cultural differences (Maddux et al., 2010). It is difficult to settle the debate because many studies may affect prices through elicitation, use imperfect markets, and/or limited to a single culture. Analyzing thousands of trading decisions from experimental markets in several cultures, we show that such asymmetries are more widespread than assumed and benefit sellers by raising prices. To assess the competing explanations for valuation asymmetries, we utilized an experimental spot asset market where participants trade stocks to earn money. We find that asymmetries appear even before traders possess the stocks. And although the stocks are both virtual and meant to be traded, the asymmetries persist in actual trading. Because offers are much farther from intrinsic values than bids are, the markets suffer from reduced liquidity that causes prices to rise. Rising prices benefit sellers, and therefore pricing according to the endowment effect may be a rational strategy rather than an emotional reaction. Author(s): Sheen S. Levine; David Stark; Joyce Wu Keyu Gender Differences in Preferences at a Young Age? Experimental Evidence from Armenia Karen Khachatryan Stockholm School of Economics ABSTRACT: Many high-­‐profile, high-­‐earning occupations often take place in highly competitive settings where winners and losers are singled out and winners are disproportionately rewarded. A few recent experimental papers have proposed a new explanation for why women may be relatively under-­‐represented in those occupations. These papers suggest that women may systematically under-­‐perform relative to men in competitive environments and that many women, even among the most able, may simply prefer to stay away from such environments. A few natural questions revolve around the underlying determinants of these documented competitive differences. Are women simply born less competitive, or do they become so through the process of socialization? Do these differences depend on the nature of the experimental task? What is the role of risk aversion? To this end we examine gender differences in preferences for competition and risk in a large sample (825 subjects) of children aged 8 to 16 in Armenia. We include four types of tasks that vary at gender stereotyping when looking at competitiveness: running, skipping rope, math and word search. We find that: boys and girls are equally competitive at math and word search tasks along the whole age spectrum; younger boys and girls (8 to 11/12 year olds) are equally competitive at running and skipping rope, while older girls (12/13 to 16 year olds) are more competitive. Our data also suggest that gender differences in risk preferences emerge at the onset of puberty, with girls becoming more risk averse. Author(s): Karen Khachatryan Rational and Emotional Motivation of Crime -­‐ How Deterrence Policy Affects Criminal Behavior Menusch Khadjavi University of Hamburg ABSTRACT: This paper analyzes deterrence schemes and their impact on criminals stealing from victims by means of computer laboratory experiments. Importantly, the novel experimental design complies with marginal deterrence theory and offers interior predictions. Results confirm Becker’s deterrence theory and substitutability of punishment size and probability. Analysis of regime change shows that incentive history matters: when deterrence incentives are removed criminals steal more than without this history. Crowding out of social preferences is a common explanation for this result. This paper offers evidence that (at least part of this) crowding out takes place via change of emotionally motivated behavior. Without deterrence incentives in place, a variant of the dictator game, criminals with praising (pro-­‐social) emotions steal less from victims. When criminals face expected punishment pro-­‐social emotions do not decrease stealing; in this case self-­‐centered emotions motivate more stealing. These findings support theory that incentives both directly and indirectly impact behavior via rational and emotional motivation channels. Hence, this paper provides support for recent theory on behavioral criminal law and economics and offers new insights for deterrence policy. Author(s): Menusch Khadjavi Information Acquisition and Voting Mechanisms: Theory and Evidence SunTak Kim Univ of Pittsburgh ABSTRACT: This paper investigates the properties of optimal voting mechanisms, defined by group size and voting rule, with endogenous information acquisition. The standard model of jury voting with exogenous information predicts that the efficiency of group decision increases unambiguously with group size. However, once information acquisition becomes a costly decision, there is an important free-­‐riding consideration that counterbalances the information aggregation effect. Since information is costly and acquired information is a public good, rational voters will purchase information with a lower frequency. An implication of the trade-­‐off between information aggregation and free-­‐riding is that there exists an optimal group size for a given voting rule. We thus compare the efficiency of group decisions under different group sizes to test whether we can observe significant decreases in both information acquisition and efficiency as the group size moves from the optimal size to a larger group size. We also consider the optimality of a given voting rule. If information is exogenous, then we can make any voting rule (asymptotically) optimal by increasing the group size regardless of the precision of the information. On the other hand, with endogenous information acquisition, the above trade-­‐off implies that consensual voting rules are optimal only when the quality of acquired information is very high. We thus explore how the performance of different voting rules varies with the precision of information. Author(s): SunTak Kim; John Duffy; Sourav Bhattacharya Opportunity Cost and Overbidding in 2nd Price Auctions Erik Kimbrough Simon Fraser University ABSTRACT: In a recent paper, Lee and Malmendier (2011) provide evidence that bidders on eBay frequently pay auction prices in excess of simultaneously available, posted price options (Buy-­‐It-­‐Now prices), behavior which they argue stems from a failure to pay attention to these options. They term this example of apparent irrationality ‘the bidder’s curse’, but they do not explain the source of inattention. We provide experimental evidence that many examples of overbidding in auctions can be explained by the opportunity cost of researching the price of outside options. We find that the probability that a bidder overbids (and overpays) relative to an identical, simultaneously available, posted price option increases with the opportunity cost of researching the posted price. However, the losses from overpaying are more than offset by the gains from not researching the posted price. A similar opportunity cost argument can account for other puzzles such as the prevalence of jump bidding. Author(s): Erik O. Kimbrough; J. Philipp Reiss Behavior in Symmetric Public Goods and Common Pool Resource Games David Kingsley University of Massachusetts ABSTRACT: The failure of groups to cooperate and provide a socially acceptable level of public goods or to properly manage a common pool resource has tremendous importance to policymakers. While public good and common pool resource problems appear similar they are intuitively different. In a PG choice environment it is the voluntary contribution which benefits others while in a CPR choice environment it is the voluntary restriction which benefits others. In this paper symmetric PG and CPR games are used to explore behavior across these choice environments. Specifically, both games have identical interior Pareto optimal equilibriums, symmetric Nash equilibriums which are interior and equi-­‐distant from the Pareto optimal allocation and provide the same earnings potential. This research first explores the impact of providing explicit information regarding group earnings. When no group earnings are provided results suggest, consistent with the previous literature, that over the course of the game self interest dominates and cooperation decays. When group earnings are presented contributions increase in the PG game but there is no impact in the CPR game. However, in both games results suggest that the decay of cooperation is completely eliminated. Ongoing research explores the impact of informal sanctions and endowment heterogeneity in these choice environments. A fee to fine ratio of 1:3, implying that each player can pay 1 ED (Experimental Dollar) for each fine of 3 ED they choose to impose on other players, is used because this has been shown to effectively enhance cooperation in linear PG games. In the endowment heterogeneity treatments participants are shown that despite the initial heterogeneity it is possible to allocate their endowments such that equality of earnings across group members is ensured while continuing to maximize group earnings. Each endowment treatment is also ran with and without informal sanctions. Author(s): David C. Kingsley Waiting for the Greater Fool -­‐ Trader Inflow and Heterogeneous Information in Experimental Asset Markets Michael Kirchler University of Innsbruck ABSTRACT: We investigate the impact of trader inflow and heterogeneous information on bubble formation in experimental asset markets. To model heterogeneous information, the traded asset has two equally likely buyback prices. Half of the traders receive information solely about buyback price A, while the other half is only informed about buyback price B. In a 2x2 design we vary ‘trader inflow’ (‘yes’ or ‘no’) and cash-­‐asset ratio (‘constant’ or ‘increasing’) in four treatments. Similar to Miller (1977) and Harrison and Kreps (1978) we observe that prices can exceed the beliefs of the most optimistic traders. In particular, we find (i) strong price bubbles when new traders enter the market with cash only. (ii) When we mimic the same cash inflow over time without letting new traders enter, mispricing is significantly lower, which indicates that the reported effect is not driven by excess cash but by new traders. (iii) In markets with new traders entering over time, many ‘old’ traders follow a speculative motive as they frequently buy at prices exceeding even their highest belief of the buyback price. The latter was elicited using an incentivized questionnaire in each period. Similarly, we investigate subjects’ beliefs about future market prices. We observe that subjects expect future prices to rise even above overvalued prices at which they bought assets in the past. This indicates their hope to sell the assets later at even higher prices to new traders. Author(s): Caroline Bonn; Jurgen Huber; Michael Kirchler Communication and Trust in Principal? Team Relationships: Experimental Evidence Marco Kleine Max Planck Institute for Research on Collective Goods ABSTRACT: We study gift exchange in repeated interactions between a principal and a team of two agents, varying the information about the agents´ individual effort available for the principal. The principal is either fully informed about the agents´ individual effort contributions to team effort (‘FullInfo’), not informed at all (‘NoInfo’) or receives “cheap talk” messages about individual effort by the agents (‘Communication’). We find that gift exchange relationships can generally be established under all three conditions. Moreover, in line with recent evidence about the importance of the adherence to the equity norm for the agents´ work morale, agents exert substantially more effort in the ‘FullInfo’ treatment than in the ‘NoInfo’ treatment. In the former treatment, principals motivate their agents to exert high effort by conditioning wages on individual efforts such that the equity norm is adhered to. I.e. they pay higher wages to the agents who perform better than their co-­‐agents and equal wages for equal performance. In the ‘NoInfo’ treatment the principals violate the equity norm far more often with detrimental effects on the agents´ effort provision. The agents´ messages in the ‘Communication’ treatment are partially truthful, provide the principals with some additional information about individual performances and therefore contribute to reducing equity norm violations. Surprisingly, this positive aspect does not improve the agents´ work morale. On the contrary, the agents´ frequent misreporting spreads additional distrust in the principal-­‐team setting, resulting in substantially lower efforts and thus efficiency than in the other two treatments. Author(s): Marco Kleine; Sebastian Kube Bonuses for traders distort prices of assets. Daniel Kleinlercher University of innsbruck ABSTRACT: Whenever bubbles on financial, real estate or commodities markets burst, negative spillovers to the real economy occur. The meltdown of the subprime mortgage market in the United States, which unleashed a major financial crisis, is the latest example of such market failures. In this context, the G20, various academics, regulators and politicians have identified bonus payment systems of financial professionals as one reason for the development of the financial crisis. In this paper we evaluate the impact of four different incentive structures on risk taking and prices in laboratory asset markets. In particular, we selected the following trader incentive structures: bonus incentives, bonus incentives with cap, linear incentives and penalty incentives. We find excessive risk taking and massive overvaluation when subjects have bonus incentives. In contrast, when paid according to penalty incentives, subjects act very conservative with the result of prices falling below fundamentally justified values. Author(s): Juergen Huber; Michael Kirchler; Daniel Kleinlercher Meritocracy, Inequality and Corruption: an Experimental Study Reuben Kline Stony Brook University ABSTRACT: This experimental study-­‐-­‐conducted with student subjects at the University of Bologna and Stony Brook University-­‐-­‐ examines whether income inequality resulting from equitable and meritocratic institutions (i.e., high-­‐performers on an effort task earn more than low-­‐performers) leads to a lower incidence of corruption than inequality arising from inequitable institutions, defined as those in which low-­‐performers earn more than high-­‐performers. In both treatments high-­‐performers are anonymously paired with low-­‐performers. A lump sum, X, is then provided to each pair, with its division to be determined by a series of 25 binary signals (generated randomly from a symmetric binomial distribution). The realization of each signal is observed only by the high-­‐performer. A green signal indicates that the high-­‐performer is entitled to transfer a portion (X/25) of this sum to themselves, and red indicates that this same portion (X/25) should be transferred to the low-­‐performer. The actual transfer depends on the reporting of the signal(s) by the high performer, giving them an opportunity to be dishonest (corrupt) in their reporting and skewing the distribution in their own favor. The private nature of the signals ensures that the low performer cannot directly observe any dishonesty. We find that in the U.S. sessions the subjects are significantly more honest in the equitable treatment than in the inequitable one, but we find no significant difference in the Italian sessions, with levels of dishonesty roughly equal across treatments. Author(s): Dr. Reuben Kline; Fabio Galeotti; Dr. Raimondello Orsini Complementary Institutions and Economic Development: An Experimental Study Andrew Kloosterman New York University ABSTRACT: Why do some economies develop while others do not? One of the leading theories of economic development is that differences in development are due to differences in political and economic institutions societies choose as they develop. We model this development as a dynamic game of society first choosing institutions, and then ultimately playing a final developed society game, a game in which the set of actions available depends on the chosen institutions. This paper reports results of a laboratory experiment of this dynamic game. We find that while obtaining a good final game is difficult, one clear comparative static result is that patient players leave themselves with better final games. Author(s): Andrew Kloosterman; Andrew Schotter Experimental Evidence On The Robustness Of Signal Jamming Equilibria Brian Kluger University of Cincinnati ABSTRACT: The paper develops a simple signal jamming game in which a manager publically reports on an unobservable component of the terminal value of an asset prior to selling that asset in the market. The manager may be truthful or may misstate the value of the component, but must pay an expected penalty for misstating. Under rational expectations, the game has a unique signal-­‐jamming equilibrium (SJE) in which the manager’s report exaggerates the true value by a constant and the market corrects the manager’s report by this same constant. An experimental version of this game is developed and implemented using subjects interacting via a computerized platform. The behavior of the experimental subjects is roughly consistent with the SJE prediction in that, on average, managers exaggerate and market participants correct for the exaggeration. However, systematic deviations from optimal behavior are documented. In particular, (1) the extent of the manager’s exaggeration and the market participant’s correction is not independent of the value of the component as predicted, and (2) both manager’s and market participants use mixed strategies that are dominated by pure strategies. Using these observed behaviors, the model is re-­‐solved under bounded rationality, resulting in solutions that are more predictive, indicating that some of these behaviors may be optimal responses to non-­‐
rational behavior. Author(s): Brian D. Kluger; Steve L. Slezak Strong, Bold, and Kind: Self-­‐Control and Cooperation in Social Dilemmas Martin Kocher University of Munich ABSTRACT: We develop a model that relates self-­‐control, risk preferences and conflict identification to cooperation patterns in social dilemmas. We subject our model to data from an experimental public goods game and a risk elicitation task, and we measure conflict identification and self-­‐
control. As predicted, we find a robust association between self-­‐control and higher levels of cooperation, and the association is weaker for more risk-­‐averse individuals. Free-­‐riders differ from other contributor types only in their tendency not to have identified a self-­‐control conflict in the first place. Our model accounts for the data patterns at least as well as previous models. Author(s): Martin Kocher; Peter Martinsson; Kristian Myrseth; Conny Wollbrant How Feedback and Peer Pressure Affect Self-­‐Control at the Gym Yvonne Kohnle University of Konstanz ABSTRACT: We investigate the role of feedback and peer pressure on gym attendance and personal fitness in a field experiment. New members of a gym are randomly allocated to one of three treatments: Feedback, Peer Pressure and Control. Every treatment group decides how often they want to exercise at the gym per week (training goal). In the Feedback treatment, members receive feedback about their weekly gym attendance via email. In Peer Pressure, feedback is shared with a self-­‐chosen fitness buddy. In Control, members receive no feedback. We complement the experiment with a survey which additionally allows us to study how the treatments affect customer satisfaction and loyalty. Our design allows us not only to draw conclusions on how small nudges such as simple feedback affect gym attendance, but also whether it helps to increase club member’s health (e.g. body fat, strength and endurance performance). Author(s): Yvonne Kohnle; Simeon Schudy The (Dis)Incentive Effect of Relative Pay Comparison: Experimental Evidence Jenny Kragl EBS UniversitÀt fÃ_r Wirtschaft & Recht ABSTRACT: Relative pay matters. Evidence from the lab and the field suggests that relative pay comparisons strongly affect individual satisfaction and behaviour. More specifically, existing studies tend to confirm the conjecture that workers decrease efforts when they experience disadvantageous wage discrimination. Most of the work investigates gift-­‐exchange games with principals and fixed wage offers. Yet wages are typically not perfectly fixed because firms try to motivate their workforce by offering performance-­‐dependent pay components. Individual incentive schemes and abilities, however, often substantially differ among employees of the same firm. Moreover, incentive pay is usually tied to some imperfect performance measure that is not affected by the worker’s effort alone but also by some factors outside the worker’s control. As a result, performance pay is an important source of wage inequality among co-­‐workers. In the present paper, we experimentally investigate how relative pay comparisons affect individual effort incentives in a pay-­‐for-­‐performance setting. Using a real-­‐effort task, we find that workers of similar ability exert significantly less effort if they are informed about a peer’s payoff as compared to isolated agents. Our results suggest that social income comparisons have important effects on the efficiency of individual incentives even if performance is independent. Additionally, the effort sensitivity to social comparison seems to be more pronounced for workers with relatively high abilities. Our findings have important implications for optimal organizational and wage contract design. For instance, they may help explain firms’ preference for wage secrecy norms or wage compression. Author(s): Jenny Kragl; Petra Nieken Testing game theory without the social preference confound Michal Krawczyk Warsaw University ABSTRACT: We propose an experimental method whose purpose is to induce selfish behavior in games for a broad class of social preferences. It provides a theoretical framework for testing game theoretical predictions by confronting subjects with a commonly known payoff matrix actually representing their preferences. The paper describes the empirical tests of this method based on the comparison of results from several popular experimental games played with and without our methodology. Apart from it being a test of validity of the method, our experiment helps answer the question of how useful social preferences could be in explaining commonly observed deviations from selfish rationality. Results suggest that our method does induce more selfish behaviors: a substantial part of the difference between predictions based on selfishness and observed behaviors seems indeed driven by such preferences. But they also indicate that a considerable share is left untouched, perhaps giving weight to alternative explanations. Author(s): Michal Krawczyk; Fabrice Le Lec Turning a Blind Eye, but Not the Other Cheek Peter Kriss Carnegie Mellon University ABSTRACT: While costly second-­‐ and third-­‐party punishment have been demonstrated in numerous studies, the motives underlying such behavior are not entirely understood. In this paper, we study the strength of the underlying preference for punishing, by presenting those who have stated a willingness to pay to punish a transgressor with an excuse to avoid having to implement the punishment, and bear the corresponding cost. In separate treatments, we elicit both second parties’ (those directly impacted by the unfair action) and third parties’ (bystanders) preferences to punish (at a cost to themselves) those who act unfairly. Using a subtle experimental elicitation, we then allow the second and third parties to seek out or avoid the opportunity to enact their punishment decision without explicitly signaling to the other participants or to the experimenter that they are doing so intentionally. We find that second parties seek to be able to enact costly punishment, while third parties seek to avoid the ability to enact costly punishment. Our interpretation is that while second parties have an intrinsic preference to punish wrongdoers, third parties that choose to do so are often motivated primarily by second-­‐order concerns, such as social image or sense of obligation. Author(s): Peter Kriss; Roberto Weber; Erte Xiao Effects of design and sample size on the detection of social preferences in the field Sabine Kroger Department of Economics, Laval University ABSTRACT: This paper is concerned with the effects of design (both within and between subject designs) and sample size on the probability of detecting social preferences in experiments. The research question is inspired by studies on reciprocity in labor market field experiments that provide conflicting results concerning the effect of gift giving on worker effort. Those studies vary in their choice of experimental design, i.e., within and between subjects design, as well as in their sample size. We conduct a simulation study to investigate the effect that the choice of design and the sample size have on the ability to detect social preferences in experiments. The data generating process relates decisions to a treatment variable, individual unobserved heterogeneity, as well as idiosyncratic noise. We calibrate the parameters of the data-­‐
generating process using existing field experiments on social preferences. We compute power curves for both within and between subject designs and various sample sizes. Author(s): Charles Bellemare; Sabine Kroger Tax Compliance and Income Origination in the Laboratory: Do Windfall Gains Corrupt Pro-­‐Social Behavior? Eike Benjamin Kroll KIT -­‐ Karlsruhe Institute of Technology ABSTRACT: The present paper adds to the experimental studies on pro-­‐social behavior and norm compliance by highlighting the way individuals acquire the income that is subject to taxation. We distinguish between income that is the result of effort and performance, and income in the form of a pure windfall. We further consider the effect that is associated with the use of tax revenues. Tax revenues are either distributed evenly among participants or they are donated to an actual charitable cause. Our laboratory results show that tax compliance is significantly higher when the acquisition of income requires expenditure of effort and demonstration of performance. Windfall gains, in contrast, appear to strengthen the disposition to evade taxes. It seems that the appropriation of effortless income has a tendency to corrupt pro-­‐social behavior. Our results further suggest that taxes used for redistribution lead to significantly lower tax compliance than taxes used for public good financing. Author(s): Eike B. Kroll; Berthold U. Wigger Introducing the Drag-­‐Along Property to the Public Good Index with A Priori Unions Peter Kulasza ILE Graduateschool ABSTRACT: By alluding to the fact that the regulation of a good is strongly affected by the power of its users, I give up the symmetry assumptions of the Public Good Index. Taking this view, I model the relative strength of a player within the game by introducing the drag-­‐along property to the Public Good Index. Depending on the strength of a defecting player, others might follow the decision to abstain from the provision of the good. Power is then assigned proportionally to this drag-­‐along property. Author(s): Peter Kulasza Coordination in the Presence of Background Uncertainty Anthony Kwasnica Penn State University ABSTRACT: Many economic contexts require the coordinated actions of interdependent agents. For example, the profitability of investment is often based on the actions of other investors, and production in firms can require complementary inputs by several workers. These situations are often characterized by multiplicity of equilibria, confronting decision makers with strategic uncertainty. At the same time, such strategic choices are also often embedded in broader contexts with varying characteristics. For example, production and investment often take place in the presence of exogenous uncertainty about other important variables. While having little bearing on the strategic (best-­‐response) nature of the coordination problem, the presence of such ‘background uncertainty’ may have significant behavioral impact on the actions of individuals simultaneously confronted with strategic uncertainty. Here, we use a laboratory experiment to explore the degree to which such background uncertainty affects behavior in strategic coordination games. In our experiment, subjects in fixed groups repeatedly play a minimum-­‐effort (‘weak-­‐link’) coordination game. Across periods, payoffs are also determined by a binary random variable that potentially adds a constant to all player’s payoffs. This random component to payoffs does not affect the best response properties of the game, and therefore has little direct strategic relevance. In one (‘No uncertainty’) treatment, players find out the realization of the random variable before every play of the game. In another (‘Background uncertainty’) treatment, players do not observe the realization of the random variable until after they play the game. Thus, the distinction between treatments affects whether background uncertainty is present at the time subjects make strategic choices. We find that, in the presence of background uncertainty, choices are significantly less efficient in the coordination game and behavior converges to less efficient equilibria. We conclude that background uncertainty can affect behavior in contexts with strategic complementarities, to an extent that cannot be entirely accounted for by theory. Author(s): Anthony M. Kwasnica; Roberto Weber; Lan Yao A Game Theoretic Approach to Randomized Response: Theory and Experiments Ernest Lai Lehigh University ABSTRACT: This paper analyzes survey response as a game of strategic information transmission and reports findings from an experimental study of the randomized response survey method. This method uses random questions to elicit sensitive information from survey respondents. We model survey response involving sensitive information as a psychological game: a respondent answers a question about his type, and an interviewer updates her beliefs in light of the answer which affects the respondent’s payoff. The respondent’s incentives are shaped by a preference for truth-­‐telling and a distaste for being identified by the interviewer as a stigmatized type. When the distaste for stigmatization is sufficiently strong relative to the preference for truth-­‐telling, no information can be transmitted in equilibrium when the question is deterministic and known to the interviewer. However, when the question is random and not known to the interviewer, information can be transmitted no matter how strong the respondent’s distaste for stigmatization is. This is in line with prior findings in the theoretical literature on strategic information transmission that suggest that introducing noise can improve communication. Our experimental data confirm the equilibrium predictions: subjects truthfully reveal their types substantially more often in the randomized response treatments than in the deterministic response treatments. They also reveal their types more often when the induced distaste for stigmatization is relatively weaker. Author(s): Andreas Blume; Ernest Lai; Wooyoung Lim Unethical Minds: A Profile of Cheating Behavior Fabio Landini University of Siena ABSTRACT: We ran a field experiment to investigate which individual characteristics correlate with unethical behavior. We randomly approached passengers who were getting off the bus and we rewarded with a free bus ticket only those holding a valid travel document (either stamped ticket or subscription). For all passengers -­‐with or without a valid ticket-­‐ we collected a set of information, including age, gender, nationality, whether they were travelling alone and eventually with whom. Our results show that younger, male and foreign individuals are more likely to travel without a ticket. Moreover, we find evidence of group effects: while travelling with friends has no bearings on unethical behavior, travelling with relatives increases the probability to hold a valid ticket except for the time when most passengers are students. In addition, we asked additional questions to a subsample of passengers, finding that those who travel without ticket do so repeatedly, they perceive ticket controls as more frequent than those travelling with a valid ticket, and they are likely to interact with people also without ticket. Author(s): Alessandro Bucciol; Francesca Gino; Fabio Landini; Marco Piovesan The Endogeneity of Cooperation Preferences in the Provision of Public Goods -­‐ An Experimental Study on the Effects of Social Identity Matthias Lankau University of Göttingen ABSTRACT: The pattern of positive, yet decaying, contributions to public goods may be explained by (1) subjects’ heterogeneous cooperation preferences (frustrated interaction of conditional cooperators with free-­‐riders) and (2) by the fact that even conditional cooperators exhibit a self-­‐
serving bias (i.e. imperfect reciprocity). Consequently, this study is focuses on whether and to what extent social identity impacts on social preferences for reciprocity. While it has been shown in many different types of games that identity increases individual’s reciprocity towards in-­‐group matches, it is unclear in how far this holds for the public good environment. Accordingly, we devised a public goods experiment with induced social identity. Subjects played three one-­‐shot games in strategy method using random, in-­‐group and out-­‐group matching. Our results suggest that while the propensity to be a conditional cooperator remains stable under each treatment, we find a significantly elevated probability to be a free-­‐rider when matched with individuals of a different identity. We also investigated identity effects on the individual degree of reciprocity. Here, subjects show significantly less self-­‐serving bias when matched with in-­‐group members than with out-­‐group members. Additionally, even subjects initially identified as free-­‐riders behave in a conditionally cooperative pattern when matched with ingroup members. In conclusion, institutions that affect the salience of group member’s identity may trigger endogenous changes in cooperation preferences targeting (1) the individual degree of conditional cooperation and (2) the likelihood to be a free-­‐rider. Conversely, it can be reasonable to device policy institutions that enhance feelings of belonging to a particular group in order to achieve better social outcomes. Author(s): Matthias Lankau; Kilian Bizer; Marianna Bicskei Asset Integration and Attitudes to Risk: Theory and Evidence Morten Lau Durham University ABSTRACT: Measures of risk attitudes derived from experiments are often questioned because they are based on small stakes bets and do not account for the extent to which the decision-­‐maker integrates the prizes of the experimental tasks with personal wealth. We exploit the existence of detailed information on individual wealth of experimental subjects in Denmark, and directly estimate risk attitudes and the degree of asset integration consistent with observed behavior. The behavior of the adult Danes in our experiments is consistent with partial asset integration: they behave as if some small fraction of personal wealth is combined with experimental prizes in a utility function, and that this combination entails less than perfect substitution. Our subjects do not perfectly asset integrate. The implied risk attitudes from estimating these specifications imply risk premia and certainty equivalents that are a priori plausible under expected utility theory or rank dependent utility models. These are reassuring and constructive solutions to payoff calibration paradoxes. In addition, the rigorous, structural modeling of partial asset integration points to a rich array of neglected questions in risk management and policy evaluation in important field settings. Author(s): Steffen Andersen; James C. Cox; Glenn W. Harrison; Morten Lau; E. Elisabet Rutstrom; Vjollca Sadiraj Sponsoring Public Goods Lotteries in the Lab Michael LeGower University of Pittsburgh ABSTRACT: It has been shown recently that funding public goods through lottery mechanisms can secure higher levels of provision and greater social welfare than traditional voluntary contribution mechanisms. Traditionally in such models, however, the prize is taken out of the raised funds. While in equilibrium the raised funds cover the prize, off the equilibrium path the potential exists for the fundraising mechanism to actually lose money. As a result, real fundraisers often secure prizes from corporate sponsors before using this strategy. The goal of this paper is to examine an environment in which firms have an incentive to provide prizes to such lotteries and then investigate the choice of solicitation mechanism for the fundraiser. Specifically, I will examine the expected total prize offers from firms under three mechanisms: an exclusive fundraiser with fixed advertising, where only the most generous offer is accepted; an non-­‐
exclusive fundraiser with fixed advertising, where all offers are accepted but the most generous firm is rewarded disproportionately; and a non-­‐exclusive fundraiser with proportional advertising, where all offers are accepted and the firms share proportionally in the benefits. After solving for the equilibria of the game, I will test the predictions of the model in the laboratory, with subjects playing the roles of both firms, who decide if and how much to contribute to the prize pool, and consumers, who observe prize amounts and make contribution decisions. Author(s): Michael LeGower Peer Effects on Ideologies: Evidence from Classroom Random Assignments in College Fernanda Leon University of East Anglia ABSTRACT: This study investigates how peers affect young adults’ political behavior. The data comes from a self-­‐collected survey among freshmen from the largest university in Brazil over the election year of 2010. We explore the fact that students are randomly assigned to different classrooms to attend introductory classes, and test how the composition of classmates’ characteristics and preferences (by the time of their entrance in college) correlates with subsequently students’ behavior. The results show that interest in politics is contagious among classmates. This in turn makes students more informed and less parcial. They become more likely to have a balanced knowledge about the different candidates, less politically polarized and more politically homogeneous. Other mechanisms of influence like social pressure or reinforcement of own preferences are not found in the data. Author(s): Camila Campos; Fernanda Leite Lopez de Leon Personal Values Explain Differences in Cooperation across Individuals and Societies Sheen Levine Columbia University ABSTRACT: Humans are the champions of cooperation: people bear costs to help others, even strangers. Cooperation is essential to global challenges today as it was to human evolution, yet why we cooperate remains a puzzle. One explanation is that cooperation emerges in situations that reward and punish through reputation. Another explanation points to culture by showing that cooperation vary across societies. But people diverge in their cooperation even in the same situation in the same society, suggesting that situation and societal culture are incomplete explanations. Here we show that differences in cooperation across individuals and societies are explained by personal values. Analyzing thousands of cooperation decisions in a trust game, we provide and empirically validate a model that integrates situational and cultural explanations. According to the model, personal values not only predict cooperation, but also determine the relative influence of situation and culture in people’s cooperation. We define three types of cooperators based on their sensitivity to situation vs. societal culture. Understanding such types helps explain why some cultural groups are more cooperative than others, why some people cooperate more than others, and why some people alter their behavior as the situation changes while others even in the same society do not. Author(s): Wayne E. Baker; Sheen S. Levine Efficiency and gains from trade in an electronic call auction with insider trading -­‐ An experimental analysis Rene Levinsky Max Planck Institute of Economics ABSTRACT: The electronic call auction is a useful trading mechanism to aggregate dispersed information. Its ability to incorporate information of a single informed insider, however, is less well understood. We analyse this question by presenting a simple call auction game where in equilibrium auction prices reflect potential insider information. The predictions of the model are tested in the laboratory. While the likely presence of an insider improves the call auction outcomes in terms of increasing trading volume, uninformed traders fail to incorporate the (potential) insider information in their limit prices. Therefore informational efficiency is low relative to the theoretical benchmark. Moreover, we find that experimental subjects overvalue the asset. Author(s): Tobias Brunner; Rene Levinsky Information and Repeated Decisions from Experience Under Risk: an inquiry into bounded rationality Louis Levy-­‐Garboua University Paris 1 ABSTRACT: We study repeated pairwise decisions from experience under risk. Each choice is repeated 10 times, the random outcome of the preferred bet is observed and the satisfaction reported. We manipulate observation of counterfactual, choice (free or imposed for a sequence) and the freedom to choose the information regime. A single choice is randomly selected to determine payoffs. On average, decision makers switch 2.2 times during a sequence of ten choices and the rate of switching hardly declines over time. All subjects switched at least once. Subjects accept more risk under the full information regime. Choices exhibit a low efficiency loss relative to the expected-­‐gain-­‐maximizing choices but do not lose money relative to counterfactuals. Subjects prefer to observe the counterfactual even though this does not increase their gains. Thus, we find no evidence of regret avoidance. Switching to another bet is caused essentially by the satisfaction derived from past choice. Our data are consistent with a model of reinforcement learning. Under repeated decisions from experience, decision makers facing objective risk maximize a subjective expected utility in which the probability of winning the big prize is constantly revised in light of perceived information. Author(s): Louis Levy-­‐Garboua; Claude Montmarquette; Dorra Riahi Opt-­‐in or Opt-­‐out? Testing the Organ Donor Registration System in the Lab Danyang Li Georgia State University ABSTRACT: Numerous lives with organ failure could be saved by organ transplants. Despite the large need in transplantable organs, most U.S. residents over the age of 18 are not registered organ donors. This raises an important question for empirical research: what potential policy changes may increase the organ donation registration rate in the U.S.? This project utilizes a laboratory experiment to evaluate the relative effectiveness of alternative policies for human organ donation registration. In the experiment design, each subject lives a virtual life where the only decision is to be an organ donor or not. The experiment includes treatments across different default choices (opt-­‐in versus opt-­‐out) and organ allocation rules (without versus with donors’ priority rule) inspired by registration systems applied in different countries. Further, the experiment includes a controlled treatment to measure effects of neutral versus descriptive framing of the decision task on experimentally-­‐observed registration rate. Results from the experiment will inform discussion of possible changes in public policy towards organ donation. Author(s): Danyang Li; Zackary Hawley; Kurt Schnier Crowding-­‐out Effects in Social Preferences Jingping Li National University of Singapore ABSTRACT: Individual's social preferences are expressed in many forms, such as altruism, trust, reciprocity, fairness, equity, conformity to social norms and so on. As a result, they are measured using various experimental games. Previous studies mainly focus on one form of social preferences in a particular type of game, such as prisoner's dilemma, public goods games, dictator games, trust games and so on (Camerer and Fehr, 2004). Few studies allowed subjects to make more than one decision that reveals one’s social preferences in a more general way. In this experimental study, we focus on the conflicts of different forms of social preferences. By offering subjects more than one choices that might elicit different forms of social preferences (trust, reciprocity, and altruism), we observe how the subjects make trade-­‐offs between their own payoffs and the payoffs of others, and between the payoffs of different type of recipients. The result shows that individuals do not have a fixed share for oneself from the budget. The allocations between self and others mostly depend on who are the 'others'. Furthermore, within the allocation for 'others', different forms of social preferences (trust, reciprocity, and altruism) crowd out one another. Measures of social preferences of different forms imply that 'others' are not treated identically: altruism is less a concern compared to trust or reciprocity. Author(s): Jingping Li; Yohanes E. Riyanto Group Identity and Partnership Sherry Li University of Texas -­‐ Dallas ABSTRACT: In this study we present a laboratory experiment that measures the effects of induced group identity on partnership formation, effort, and division of surplus. In the experiment subjects face choices of forming business partnership that involves a real effort task with a hidden action. We find that agent’s effort and division of surplus are both significantly affected by their group membership. Compared to the control treatment without group categorization or the outgroup pairing, the division of surplus between ingroup partners exhibits less inequality; ingroup partners show a lower tolerance level for unfair offers but a higher wage elasticity of labor supply. This study sheds light on how group identity may interact with monetary incentives in labor contracts, and influence agent’s behavior and economic efficiency. Author(s): Jiang Jiang; Sherry Li Time is Money:A Natural Field Experiment on How a Non-­‐Monetary Timely Feedback System Motivates Volunteers Xiaoye Li National University of Singapore ABSTRACT: A natural field experiment is designed to explore the crowding effects of a timely feedback system on voluntary contributions according to their original dominant motivation (intrinsic/extrinsic). This paper finds crowding-­‐in effect of a one-­‐way timely feedback for both types of volunteers. However, a two-­‐way timely feedback has a detrimental effect for intrinsic volunteers (crowding-­‐out), while enhances extrinsic volunteers’ performance (crowding-­‐in). Furthermore, the crowding effect is robust to those aware of the feedback in advance or with access to the median score. These findings show the significance of original dominant motivation, since it will affect the change rate of intrinsic/extrinsic motivation. Author(s): Xiaoye li Moral Courage In The Laboratory Jannis Liedtke University of Hamburg ABSTRACT: We run a series of experiments to test the determinants of moral courage in the laboratory. In the baseline experiment, subjects are assigned to groups of four and endowed with 10€. Then, one subject is randomly selected to lose his endowment and the others can decide to help or not. The payoff for the ‘loser’ and the ‘helpers’ is the same. It increases in the number of helpers up to 10€, so that there are no efficiency gains from acting prosocially. Helping bears the risk of high loss. A coordination of several spectators is required to decrease this loss. By implementing these features, we create a dilemma between prosocial motives and risk aversion. We expect that, in the repeated game, varying norms will be established among groups. We test which of these norms found in the local group persist when two groups are combined as one global group and vice versa. Changing the sizes of the two combined groups to four and six players, we test whether the persistence or adaption of norms depends on the sizes of the groups. In addition, we investigate whether helping increases when a high degree of group identity has been formed in a quiz solved through teamwork. One treatment gives a time window for the decision; thus, our design also raises the question as to what kind of people initially start helping, contributing to the literature on the leadership of taking initiative. Author(s): Jannis Liedtke; Jan Schmitz Financial behavior of the homeless population: a field experiment Sera Linardi University of Pittsburgh ABSTRACT: This paper aims to understand the financial situation of transient populations and designing policies that would increase the effectiveness of social services in reaching them. We conduct our experiment at the largest homeless shelter in Arizona. The first part of this study focuses on the first stage where we conducted time discounting experiments and in-­‐depth survey with 51 shelter residents. We find that survey variables such as the gap between the ideal and predicted saving is strongly correlated with present bias elicited by the time discounting experiment. In the second part of this study, we introduce a competition for monetary rewards into the saving program. We randomly assigned our subjects into a control group and a competition group, where the subject that saved the largest portion of income in the first month of the competition receives $100. The results of the policy intervention suggest that the competition induced subjects to temporarily delay expenditures, resulting in an $89 increase in amount saved in the first month. Preliminary analysis of second month data seems to suggest that the increase in savings is temporary. However, we see strong evidence that savings is largely determined by income, and income is largely determined by number of meetings with case manager (possibly through better job leads). This suggest that the most effective way to increase savings is to help clients find better paying jobs and to meet regularly with case managers. Author(s): Sera Linardi; Tomomi Tanaka Multi-­‐round minority game strategy experiment Jona Linde UvA-­‐Creed ABSTRACT: Minority games are widely studied using simulations as well as laboratory experiments. The results of simulation analyses depend on the type of strategies used. So far experiments provided little guidance into the type of strategies people use because the set of possible strategies is very large. We use a strategy method experiment to directly elicit people’s strategies. These strategies are then subjected to an evolutionary competition. We find that the strategies people use are very heterogeneous although aggregate outcomes resemble the symmetric Nash equilibrium. The strategies that survive evolutionary competition achieve much higher levels of coordination than the complete pool of strategies. Author(s): Jona Linde; Joep Sonnemans; Jan Tuinstra Reward, punishment, or both? An experimental investigation of three tournament mechanisms Florian Lindner University of Innsbruck ABSTRACT: Tournaments are widely used in organizations, explicitly or implicitly, to reward the best-­‐
performing employees, e.g., through promotion or bonuses, to punish the worst-­‐performing employees, e.g., through firing or unfavorable job assignments, or in a combination of both types of incentives. We use a principal-­‐agent model to compare three tournament incentive schemes: reward tournament, punishment tournament, and reward & punishment tournament, which, respectively, reward the best performer, punish the worst performer, or reward the best and punish the worst performer. We show theoretically that the three schemes are equivalent when the agents are symmetric in their ability, independent of the number of agents. We then test our theory experimentally using a chosen effort task. With six agents the reward and the punishment tournament are equal in terms of total effort, but differ in their distribution; in the reward & punishment tournament agents choose significantly higher effort levels. In the three agents case we find that total effort is significantly lower in the reward tournament, compared to the other two schemes. Author(s): Loukas Balafoutas; Glenn Dutcher; Florian Lindner; Dmitry Ryvkin; Matthias Sutter Stabilising the Economy: Market Design and General Equilibrium Luke Lindsay University of Zurich ABSTRACT: We study the stability of competitive equilibrium in Scarf's (Int. Econ. Rev. 1 (1960) 157) economy in the laboratory under two market mechanisms. In this economy, Tatonnement theory predicts prices will be globally unstable. That is, unless prices start at the competitive equilibrium, prices will oscillate without converging to equilibrium. This prediction follows from the simple assumption that the price of a commodity increases if demand for the commodity exceeds supply and that the rate of change of the price is proportional to the excess demand. Anderson et al (J. Econ. Theory 115 (2004) 209) have found that in laboratory double auctions, prices in Scarf’s economy do indeed oscillate. We replicate these experiments and find Tatonnement theory predicts the direction of price changes remarkably well. Then we test a novel market mechanism where participants submit demand schedules. Using this new mechanism, instead of oscillating prices quickly converge to the competitive equilibrium. Furthermore, efficiency is significantly higher: 95 percent of the potential gains from trade are realised compared to 77 percent under the double auction. Author(s): Jacob K. Goeree; Luke Lindsay Opt-­‐In vs Opt-­‐Out in Voluntary Contribution Game: Do Default Options Make a Difference? Jia Liu Nanyang Technological University ABSTRACT: We evaluate the impact of non-­‐binding default options on cooperation in a setting whereby individuals collectively engage in the provision of public goods. In deciding whether to cooperate, individuals face a trade-­‐off between self-­‐ and collective interests. We consider two non-­‐binding default options; the opt-­‐out and opt-­‐in options. Under the former, by default, individuals agree to make non-­‐zero contributions, but they may opt-­‐out of the default arrangement. Under the latter, by default, individuals do not contribute, but they may decide to contribute if they want to do so. We show that the opt-­‐out option significantly increases the overall group contributions and the individual’s earnings relative to both the opt-­‐in option and the standard voluntary contribution arrangement. Societal welfare is also maximized under the opt-­‐out treatment. Author(s): Jia Liu; Yohanes E. Riyanto Crowdsourcing with All-­‐Pay Auctions: a Field Experiment on Taskcn Tracy Liu University of Michigan ABSTRACT: To understand the effects of incentives on crowdsourcing participation and submission quality, we conduct a field experiment on Taskcn, a large Chinese crowdsourcing site using all-­‐pay auction mechanisms. We systematically vary the size of the reward, and the presence of a soft reserve in the form of the early entry of a high-­‐quality submission. We find that a higher reward induces significantly more submissions and attracts higher quality users. However, unpredicted by theory, we find that high-­‐quality users are significantly less likely to enter tasks where a high quality solution has already been submitted, resulting in lower quality in subsequent submissions in such soft reserve treatments. Author(s): Tracy Xiao Liu; Jiang Yang; Lada A. Adamic; Yan Chen What can we learn from laboratory public good games about climate change mitigation: Evidence from an artefactual field experiment Johannes Lohse University of Heidelberg ABSTRACT: Climate change mitigation is one of the largest environmental public good problems present in the real world. The public goods literature has produced ample experimental evidence under which conditions public goods problems can be alleviated. However, much of this research relies on non-­‐representative samples of student subjects observed in abstract laboratory settings. Its external validity regarding the real world problem of climate change mitigation is therefore in question: Students are not only a special subpopulation that might behave differently than the general public, but it is also unclear if behavior in the abstract public goods game is a good indicator for real world behavior. We test these propositions experimentally by comparing students to members of the general public in an artefactual field experiment. We employ a standard public good game and a real contribution task in which participants can reduce CO2 emissions in a within subjects design. Furthermore we collect a general measure of cognitive ability. We find that non-­‐student subjects give higher amounts both to the abstract and the real public good. Hence observing students underestimates real world contributions. Additionally high cognitive ability is a strong predictor for high contributions in both tasks, which is a counterfactual to the common claim that deviations from Nash-­‐behavior in public games are due to mistakes. On the individual level we can reject that behavior in the laboratory public goods game is consistent with behavior in the contribution task. This casts doubt on whether the stability of preferences holds across different decision frames. Author(s): Timo Goeschl; Sara Elisa Kettner; Johannes Lohse; Christiane Schwieren A violation of the monotonicity axiom Sandra Ludwig University of Munich ABSTRACT: The monotonicity axiom underlies economic decision theory and ordinary inferences about people’s motivations. We test and reject the monotonicity axiom in a setting where choices involve wealth transfers. We do not find evidence that the rejection of the monotonicity axiom in our setting is driven by signaling or anchoring. Our results suggest that an announcement of a possible wealth transfer may affect preferences and dampen down sentiments of generosity. Author(s): Wolfgang Hoechtl; Rudolf Kerschbamer; Philipp Kircher; Sandra Ludwig; Alvaro Sandroni Rational Inattentiveness in a Forecasting Experiment Wolfgang Luhan University of Bochum ABSTRACT: Mankiw and Reis (2002) proposed the sticky-­‐information Phillips curve as an alternative to the standard sticky-­‐price Phillips curve. We test its microeconomic foundation, the rational-­‐
inattentiveness model in a controlled laboratory environment. Our design is a forecasting task with costly information and a clear cost-­‐benefit structure. While we find numerous deviations from the model predictions on the individual level, the aggregate results show clear evidence in favor of rational inattentiveness and sticky-­‐information models rejecting simpler behavioural heuristics. The only contra-­‐ diction found is a significant impact of the forecasting duration. Author(s): Henry Goecke; Wolfgang J. Luhan; Michael W.M.Roos Should feedback be public in trust game? Marianne Lumeau university de Rennes 1, CREM ABSTRACT: This paper examines the reasons for the relative efficiency of a reward and sanction system such as online decentralized public feedback systems on the emergence of cooperation. This could be due to two different effects. The first one is a reputational effect. In other words, agents would like to avoid the reputational cost of receiving a sanction. The second effect is emotional: agents seek to avoid disapproval from their peers. To distinguish between these two effects, we use an experimental approach comparing a public feedback system to a private feedback system in a trust game context. Our experimental results indicate that either one of these effects influences trust and trustworthiness behaviors. We then examine the role of agent’s emotions on their decision-­‐making. Based on self-­‐report of emotional reaction, our analysis shows that some specific emotions are associated with some agent’s decisions, such as anger and rating decision. Author(s): Marianne Lumeau; David Masclet; Thierry Penard Lying aversion and social preferences in children Valeria Maggian University of Padua ABSTRACT: We run an experiment to study the relationship between the development of social preferences and lying behavior in children. Besides contributing to the recent debate investigating the role played by age and gender in the development of social preferences, this study investigates how these characteristics affect the early emergence of dishonest behavior. In our experiment, 567 children, aged seven to eleven years, are matched in couples and play a modified version of the dictator game. After eliciting children’s social preferences, we give them the opportunity to cheat in order to obtain their preferred allocation option. We find that females are more likely to choose an equal payoff distribution both when the alternative one increases their own payoff at the expense of the other and when it is Pareto efficient. Our results show that the probability of cheating is uniform across age-­‐levels: a large fraction of children are reluctant to tell lies, independently of any social preferences over outcomes. Interestingly, a non-­‐negligible fraction of children are willing to tell an ‘envy lie’ that equalizes outcomes by reducing other’s payoff. Author(s): Valeria Maggian; Marie Claire Villeval Limited backward induction Marco Mantovani Universite Libre de Bruxelles ABSTRACT: This paper presents an out-­‐of equilibrium model for sequential games, Limited Backward Induction, and provides an experimental investigation of the subject. Players are characterized by limited foresight over the game tree and perform backward induction at every decision node only on a limited number of stages. The extent of their strategic reasoning is affected by how relevant the stakes are -­‐ i.e. how beneficial strategic reasoning can be -­‐ and how complex the game is -­‐ i.e. how costly strategic reasoning can be. The predictive power of LBI and its functioning is shown through examples. In the second part of the paper we run an experimental investigation of LBI using a variant of the Race Game. In the base game two players choose sequentially a number within a range. Those number are summed up until a target number is reached. The player who reaches the target wins a prize, the other gets nothing. One of the two players has always the opportunity to secure his victory at the very first stage using a dominant strategy. If he doesn’t, however, the same opportunity is given to his opponent at his next turn. We add treatments with small prizes at different distances from the target, by pursuing which the players are driven away from their dominant strategy and allow the players to claim victory over prizes during the game. The results support the LBI hypothesis and provide a number of other interesting insights on strategic reasoning in sequential games. Author(s): Marco Mantovani Losing Face An Experiment on Self-­‐Image and Image Reciprocity Lei Mao University of Lyon -­‐ GATE ABSTRACT: Theoretical models have shown that individuals value their self-­‐image and try to preserve it. But do people also care about preserving the self-­‐image of others? Does reciprocity exist when social interactions involve image and do not affect monetary payoffs? In this paper, we provide experimental evidence on the willingness of individuals to pay for preserving their self-­‐image and other’s self-­‐image when facing a risk of losing face. In the baseline treatment of our laboratory experiment, participants perform a real-­‐effort task in a triad and the least performer is publicly exposed, except if two triad members are sacrificing money to avoid public exposure. Our 2×2 design manipulates group identity and praise. In the praise treatment, the best performer in the triad is publicly praised after the least performer has been exposed. The main results reveal, firstly, a majority of the lowest ranked sacrifice money to avoid being publicly exposed. Secondly, individuals are also willing to preserve the image of others at a personal cost. Thirdly, such a sacrifice is more likely when group identity is salient and groups homogenous. Fourthly, the proportion of individuals who sacrifice to preserve other’s image is decreased in the treatment where public praise follows public shaming. Self-­‐reported feelings provide evidence of gloating: the best performers report a higher level of happiness when publicly praised after the least performer has been publicly exposed than in the absence of public shaming. Finally, we measure the impact on productivity of the anticipation of shame and praise. Author(s): Tor Eriksson; Lei Mao; Marie Claire Villeval Making the Headlines: Evidence from a Donation Field Experiment Marta Maras Bocconi University ABSTRACT: People act more generously when their donations are announced publicly. Where are the limits of publicity in encouraging people to give more? This paper contributes to the literature by exploring the varying degrees of publicity in a natural field experiment. A unique household dataset includes individual weekly donations to a fundraising campaign for a new neighborhood church of 1597 households over the 1994-­‐2000 time period. Individual donations collected the previous week were first publicly announced by the fundraiser at the end of the Sunday mass (117 weeks). In the following 106 weeks they were only posted on a board on the side church wall. Finally, both individual and total amounts donated so far by the household were announced publicly (112 weeks). Announcing the donations publicly significantly increased the number of donors and the aggregate weekly donations compared to only posting them. However, the average household donations remained unchanged. Households had an anonymous donation option, still the vast majority preferred to donate publicly. Significantly more donations were raised during holidays, but this holiday effect does not explain the increased public giving. Interestingly, the donation announcement order additionally raised the donations in the last and the most public treatment. The order was determined according to the alphabetical order of the street, so that the donations from households in street A were read earlier than those from households in street Z. Households that were announced earlier donated more on average. Key Words: Fundraising Behavior, Field Experiment, Public Goods, Social Interaction JEL Classification: C93, D64, H41, Z12 Author(s): Marta Maras; Bocconi University The Drop Out Effect Daniel Martin New York University ABSTRACT: When choice options are complex, policy makers may seek to reduce decision making errors by making a high quality option the default. We show that this positive effect is at risk because such a policy creates incentives for decision makers to "drop out" by paying no attention to the decision and accepting the default sight unseen. Using decision time as a proxy for attention, we confirm the importance of this effect in an experimental setting. A key challenge for policy makers is to measure, and if possible mitigate, such drop out behavior in the field. Author(s): Daniel Martin; Andrew Caplin Outsourcing in an Incomplete-­‐Contract Labor Market Patrick Martin University of Pittsburgh, Katz Graduate School of Business ABSTRACT: We report the results of experimental labor markets designed to test the impact of firms replacing higher cost workers with lower cost workers (i.e., outsourcing). The trend toward outsourcing appears to be driven by the financial benefit of moving operations to locations with lower labor costs. However, we predict that in an incomplete contracting environment, outsourcing may harm the existing positive reciprocal relationship between the firm and the remaining incumbent workers. The negative effect on the established positive reciprocal relationship with any remaining incumbent workers could impact firm profit, worker payoffs and overall social welfare. We find that firms are willing to hire new workers at a lower wage than incumbent workers, and they when they do, the positive reciprocal relationship with the remaining incumbent workers deteriorates. Specifically, when firms chose to outsource, both wages and effort decline for the remaining incumbent workers. As a result, hiring new workers at a lower wage does reduce labor costs, but does not result in greater firm profit. In addition, outsourcing results in lower wages for all workers and, consequently, lowers overall social welfare. Our results suggest that firms should consider both the wage savings and the effect on the relationship with its remaining incumbent workers when deciding whether to outsource jobs. Author(s): Jason Brown; Patrick Martin; Donald Moser; Roberto Weber Gender Differences in (Non) Competitive Environments: A New Experimental Evidence. David Masclet University of Rennes 1 ABSTRACT: We present a new experimental design that permits us to explore gender differences in both remuneration choice and performance. We design a game in which participants are asked to perform a real effort task under different remuneration schemes, either a flat wage or a two-­‐
player tournament remuneration scheme. We also allow participants to choose their remuneration scheme. To our knowledge, our paper is the first one to examine gender differences under a flat-­‐wage versus a tournament. Our data indicate that there exist some gender differences both in performance and remuneration choice. First, we find that females are more likely to choose the flat wage scheme than males. This finding seems to be due to the fact that females are both more risk averse and inequality averse than males. Furthermore, we observe that this gender gap tends to disappear with experience, which may be explained by male’s overconfidence. Second, we find that females provide significantly more effort than males under the flat wage scheme while males increase more their effort when a tournament scheme is introduced. These gender differences seem to be explained by differences in risk aversion. Author(s): Sophie Larribeau; David Masclet; Emmanuel Peterle Insurance, beliefs and affects: using psychophysic paradigm to explain insurance behaviors facing catastrophic risk Sebastien Massoni CES -­‐ University Paris 1 ABSTRACT: This experiment studies insurance behaviors facing catastrophic risk (i.e. low probability and high consequences events) in an emotional framing. Emotions are induced by house money effect and loss aversion. We use a perceptual task to predict subject’s performances and beliefs by Signal Detection Theory. In a first part, the subjects perform the task to accumulate their own rewards; then they play some bingo periods with gain or loss framings. To succeed the bingo, the subjects have to reach a minimal score over a fixed number of trials. In the loss bingo, they can lose all their money, whereas they can win it in the gain bingo. Facing these risky bets, the subjects can chose to take an insurance to obtain a sure amount of money (sure gain or sure loss). The probability of failure (1%, 10% or 25%) and the stakes (20 or 200‚¬) vary. We elicit the level of coverage by a BDM mechanism, the subjective beliefs by probabilities matching, and the level of concern by a scale of worry before each bingo. Preliminary results show that insurance behaviors are affected by the loss framing, the level of worry and the stakes. Low probability events lead to inaccurate levels of coverage. Interestingly, metacognitive abilities (calibration and discrimination) are positively correlated with more efficient insurance behaviors. Author(s): Olivier Chanel; Sebastien Massoni; Jean-­‐Christophe Vergnaud Stochastic Asymmetric Blotto Games: Theory and Experimental Evidence Alexander Matros Lancaster University Management School ABSTRACT: We consider an election model where two risk neutral candidates have to allocate their campaign budgets (for media advertisement) among several states. If candidates allocate different amounts for the same state, the candidate with higher amount has higher chance to win this state. In general, we assume that a winner of each state is determined stochastically by means of a lottery in which a candidate’s winning chance is proportional to her spending on this state. Different states can have different values (number of votes) and candidates might have different campaign budgets. Each candidate wants to maximize her chance to win a majority of all states. How should candidates allocate their budgets? How different the behavior of candidates with different budgets should be? Will candidates compete for all states? These and similar questions are analyzed in this paper. We find a unique Nash equilibrium of the Election game. There are two important equilibrium properties. First, candidates allocate their budgets in the same proportion. Second, spending on a particular state is proportional to ‘pivotality’ of this state. Friedman (1958) was the first to analyze a related advertising model in which both players allocate their budgets across all the items in order to maximize their expected payoffs. He also finds a unique Nash equilibrium. We compare our equilibrium prediction with the Friedman’s equilibrium. It turns out that Friedman’s equilibrium spending on a particular item is proportional to the item’s value. It means that players compete for all items in the equilibrium. It contrasts our finding that candidates spend resources only on pivotal states. Therefore, equilibrium behavior in two models might be significantly different. This observation is similar to Snyder’s (1989) result. Snyder (1989) also considers a model of a two-­‐party competition for legislative seats. He models a competition for each seat as a stochastic contest. The main difference between his paper and the present paper is that Snyder’s parties have to incur costs for their spending, unlike our model, in which the budgets are given. We take our model into the experimental laboratory to test whether subject’s behavior is consistent with the theoretical predictions. We observe strong support for theoretical predictions in the experimental laboratory. Author(s): John Duffy; Alexander Matros Positive Productivity Spillovers and their Effect on Voluntary Leadership in Teams Susanne Mayer Vienna University of Economics and Business ABSTRACT: Recent game theoretic articles include behavioral considerations and information asymmetries to account for the benefits of leading in team efforts. This paper, instead, examines a previously unconsidered rationale for voluntary leadership in the public goods provision. In our model, we prove that leadership may emerge as the first mover realizes that, through her own contributions, she boosts her follower’s productivity. By allowing for such positive productivity spillovers in the production technology of a public good, we show that a leader can effectively increase total contributions, yielding a Pareto-­‐superior outcome for all team members involved. To date, we are not aware of any experiment taking account of positive productivity spillovers as another explanation for the endogenous emergence of leadership. One may therefore also regard our theory itself as a potential trigger of positive productivity spillovers in experimental economics. Author(s): Susanne Mayer; Christian Hilbe The Emergence of Cooperation in an Adversarial Game: Experimental Evidence Michael McBride UC Irvine ABSTRACT: The evolution of cooperation has been widely studied, but with less attention given to directly adversarial settings. We experimentally study the emergence of cooperation in the adversarial game introduced by Short, Brantingham, and D’Orsogna (2010). We find the likelihood that cooperation emerges to be dependent on two factors: first, the costs and benefits of defection and punishment; and second, the presence of a behavioral type that simultaneously defects and punishes defectors. We develop more general lessons about the success rate of intervention policies intended to increase cooperative behavior. Author(s): Michael McBride; Ryan Kendall; Martin Short; Maria D'Orsogna Whoever you want me to be: Personality and Incentives Andrew McGee Simon Fraser University ABSTRACT: Employers often screen potential employees using personality profiles such as the Big 5 to identify workers with characteristics the firm values or traits possessed by prior applicants who have gone on to successful careers with the firm. The Big 5, however, was developed by psychologists to measure personality traits in an environment devoid of incentives that work against truthful reporting. Job applicants, however, have incentives not to respond truthfully but to make themselves appear desirable to employers. This experimental study examines how responses to the Big 5 personality profile change when the incentives facing the respondents change. Having completed the Big 5 personality battery prior to their arrival in the laboratory, subjects were incentivized to exhibit particular traits on a Big 5 personality profile administered in the laboratory. Our study examines (1) whether subject’s responses to the Big 5 change in a predictable fashion in response to incentives that may be at odds with truthful reporting and (2) whether any such changes are related to measurable factors such as intelligence. Author(s): Andrew McGee; Peter McGee From Rival to Boss: Promotions with Continued Interaction Peter McGee Ohio State University ABSTRACT: Tournament models of corporate promotion have been widely studied by economists because they have the appealing feature that even when output is noisy, principles need only to be able to observe relative output of employees. It is also well known that tournaments sometimes introduce perverse incentives to engage in inefficient behavior, such as sabotaging a rival. Existing tournament models, however, have stopped at the tournament’s edge. The combination of evaluation criteria that may be unclear to tournament participants and animosity from the competition may introduce feelings of injustice that spill over into the workplace when an employee is supervised by a former peer. When subordinates have the opportunity to withhold effort to punish a manager, tournaments may further distort effort provision. This paper presents the results of a laboratory experiment where feelings of procedural injustice are induced. Subjects first engage in a standard tournament with one other subject, followed by followed by a piece-­‐rate stage in which the winner from the tournament receives a fraction of the loser’s output in the second stage. We find that lingering feelings of injustice lead to a significant reduction in effort by tournament losers in the second stage that can only be explained as spiteful behavior. In additional treatments, we consider potential structural changes designed to minimize the counterproductive behavior, such as rotating managers, taking into consideration concerns about specific human capital. Author(s): Peter McGee; Andrew McGee Incentives and Group Identity Stephan Meier Columbia University ABSTRACT: Groups and teams are very prevalent in society and organizations. We investigate the impact of group membership on the effectiveness of incentives, i.e. whether incentives work differently when the ‘principals’ and ‘agents’ either belong and identify with the same group or with different groups. It has been shown that incentives can have detrimental effects, but those ``hidden costs of incentives’’ might critically depend on the relationship between the principal and the agent. Incentives might be perceived very differently by an agent if they were set by an ‘in-­‐group’ principal. Our experiment uses a stylized principal-­‐agent relationship in which the principal can incentivize the agent by imposing a control mechanism. We do not find evidence that the ‘hidden costs of incentives’ are larger in within than in between-­‐group matchings. However, the mechanism originating the hidden costs critically depends on the type of matching. In within-­‐group matchings agents expect principals not to control and the decision to control therefore is perceived by agents as a signal of distrust of the principal and consequently punished. On the other hand, keeping agent’s beliefs constant, control devices have a detrimental effect because agents react more hostile to being controlled by a principal belonging to a different group. Author(s): Paolo Masella; Stephan Meier;Philipp Zahn Identity, Homophily and In-­‐Group Bias Friederike Mengel University of Nottingham ABSTRACT: Many Social Interactions display either or both of the following phenomena. People tend to interact with similar others homophily. And they tend to treat others more favorably if they are perceived to share the same identity in-­‐group bias. While both phenomena involve some degree of discrimination towards others, a systematic study of their relations and interplay is yet missing. In this paper we report the findings of an experiment designed to address this issue. Participants are exogenously and randomly assigned to one of two groups. Subsequently they play a sequence of eight games with either an in-­‐group or an out-­‐group member. We find strong evidence of in-­‐group bias when agents are matched exogenously. When agents can affect who they are matched with, we find strong evidence of homophily. However, in-­‐group biases either decrease or disappear altogether under endogenous matching. This is due both to the self selection of homophilous agents into in-­‐group matches and to a change in behavior. We also show that homophily is strongly correlated with risk aversion, and we use this fact to provide an explanation for both the existence of homophily and the disappearance of in-­‐group biases under endogenous matching. Author(s): Sergio Currarini; Friederike Mengel Why Blame? Joshua Miller Bocconi University ABSTRACT: We provide experimental evidence that subjects blame others based on events they cannot be considered responsible for. An agent chooses between a lottery and a safe asset on behalf of a principal who then decides how much of a fixed amount to allocate between the agent and a third party. We observe widespread blame: regardless of what agents choose they are blamed by principals for the outcome of the lottery, an event they are not responsible for. We provide and test an explanation of this apparently irrational behavior with a delegated-­‐expertise principal-­‐agent model, the subjects’ salient perturbation of the environment. Author(s): Joshua Miller; Mehmet Y. Gurdal; Aldo Rustichini Estimation Of The House Money Effect Using Hurdle Models Peter Moffatt University of East Anglia ABSTRACT: Evidence from an experiment (Clark, 2002) investigating the ‘house money effect’ in the context of a public goods game is reconsidered. Analysis is performed within the framework of the panel hurdle model, in which subjects are assumed to be one of two types: free-­‐riders, and potential contributors. The model is extended in a number of ways: upper censoring, a standard feature of contribution data in interactive experiments, is incorporated; a tremble parameter is introduced, which allows for a free-­‐rider type to contribute occasionally for reasons of concentrational lapse; finally, the error terms in the two hurdles are assumed to have a non-­‐
zero correlation, this parameter being identifiable by virtue of the panel structure of the data. The effect of house money is seen to be significant in the first hurdle: specifically, house money makes a subject more likely to be a potential contributor. Hence we find that the effect of house money is more than just an effect on behaviour; it has the effect of transforming a subject from one type to another. This result is potentially important in the external validity debate. Author(s): Christoph ENGEL; Peter G. MOFFATT The Effects of Social Ties in a Coordination Game Frederic Moisan University of Toulouse ABSTRACT: We experimentally study the influence of positive social ties on coordination behavior. Ties are defined based on two different dimensions of social proximity, namely similarity and importance of social values (e.g. political orientation or religion), and experience-­‐based proximity (i.e. past interactions between individuals). Participants were members of a sport club involved in a team sport (volleyball). They interacted in a two-­‐player asymmetric coordination game, with and without some outside option. During this game, they faced the dilemma to choose between maximizing their own earnings and maximizing others’ earnings. In an in-­‐subject design we compare play with a teammate, another club member (but from another team) and a stranger. Every interaction was anonymous and not repeated. Our results demonstrate a significant impact of social ties on behavior and emphasize the importance of experience-­‐based proximity. Moreover, we observe that the subjects generally act according to some internalized norm, which discards all relevant theories of social preferences requiring any strategic reasoning (i.e. inequity aversion or fairness). However, we show that the observed behavior clearly supports the model of homo moralis. As an alternative approach explaining our experimental findings, we also consider Bacharach’s theory of team reasoning, which relies on group identification, thereby interpreting well the subjects’ behavior in the game. Author(s): Giuseppe Attanasi; Astrid Hopfensitz; Emiliano Lorini; Frederic Moisan Does Reciprocity Foster Mediocrity? Natalia Montinari Max Planck Institute of Economics ABSTRACT: We report evidence from an experiment where a principal has to select one between two agents to perform a task for a fixed compensation. Agents’ productivity and the principal’s payoff depend on an exogenous ex-­‐ante component which is common information and a non-­‐
contractible effort to be exerted after the selection. We run three treatments which differ in how the selection is conducted. We find that a significant share of principals select the agent who has the lowest ex-­‐ante productivity. When the principal can send a message, selected agents with the lowest ex-­‐ante productivity exert more effort than agents with the highest ex-­‐
ante productivity and principals who selected agents with the lowest ex-­‐ante productivity have on average a larger payoff than principals who selected agents with the highest ex-­‐ante productivity. This difference in effort over compensates the difference in ex-­‐ante productivity. Therefore we provide a rationale for the selection of mediocre (ex-­‐ante less productive) workers in organizations: mediocre agents reciprocate more than agents who are ex-­‐ante more entitled, when the principal is able to make them feeling indebted. Our evidence is relevant for organizations, like civil service, where i) reciprocity cannot be induced by means of higher wages and ii) non-­‐contractible effort is a relevant component of the employer’s production function. Author(s): Natalia Montinari; Antonio Nicolo; Regine Oexl Ethical versus Selfish Motivations and Turnout in Small and Large Electorates Rebecca Morton NYU ABSTRACT: We evaluate voter motivations using both small and large electorates with a diverse subject pool via two virtual laboratory experiments. We find little support for selfish instrumental turnout; that is, abstention does not increase with electorate size and voters who are in the minority abstain more than those in the majority. Expressive motivations, related to pro-­‐social behavior, appear to explain the majority of voter turnout. Most of these voters make selfish voting choices. A smaller significant number of voters make ethical choices, which are best explained by instrumental ethical motivations or large expressive ethical utility. However, we also find a significant minority of voters appears to engage in bandwagon voting, which onsets the possible moral bias in electoral outcomes from ethical voting. Moreover, the percentage of ethical voting is unrelated to electorate size, so although we find a slight moral bias in electoral outcomes, it does not increase in electorate size. Author(s): Rebecca Morton; Jean-­‐Robert Tyran Training and Planning in Saving Decisions Meryl Motika UC, Irvine ABSTRACT: I introduce a new experimental design for studying saving decisions in the laboratory and investigate the effects of information or training and planning on short-­‐term saving behavior. Subjects choose when to use tokens to watch video clips. Boredom encourages subjects to use tokens immediately, while increased video length in later rounds creates an incentive to use them later. Preliminary results show that when subjects are given intensive training in their options, they choose to either maximize the total amount of video watched or smooth video-­‐
watching over the session. Present bias and self control do not affect subjects’ choices in this treatment. Inducement to plan ahead does not change behavior, and subjects tend to either follow their plans or save more than they expected. Future sessions implement a no-­‐training treatment in which subjects are given complete but minimal instructions, without intensive training. Author(s): Meryl Motika Status, Risk, and Catching Up Pallab Mozumder Florida International University ABSTRACT: Other-­‐regarding behavior in decision-­‐making under risk remains relatively unexplored in experimental economics. To extend research in this particular direction, we design an experiment to test whether individual’s risk preferences changes when their status (relative wealth distribution) varies. In a simple experimental design we study subject’s risk preferences where we vary subject’s status while keeping their absolute wealth unchanged. Results from the experiment provide evidence that people behave differently in making decisions under risk when they are standing at different points in the relative distribution of wealth. The experimental evidence supports a behavioral argument by Robson (1992) that status affects willingness to bear risk. Author(s): Pallab Mozumder; Michael McKee; Robert Berrens Identity, Bounded Rationality, and Coordination in the Inter-­‐Group Prisoner's Dilemma Vai-­‐Lam Mui Monash University ABSTRACT: Nonstandard preferences can transform a game form from its original interpretation based on monetary payoffs. For example, social preferences can transform a Prisoner’s Dilemma in monetary payoffs into a Stag Hunt game in utilities. Social interactions, by affecting group identity and strengthening social preferences, can therefore affect cooperation in an Inter-­‐group Prisoner’s Dilemma. This paper reports an experimental study of the effects of different social interactions among members of two groups prior to their play of a one-­‐shot Intergroup Prisoner’s Dilemma. We show theoretically how inter-­‐group social interactions can affect the incidence of cooperation in this game with multiple equilibria in the presence of social preferences, strategic interactions of in-­‐group and out-­‐group members, and bounded rationality. The experimental results show that in the Baseline treatment without inter-­‐group social interaction, only 8.3% of teams cooperate in the Intergroup Prisoner’s Dilemma. Inter-­‐
group play of a coordination game, however, increases the cooperation rate by 400 percent, and adding pre-­‐play communication across groups increases cooperation by 700 percent. Author(s): Timothy N. Cason; Sau-­‐Him Paul Lau; Vai-­‐Lam Mui What can the Big Five Personality Factors Contribute to Explain Small-­‐Scale Economic Behavior? Julia Muller Erasmus University Rotterdam & Tinbergen Institute ABSTRACT: Growing interest in using personality variables in economic research leads to the question whether personality as measured by psychology is useful to predict economic behavior. Is it reasonable to expect values on personality scales to be predictive of behavior in economic games? It is undoubted that personality can influence large-­‐scale economic outcomes. Whether personality variables can also be used to understand micro-­‐behavior in economic games is however less clear. We discuss reasons in favor and against this assumption and test in our own experiment, whether and which personality factors are useful in predicting behavior in the trust or investment game. We can also use the trust game to understand how personality measures fare relatively in predicting behavior when situational constraints vary in strength. This approach can help economists to better understand what to expect from the inclusion of personality variables in their models and experiments, and where further research might be useful and needed. Author(s): Julia Muller; Christiane Schwieren Buridan's Ass in the Lab? A cautionary note on endowment effect experiments. Sean Muller University of Cape Town ABSTRACT: Many experimental studies have found evidence that supports the presence of an endowment effect, or status quo bias, in individual choice behaviour. While there has been some recent debate regarding the way in which the design of such experiments may influence findings, the basic conceptual rationale behind the design is widely accepted. In this paper we revisit the revealed preference foundations of that rationale, focusing on exchange-­‐based (as opposed to valuation-­‐based) endowment effect experiments. In particular, we show that results appearing to support the endowment effect hypothesis only do so under very particular, contestable assumptions regarding choice under indifference. We link this issue to the so-­‐called Paradox of Buridan's Ass' in philosophy, and argue that, for empirical analysis, resolutions proposed to the indifference problem in the early theoretical literature are inadequate. In addition, we show that some recent work on the theoretical problem of revealed indifference is unhelpful in separating-­‐out the competing explanations for the results of endowment effect experiments offered by prospect theory and indifference. These points are demonstrated by examining results from previous, published experiments. Author(s): Sean M. Muller Competition and Public Goods of the Ultimatum Game Noah Myung Naval Postgraduate School ABSTRACT: We create a competitive ultimatum game, which is auction like, and a public goods version of the ultimatum game and compare with the standard ultimatum game. Competitive UG lets the receiver choose which of the two sender’s offer to accept or reject while the unselected sender receives nothing. Public goods version pools the two sender’s offers and the receiver decides to accept or reject the offer as a whole. We find that 1. Sender’s send 41.8% more in competitive UG, 2. Sender’s send half as much in the PG version, which as a sum is as much as the standard UG. Receiver generally selects the higher sender in the competitive UG. However, receivers are less likely to accept in both the competitive and the PG version for the same dollar amount to the standard game. In the PG version, conditioning on the same total offer, the receivers are more likely to accept an egalitarian contribution compared to the inegalitarian contribution. Author(s): Noah Myung The ESA Executive Committee Composition Guessing Game and Proclamation of the Winners Rosemarie Nagel UPF, ICREA ABSTRACT: In January 2012 we started the survey ESA Executive Committee Composition Guessing Game in order to raise awareness about the gender composition in the ESA Executive Committee (EC from now on). In that survey we also asked for suggestions on how to improve the gender composition in EC. One of our goals is to get women more involved in ESA. In some dimensions this has already happened thanks to the actions of some members of the EC (some actions were prior to our survey). Our motto is Women need ESA and ESA needs Women. Women being elected into the EC can serve as role models for young women, however, only if they are aware that there are women in the EC. We present the results of the survey and proclaim the winners of the guessing game. We will also present the voting results of the ESA committee for 2012, if available. Keywords: Field experiment, Policy advise, Gender, Awareness, Guessing game Author(s): Humberto Llavador; Mechthild Nagel; Rosemarie Nagel; Alvaro Perdomo Strauch Nudging with Information: a Randomized Field Experiment on Reminders and Feedback Mattia Nardotto University of Bologna ABSTRACT: Can people be helped to stick to their plans with a little help from information? We provide a theoretical and empirical analysis of the effects of reminders and feedback on investment activities involving up-­‐front costs and delayed benefits, such as education and healthy behavior. By means of a randomized field experiment, we show that simple weekly reminders induce users of a gym to substantially increase their levels of physical exercise. We show that limited attention helps explain our results, and we find evidence of mental accounting in users’ response to the stimulus of reminders. These results show that virtuous behavior, such as following a healthy life style, can be promoted without the need for monetary incentives: providing incentives through information is both effective and cheap. Author(s): Giacomo Calzolari; Mattia Nardotto Alpha-­‐Final Offer Arbitration and Risk Preferences Daniel Nedelescu Purdue University ABSTRACT: Conventional Arbitration (CA) and Final-­‐Offer Arbitration (FOA) are the only two methods of arbitration used for solving disputes. Theoretical predictions show that for these approaches final offers may not converge to allow an agreement. Other theoretical arbitration mechanisms have been developed that do lead to agreement, but they do not outperform CA and FOA in laboratory experiments. The most promising results are from Amended Final Offer Arbitration (AFOA) which outperforms FOA and weakly outperforms CA. This paper presents an experiment to evaluate a more general case of AFOA, ?-­‐Final Offer Arbitration (?-­‐FOA). This mechanism is similar to a ‘second-­‐price auction,’ which punishes the loser with a value proportional (?) to the difference between her final offer and the arbitrator’s fair settlement. The experiment controls for risk preferences due to the fact that the contract zone depends on the relative risk preferences of the subjects involved in negotiation. The results show that ?-­‐FOA has on average higher agreement rates than CA and FOA, but differences in mean agreement rates are not statistically significant. The agreement rate after the subjects submit their final offers for ?-­‐FOA is significantly different from the rate for FOA and weakly statistically different from the rate for CA. No difference exists between agreement rates of more and less risk averse groups of subjects for any arbitration mechanism, contrary to equilibrium predictions. Author(s): Daniel Nedelescu Forecasting of security prices by application of valuation models Tibor Neugebauer University of Luxembourg ABSTRACT: The forecasting errors of discounted cash flow valuation and relative valuation to pricing of securities of independent companies are studied in a controlled laboratory environment under almost perfect capital market conditions. The paper presents an original experimental asset-­‐
market design where production decisions by human managers are responsible for the cash-­‐
flows to the firm. We find that security price-­‐forecasts of the relative valuation approach are generally better than those of the discounted cash flow approach. Author(s): Reinhard Selten; Tibor Neugebauer Bargaining power does not matter when sharing losses Experimental evidence of fairness in the Nash bargaining game Thomas Neumann Otto-­‐von-­‐Guericke-­‐University Magdeburg ABSTRACT: Negotiations in cases of litigation, loss given default, and debt relief concern the distribution of losses among the involved parties. These negotiations are economically relevant and seem more often than not to receive a wide range of media coverage. We analyze this type of situation and implement a Nash bargaining game in an unstructured bargaining process in an experimental study. Instead of monetary payoffs we use waiting time, where we assume that less waiting time is preferred over more waiting time. Therefore, a negotiation about waiting time can be considered as bargaining over losses. Our experiment consists of four versions of the Nash bargaining game, which vary in a way that allows a comparison of classical concepts on negotiations (Nash, Equal Loss, Equal Gain, and Kalai Smorodinski), and fair solutions driven by inequality aversion. We find an equal split of waiting time for all parameter variations. Therefore, our experimental evidence shows that the fairness concept provides a better prediction than classical concepts for the outcome of an unstructured Nash bargaining game when sharing losses. Furthermore, we can show that the inequality minimizing distribution of the loss is realized at the cost of overall welfare. Author(s): Eike B. Kroll; Ralf Morgenstern; Thomas Neumann; Stephan Schosser; Bodo Vogt Risk Taking In Groups: Experimental Evidence On The Role Of Communication And Consultation Jeroen Nieboer University of Nottingham ABSTRACT: Experimental evidence from the Myopic Loss Aversion investment task (Gneezy & Potters, 1997) shows that groups are willing to take more risk than individuals, and so achieve higher expected earnings (Sutter, 2007, 2009). We replicate this difference between groups and individuals in our experiments. The reported evidence also indicates that group members explicitly use the efficiency argument to justify higher risk taking. We investigate the power of this argument by letting groups of three individuals consult during the task, a setting with all the characteristics of group decision making minus payoff commonality. We find that consultation has an effect on individual choices. First, within consultative groups variability in choices is lower than the variability in choices across isolated individuals. Second, in later periods individuals who consult invest more than isolated individuals, and invest similar amounts to groups. However, across all periods the average amount invested by individuals who consult is not significantly different from the average amount invested by isolated individuals, and is significantly lower than the average amount invested by groups. Our results underline the importance of payoff commonality for bringing about higher risk taking in groups. Author(s): Jeroen Nieboer; Martin Sefton; Spiros Bougheas Coordination in Games with Third-­‐Party Externalities Nikos Nikiforakis The University of Melbourne ABSTRACT: Many economic activities that require coordination impose externalities on third parties. These externalities may affect coordination if decision makers care about the externality they impose on third parties or believe that other decision makers may care about them. We use a laboratory experiment to study behavior in a coordination game in which player’s actions impose an externality on an inactive third party. Subjects play a number of one-­‐shot coordination games varying the sign of the externality (i.e., positive or negative) and its size. The decision-­‐makers face a conflict between maximizing their joint earnings at one of the two pure-­‐
strategy equilibria, and achieving a more equal allocation of earnings for all group members at the other. The results show that negative externalities affect coordination, but not positive externalities. Players’ actions are largely consistent with their preferences, but other-­‐regarding individuals are more likely to attempt to coordinate on the outcome preferred by selfish players. We show that this behavior is rational given the low number of individuals with other-­‐regarding preferences in our sample. Finally, we find that although negative externalities encourage pro-­‐
social behavior, efficiency is harmed because coordination failure increases. Author(s): James Bland; Nikos Nikiforakis 'I do it by myself as I knew it all along.' On hindsight bias and delegation. David Nils Danz Social Science Research Center Berlin (WZB) ABSTRACT: We investigate whether the hindsight bias can affect decisions to delegate. To test this relationship, we ran an online experiment during the Soccer World Cup 2010. Participants were asked to predict a number of outcomes of the ongoing World Cup and later had to recall their assessments after the outcomes had been realized. This allowed us to assess which participants were hindsight biased. The participants also had to make choices in a delegation game between a principal and an agent. With the help of a simple model, we show that for certain signal realizations, the hindsight bias leads to inefficient delegation decisions. Our data confirms that hindsight-­‐biased subjects more frequently fail to delegate optimally than subjects whom we have not classified as hindsight biased. Author(s): David Danz; Frank Huber; Dorothea Kubler; Lydia Mechtenberg; Julia Schmid The Restart Effect as a Focal Point Doug Norton Florida State University ABSTRACT: The restart effect in public goods games occurs when there is an increase in contributions to the group account in the period following a stoppage in play. This upward pulse in contributions behavior is often reported but has been largely un-­‐researched. I present a 2x2 experimental design to isolate potential causes of the restart effect. The treatments include the number of periods elapsed prior to stoppage in play and language in instructions. Language in the instructions about the exact pattern of the stoppages has a strong and independent effect. These findings suggest that in order for the restart effect to exist there must be something focal about the stoppage in play something that sparks a “coordination moment”. Author(s): Douglas A. Norton The Use and Effectiveness of Discretionary Sanctions and Rewards in the Inspection Game Daniele Nosenzo University of Nottingham ABSTRACT: We investigate the effectiveness of discretionary sanctions and rewards in a repeated ‘inspection game’. In this game an employee can either work or shirk and an employer can inspect or not inspect. We examine the effects of allowing the employer to sanction or reward the employee after an inspection relative to the case where neither instrument is available. Theoretically, the set of outcomes that can be sustained in equilibrium depends on which instrument is available. Only when discretionary sanctions are available can employees be induced to work without the use of costly inspections in equilibrium, attaining the employer’s preferred (and the joint-­‐payoff maximizing) outcome. In an experiment we find that rewards or sanctions both discourage shirking and have similar effects on joint earnings. In games allowing sanctions this is accomplished with a lower inspection rate and the efficiency gains accrue to employers. In games allowing rewards employers actively reward employees for working and the efficiency gains mainly accrue to employees. A treatment where employers can combine sanctions and rewards does not lead to higher efficiencies than the single-­‐instrument treatments. Author(s): Daniele Nosenzo; Theo Offerman; Martin Sefton; Ailko van der Veen Cash balances and dividend processes in markets with constant fundamental values Charles Noussair Tilburg University ABSTRACT: We consider markets for long-­‐lived assets, with a constant fundamental value, and populated with inexperienced traders. We find that prices track fundamental values when traders have a modest amount of cash, but exhibit price bubbles when cash endowments are large. Furthermore, when fundamental values are constant for part of the life of an asset and later exhibit a trend, mispricing arises at the onset of the trend. We evaluate whether there are asymmetries in behavior between markets with increasing and decreasing trends in fundamentals. The results indicate that when markets of the type we study have appeared to converge to fundamentals, this convergence is to a particular price rather than to the fundamental value trajectory. Author(s): Charles Noussair Separately Identifying the Screening, Signaling, and Sunk-­‐Cost Effects of Price Scott Ogawa Northwestern University ABSTRACT: Is there a causal effect of price? Specifically, if a person pays more for a product, will he or she use it with higher intensity? Previous research has had difficulties disentangling three distinct effects of price on usage intensity: as a screen on population valuations; as a signal of quality; and as something that induces sunk-­‐cost bias. I present a novel experimental design that separately and simultaneously identifies each of these effects. To quantify and compare these effects, I introduce a simple structural model and show that the parameters are separately identified by the experimental data. I implement the proposed experiment by selling two different goods. In a small study, I sell vouchers for coffee and then track redemption rates. I find evidence that both price signaling and sunk-­‐cost bias each increase usage separately. In a larger study, I sell internet-­‐based exam preparation tutorials. I monitor tutorial usage intensity as well as subsequent performance on the exam. Data collection for the second study will be completed in June 2012. Author(s): Scott Ogawa Power and the Privilege of Clarity: An Analysis of Bargaining Power and Information Transmission Sander Onderstal Amsterdam School of Economics ABSTRACT: We provide a game theoretic analysis of how power shapes clarity. In economics, the relation between power and communication is a relatively untouched research area. Here, we analyze information transmission in a cheap talk bargaining game between an informed Sender and an uninformed Receiver. Our main result is that the maximum amount of information transmission is increasing in the relative power of the Sender: power reduces the extent to which the Sender can be exploited with the information she provides. This prediction is corroborated by a laboratory experiment. Author(s): Adrian de Groot Ruiz; Theo Offerman; Sander Onderstal Tiger Women: An All-­‐Pay Auction Experiment on Gender Signaling of Desire to Win David Ong Peking University HSBC School of Business ABSTRACT: Women’s lower wages and underrepresentation in the most competitive professions have been well documented. Numerous experiments suggest that women are less competitive. However, currently competitiveness is measured indirectly by using the residual of the choice of competitive payment after controlling for ability, confidence, and risk attitude econometrically. More recent results with different tasks, children, single sex schools, and Muslim countries have shown no gender differences, indicating the possible influence of cultural stereotypes in prior results. Furthermore, even in the west, women’s superior academic achievement despite lower standardized test scores is not consistent with lesser competitiveness. We contribute to this literature by testing for gender differences in ‘desire to win’ in a sealed bid all-­‐pay auction at mid-­‐tier and top-­‐tier universities. Our treatments consisted of informing paired bidders of the gender and school rank of their opponent. We derived the implied valuations (desire to win) and risk attitudes using an extension of standard auction theory. In principle our method should be a direct measure of competitive attitude which eliminates ability, confidence and stereotype confounds. Contrary to the literature and our expectations, we found that women had higher desire to win, most significantly at the top tier school, where they were even less risk averse. Furthermore, this pattern of results was anticipated in across school bids, which is consistent with our signaling hypothesis. If willingness to pay to win is related to willingness to prepare, our results support other evidence showing that women’s superior academic achievement is due to greater self-­‐discipline. Our result could also help explain overbidding in apparently common value auctions. Author(s): David Ong; Zhuoqiong (Charlie) Chen Reference Dependence and the Momentum Effect: Theory and Experiment Ryan Oprea University of California, Santa Cruz ABSTRACT: A great deal of previous experimental research shows that reference dependence can seriously inhibit trade in static, non-­‐continuous settings due to the Endowment Effect. We theoretically and experimentally study how reference dependence operates in more realistic dynamic, continuous environments. We show theoretically that reference dependence can generate a dynamic Momentum Effect capable of eliminating the impact of the Endowment Effect on trade. We then design a novel experiment testing for the Endowment Effect and countervailing Momentum Effect and find that both are quite common in the subject population. Counterfactual analysis using elicited preferences suggest that on net a majority of subject pairs in our experiment would trade more in a dynamic market than in a static one. Author(s): Sean Crockett; Ryan Oprea Multidimensional Ellsberg Pietro Ortoleva California Institute of Technology ABSTRACT: Ellsberg's classical experiment involved a gamble with no ambiguity and gambles where the prize that could be won is objectively known, but the winning probability depends on the (ambiguous) urn's composition. We extend this by including gambles where the prize won depends on the urn's composition; gambles in which the date of payment depends on the urn's composition; and gambles in which combinations of the three dimensions -­‐-­‐ prize, probability and date -­‐-­‐ depend on the urn's composition. All these ambiguous dimensions depend on a unique state of the world, the composition of an ambiguous urn. We have three main findings. First, attitude towards ambiguities vary sharply across dimensions: a vast majority of subjects dislikes ambiguous prizes, a smaller majority dislikes ambiguous probabilities (Ellsberg), while only half of the subjects dislike ambiguous payments dates. Second, when one dimension is fixed and ambiguous and participants can choose whether or not to have ambiguity in another dimension, the majority of participants prefer ambiguity in both dimensions, and in particular prefer the gambles in which the two forms of ambiguity are perfectly correlated (more exposure). Third, when comparing options with ambiguity in multiple dimensions against options with no ambiguity, the majority of participants prefers the option with no ambiguity, but the proportion of participants who prefer no ambiguity to ambiguity in multiple dimensions is often significantly smaller than the proportion preferring no ambiguity to ambiguity in each single dimension. Author(s): Pietro Ortoleva Facial cues and trust-­‐related behavior Bastiaan Oud University of Zurich ABSTRACT: What influence do facial cues have on trust-­‐related behavior? We extend the literature on this by (1) assessing also the behavior of trustees (not just trustors) in trust games, (2) analyzing the role of beliefs in the translation of facial information into behavior and (3) assessing the accuracy of inferences made from the photos. In a first experiment, we exposed both trustors and trustees in a trust game to pictures of their (actual) counterpart’s neutral-­‐expression faces, very subtly morphed in shape towards facial prototypes identified as ‘trustworthy’ or ‘untrustworthy’ by Oosterhof and Todorov (2008). This affected the amount trustors sent in the expected direction. Interestingly, there was a similar effect on trustee’s behavior, even though the trustworthiness of their counterparts was strategically irrelevant. We conjecture that this may be an instance of type-­‐based reciprocity, where participants respond to what type they think a counterpart generally has, as opposed to their behavior in a particular instance. We also assessed how accurate beliefs based on facial pictures were, by presenting the unmorphed photos to a separate group of participants and incentivizing them to accurately guess the behavior of the people in the photos. Not only were these guesses uncorrelated with the actual behavior of the persons in the photos, but also, guessing certainty was uncorrelated with guessing accuracy, suggesting overconfidence. Author(s): Bastiaan Oud; Tony Williams; Jan Engelmann; Eva Krumhuber; Ernst Fehr Dynamic Effects of Redistribution: A Real-­‐effort Experiment Fabian Paetzel University of Bremen ABSTRACT: We study how heterogeneity of initial endowments affects voting behavior on redistribution in a repeated real effort task. In particular, we investigate dynamic effects of the voting outcome on individual effort in following periods. In each period, subjects take part in a real-­‐effort task (from Gill and Prowse, 2012). The earnings distributions are generated endogenously and thus vary in mean, variance and skewness. Subjects then have to decide how much they want to redistribute within the group. Redistribution is designed as a fully redistributive tax (Meltzer and Richard, 1981). Free communication is allowed. Treatments differ in the voting rule (unanimity and majority) and the matching protocol (stranger and partner). In all treatments, we find a positive relationship between the skewness of endowment distribution and the level of the tax rate (in line with Nash-­‐prediction). We found that with increasing skewness the likelihood of forming a coalition of lower endowed subjects is increasing only if the best-­‐endowed are not protected by the unanimity rule. When lower endowed subjects form a coalition, they enforce significantly higher taxes (median=100%). The effect of the tax level in previous rounds on effort in following rounds does not depend on the treatment (voting rule, matching protocol). We found that lower endowed subjects who profit from redistribution, increase effort in the next round (indirect reciprocity). Moreover higher endowed subjects who lose from redistribution do not decrease effort in following periods. Author(s): Bernhard Kittel; Fabian Paetzel; Stefan Traub Endogenous Evolution of Society: A Public Goods Experiment Stefan Palan Karl-­‐Franzens-­‐University Graz ABSTRACT: We experimentally investigate the question whether human subjects are willing to give up individual freedom in return for the benefits of societal organization. This question is motivated by the sociological literature, which posits that tribal organization in human societal evolution developed because the benefits from cooperation increased the fitness and procreational prospects of individuals who banded together in tribal groups. In our experiment, we conduct a public goods game where subjects can self-­‐select into one of two groups. In the first, each subject decides on her contribution to the public good individually. In the second, one member of the group is randomly chosen to act as a coordinator (a tribal chief), who chooses a homogeneous contribution level for all group members. We conjecture that the structure of the second group, which resembles a tribal organization, helps overcome the coordination problem inherent in the provision of a public good. In particular, we expect coordinators to choose a high contribution level to maximize individual and group payoff. Furthermore, over time we expect subjects to self-­‐select into the coordinator group, thus giving up individual freedom in return for the benefits of belonging to a group wherein one individual coordinates all group member’s actions. Author(s): Fleiss; Jurgen; Palan; Stefan Promoting Trust with Outgroups Xiaofei (Sohpia) Pan George Mason University ABSTRACT: We investigated experimentally whether group identity formed under a more cooperative process trust more in general. In particular, whether they trust a more cooperative group formation process promotes trust towards outgroups. We randomize participants into two production tasks whereas one requires more cooperation than the other to finish. Participants then play the trust game where they indicate their willingness to trust if matched with an ingroup trustee or an outgroup trustee. We find that the more cooperative group formation task involves more sharing and exchanges behaviors. In the trust game, we find that participants who have experienced a more cooperative group formation process present significant less ingroup favoritism, and this is realized through a significantly enhanced trust towards outgroup. This enhanced trust towards outgroup trustee is persistent even after they experienced betrayal's. From a belief elicitation, we conjecture that the proximate mechanism can be that a cooperative group formation process promote trust by positive affecting beliefs on outgroup trustee’s willingness to reciprocate. Author(s): Xiaofei (Sophia) Pan; Daniel Houser Cognitive hierarchies in adaptive play Ronald Peeters Maastricht University ABSTRACT: Inspired by the behavior in repeated guessing game experiments, we study adaptive play by populations containing individuals that reason with different levels of cognition. Individuals play a higher order best response to samples from the empirical data on the history of play, where the order of best response is determined by their exogenously given level of cognition. Using this learning framework, we study first the evolution of behaviour in the Nash demand game; from there, we generalize intuitive properties to the more general class of bimatrix games. As in Young’s model of adaptive play, (unperturbed) play still converges to a minimal curb set. However, with the random perturbations of this (higher order) best response dynamic, the stochastically stable states obtained may now be different, but in a deterministic manner. Perhaps counter-­‐intuitively, higher cognition may actually be bad for both the individual with higher cognition and his parent population. Author(s): Abhimanyu Khan; Ronald Peeters Reciprocating Kindness: An Experimental investigation Vittorio Pelligra Dept. of Economics, University of Cagliari ABSTRACT: This paper discusses and experimentally investigates the relation between perceived kindness and reciprocal behavior in the context of atrust game. We consider the level of kindness associated with a trustful action as treatment variable, and compare, both in a within-­‐ and in a between-­‐subject design, trustees’ responses to different degrees of kindness. Contrary to theoretical predictions, our data show that the frequency of trustworthy responses do not change significantly across treatments: different degrees of kindness elicit very similar responses from the trustees. We conclude that, despite intentions may matter in motivating cooperative behavior, they trigger reciprocal responses via a mechanism that seems to be different from that described by the reciprocating kindness assumption. Author(s): Vittorio Pelligra A Model Of Boundedly Rational Team Decision Making Stefan Penczynski University of Mannheim ABSTRACT: Team decision making has received a lot of attention in experimental economics. This paper uses a theoretical framework of limited information processing in the spirit of Lipman (1999) to model communication between boundedly rational agents and to formulate a simple theory of team decision making. We describe the characteristics of persuasive arguments and relate them to experimental results on team deliberation. Furthermore, we characterize features of optimal team deliberation. Author(s): Elena Manzoni; Stefan Penczynski A New Experimental Evidence of Statistical Discrimination Emmanuel Peterle CREM -­‐ Rennes (France) ABSTRACT: This paper reports results from a real effort experiment aiming to study discrimination in hiring decisions. We design a game in which employers are asked to hire employees in order to perform a real effort task under a tournament remuneration scheme. Our experiment consists of four treatments that differ in the level of information that employers receive about each applicant. In our baseline treatment, employers are only informed about the applicant’s demographics such as gender, ethnicity and education. In the remaining treatments, employers receive additional information about each applicant’s potential performance. These treatments differ in the accuracy of the information regarding performance that is available. Our data provide unambiguous evidence for the existence of statistical discrimination in hiring decisions. We observe both gender and ethnic discrimination when no relevant information on ability is provided. The introduction of information on performance reduces significantly discrimination, suggesting that discrimination is mainly due to a lack of information rather than preferences. However the extent of the reduction of discrimination with additional information about performance strongly depends on the accuracy of this information. Author(s): Sophie Larribeau; David Masclet; Emmanuel Peterle Commission and Omission in Tournaments Jan Philipp Wilhelm RWTH Aachen ABSTRACT: Rank-­‐order tournaments are used to allocate prizes according to the relative performance of contestants. One downside of tournaments is that, apart from inducing productive effort, competitors also have incentives to improve their relative position by decreasing the output of others. Anticipating the moral and social meaning of such acts, people perform them clandestinely. Thus, studying them in real organizations is hardly possible. Apart from actively destroying the output of contestants via sabotage activities, contestants may also decide not to prevent the reduction of other’s output frequently labeled as acts of ‘omission’, e.g. by withholding information. In a laboratory experiment, we compare treatments in which participants in identical strategic settings may either exert active sabotage (commission) or make an omission decision. Participants are competing in dyadic real-­‐effort tournaments, and they may decrease the output of their contestant at no cost. By distinguishing between active sabotage and omitting information regarding a negative external shock, we may investigate determinants for both types of activities. Previous research suggests that omission should be used more frequently. In real organizations, acts of commission or omission can typically be detected and made public by colleagues or supervisors. To understand the effect of possible detection and the potential fear of being revealed as an offender, we introduce a detection probability in the lab. When an act of commission or omission is observed, a photo of the offender as well as information on the intensity of harm is shown to the contestant. Author(s): Christine Harbring; Jan Wilhelm Bonus Schemes and Trading Activity Elena Pikulina Tilburg University ABSTRACT: Although the role of compensation packages in aligning the interests of managers with those of shareholders is widely studied, there is little research available on how different bonus schemes affect trader’s propensity to trade and which bonus schemes improve trader’s performance. In an experimental setting, we study the effects of these two bonus schemes on trader’s behavior: a linear bonus scheme, which always pays a fixed share of profit, and a threshold bonus scheme, under which a share of profit paid as bonus increases in the total profit earned by a trader if specific performance targets are met. The participants to our experiment have traded shares in an experimental stock market on the basis of fundamental and technical information (evolution of the market index, past share price evolution, realized earnings, and analyst’s earnings forecasts). We find that they trade more intensely under the threshold bonus scheme than under the linear bonus one. Furthermore, trading intensity significantly decreases when the bonus scheme’s thresholds are reached under the threshold bonus scheme. While trading intensity is higher under the threshold scheme, the participant’s performance under this scheme is worse than under the linear bonus scheme as a consequence of worse investment decisions but of not transaction costs. Author(s): Elena Pikulina; Luc Renneboog; Jenke ter Horst; Philippe Tobler The risk of contagion. An experiment on conformity to unethical behavior Marco Piovesan Harvard Business School ABSTRACT: Do people cheat more when they are informed that others cheated? And, does it matter whom the others are? In our experiment students toss a coin in private and record the outcome (white or black) on a report sheet. We only reward those who report white. In our 2x2 experimental design we manipulated the report sheets to make students believe that others were honest (or dishonest) and to make them believe that the other’s are from their own class (or not). Our results show that subjects tend to conform: they cheat more if others are cheating and they cheat less if others are not. However, this conformity effect is significantly stronger in men. Men are also more influenced by the behavior of their classmate compared to woman. Finally, we control for individual characteristics and we find that intelligence increase the probability of cheating for both genders. Author(s): Toke Fosgaard; Francesca Gino; Lars Gaarn Hansen; Marco Piovesan Ambiguity and compound risk attitudes: an experiment Laetitia Placido Paris 1 & Paris School of Economics ABSTRACT: The identification of compound risk attitudes and ambiguity attitudes has recently received experimental support (Halevy, 2007) and been incorporated in decision models (Seo, 2009; Halevy and Ozdenoren, 2008; Segal, 1987). Non-­‐reduction of compound lotteries is this literature’s explanation of Ellsberg type behavior. We conduct an experiment measuring individual behavior under simple risk, under various types of compound risk and under ambiguity. We examine how each of these behaviors changes as the probability (or size) of the winning event varies. We find that attitudes towards all three types of uncertainties move from seeking to aversion as the probability level increases. Controlling for probability level, we find that the link between ambiguity and compound risk attitudes is partial and sensitive to the type of compound risk considered. We do not support the equivalence between reduction of these compound risks and ambiguity neutrality. Author(s): Mohammed Abdellaoui; Peter Klibanoff; Laetitia Placido Social Preferences, Risk Preferences and the Hexagon Condition Giovanni Ponti Universidad de Alicante ABSTRACT: This paper reports experimental evidence from a 3-­‐player Dictator Game in which Dictators decide over the distribution of probabilites of winning a fixed, indivisible, monetray prize. This evidence is compared with a standard (control) treatment in which money is perfectly divisible, and Dictators allocate shares of the prize across the group members, and also with an ‘hybrid’ protocol, by which a fraction of the prize is allocated deterministically, and the remainder through a lottery. Dictator’s decisions are framed within (a suitably modified version of) Karni and Safra'ss (2002) model of distributional justice, which allows us to identify ‘consequentialist’ vs. ‘procedural fairness’, also controlling for (own-­‐payo?) risk aversion. Our evidence shows that both views of fairness are complementary in explaining subject’s social preferences, and how they relate with individual’s attitudes to risk. KEYWORDS: Dictator Games, Social preferences, risk preferences. JEL CLASSIFICATION: D86. Author(s): Xavier Del Pozo; Arianna Galliera; Giovanni Ponti; Iryna Sikora Disclosing Advisors' Interests Helps nor Hurts Jan Potters Tilburg University ABSTRACT: An informed advisor may have incentives to send a deceiptful message to an uninformed decisionmaker. We set up an experiment to study whether disclosure can foster truthfullness and trust. A key feature of our setup is that disclosure reveals the interests of the informed advisor to the uninformed devisionmaker but not that there is a conflict of interests between the advisor and decisionmaker. We find that disclosure hurts nor helps; deceptive advice and mistrust are equally frequent with as without disclosure. Relative to the aims of disclosure policies this is ‘bad news’; relative to studies suggesting that disclosure may actually increase the rate of deception this may be regarded as ‘good news’. Author(s): Huseyn Ismayilov; Jan Potters Do smart people cheat less? An experimental study Xiangdong Qin Shanghai Jiao Tong University ABSTRACT: Cheating is sometime used to gain competitive advantage. It is conjectured that in a competitive environment, the less advantageous people would cheat more to compensate for their deficiency. We use a series of lab experiments to test the hypothesis that smart people cheat less. Subjects go through an incentive compatible IQ test to reveal their overall intelligence. Then subjects with the similar IQ scores are randomly matched into a group of four. A depository device is used to elicit subject’s cheating levels. The overall finding confirms the hypothesis, that is, compared to subjects with low IQ scores, subjects with high IQ scores cheat less in a competitive environment. We also alter the environment to eliminate competition, we find that the difference in cheating levels re-­‐emerges, but in a different direction. Furthermore, the risk neutral assumption is relaxed in this study. We use the lottery game to assess each subject’s risk preference, and then incorporate the degree of risk preference into analyzing the subject’s cheating level. Author(s): Xiangdong Qin; Shuwen Li The Effect of Expectation on the Third Party's Decision: An Experimental Study Jianying Qiu Radboud University Nijmegen ABSTRACT: Despite the absence of material interests, third parties of an interaction typically punish selfish decisions and reward kind choices, even when the punishment or reward is costly to them. We experimentally explore the effect of expectation on the third party’s decision in an investment game with three players: a truster, a trustee, and an impartial observer who receives a fix payment and decides on the amount of reward or punishment of the trustee at own cost. As a novelty, we additionally elicit the truster’s payoff expectation of the trustee’s return and convey it to the impartial observer. Our experimental result suggests that truster’s expectation changes the third party’s decision: he punishes more (or reward less) when facing high payoff expectations of the truster, and is more tolerant of the trustee when facing low payoff expectations. Author(s): Jianying Qiu; Tobias Uske; Andreas Haupt Peer Influences On Honesty: The Role Of Social Norm And Social Preference Hong Qu Penn State University ABSTRACT: In a budget reporting experiment in which lying about privately known costs increases personal monetary payoffs, we examine how an individual’s honesty is affected by information about their peers who face similar reporting dilemma. We find that subjects decrease their honesty in response to less honest peer and increase their honesty in response to more honest peer. The response to less honest peer is much stronger than the response to more honest peer. Furthermore, the reaction to more honest peer is significant predominantly for subjects who are identified as the conventional type by a psychometric test. The conventional type is sensitive to social norms according to the Kohlberg’s theory of moral development (Kohlberg 1958). All types react to less honest peer, and the magnitude of the reaction is correlated with a measure of disadvantageous payoff aversion in the post experiment questionnaire. Author(s): Steve Huddart; Hong Qu Distributed Choices in Networks: Routing Splittable Flow Amnon Rapoport University of Arizona ABSTRACT: The Braess Paradox (BP) is a counterintuitive finding that upgrading a network which is susceptible to congestion may increase the cost of travel for each user. Our major goal is to determine if, and to what extent, network users may benefit when the population consists of separate groups that operate like unitary players. A second purpose is to determine if the BP is experimentally realized in complex networks. We answer these questions by having large groups of participants (n=18) choosing routes in an iterated network game under two conditions. In one condition, decisions are fully distributed as each user only controls a single unit of flow. In the second condition, each user controls multiple units of flow which she may split over the routes in any way she wishes. In each condition, the users independently and repeatedly choose routes in two different networks. The basic network consists of three parallel routes that connect a common origin to a common destination. This network is subsequently expanded by adding two links to form an augmented network which has five routes with shared links. In equilibrium, individual cost of travel increases in both conditions as the network is expanded. We construct the equilibrium solutions for each combination of condition and network, and then use them as benchmarks to organize the route choices. Our results show that in both conditions behavior converges to equilibrium. Author(s): Amnon Rapoport & Eyran J. Gisches Mistakes, Closure and Endowment Effect in Laboratory Experiments Anmol Ratan University of Maryland ABSTRACT: In this paper, we relax the hard closure property of experiments that have been used to study endowment effect in laboratory. We study differences in benchmark environments (hard closure) and an environment that allows participants to reverse the decisions taken in the laboratory (soft closure). We find that ‘endowment effect’ is not observed in the soft closure treatment. The procedures in our experiment allow us to circumvent the critique of altered expectations. Our results call for a careful interpretation of experiments that suggest ‘endowment effect’ in laboratory conditions. Other implications pertain to external validity of experiments with hard closure. Author(s): Anmol Ratan Dishonesty, tolerance, and social information Maria Recalde University of Pittsburgh ABSTRACT: This paper investigates the effect of social information on the individual decision to engage in and tolerate dishonest behavior. It constructs a strategic setting in which subjects play an investment game repeatedly in pairs. Half of all participants have the opportunity to engage in dishonest actions and half are harmed by the dishonest actions of others. Tolerance is studied through the behavior of the latter participants who have the opportunity to punish dishonest behavior at a cost. Results show that dishonesty gives rise to inefficiencies in a repeated setting. Observing others behave honestly or be punished increases the likelihood that individuals behave honestly. This information also increases the likelihood that individuals punish dishonest behavior at a cost. Punishment is ineffective in inducing honesty among participants; the victims of dishonesty learn to tolerate dishonest behavior. Confusion cannot explain the results observed. Subjects are honest 40% of the time when dishonesty inflicts a direct negative externality on another participant, but are never honest when they interact with the computer. Although the effects of social information could be explained by learning and strategic considerations, the fact that third party behavior is more important than partner behavior suggests that this is not the case. Individuals develop a sense of morality in the laboratory when their actions inflict a negative externality on another equal party that changes with social information and the prevalence of unselfish behavior observed. Author(s): Maria P. Recalde Severity vs Leniency Errors in Performance Appraisal -­‐ A Real Effort Experiment Tommaso Reggiani U. Cologne ABSTRACT: Supervisors can fail performance appraisal mainly in two ways: with leniency errors that assign predominantly high evaluations thus rewarding even undeserving agents that have exerted low effort or with severity errors that assign predominantly low evaluations and thus neglecting rewards to deserving agents that have exerted high effort. The basic principal-­‐agent model with moral hazard predicts both errors to be equally detrimental to effort provision. We then show this prediction fails in the lab. In fact, failing to reward deserving agents is significantly more detrimental than rewarding undeserving agents. We discuss our result in the light of different economic theories of behavior. Our result may have interesting implications for strategic human resource management and personnel economics and may also contribute to the debate about incentives and organizational performance. JEL code: C91, M50, J50 Keywords: Agency theory, Organizational justice, Type-­‐I and Type-­‐II errors, Real effort, Rater error, Leniency errors, Severity errors, Performance appraisal. Author(s): Reggiani; Rizzolli Giving and Probability David Reinstein University of Essex ABSTRACT: What is the best way to ask for assistance from someone who may get an uncertain ‘bonus’ income? When will she give more to charity: if she is asked before, to conditionally commit to donate if she wins the bonus; or if she is asked just after learning that she won the bonus? If decisions follow the rules of standard expected utility theory there should be no difference. In contrast, we present interpretations of theories of reference-­‐dependent utility, adaptation, tangibility, uncertainty aversion, and self-­‐signaling that predict more giving before, as well as theories of affective mood that predict greater generosity after. Our evidence comes from a field experiment run on a Valentine’s Day E-­‐card web site accesible at three UK universities, advertised as a fundraiser for Right to Play, a popular international charity. Students and staff who completed a survey were randomly awarded restaurant vouchers worth between £12 and £30, depending on university. Eligible participants were later sent an email stating ‘the winners have now been drawn ... Please click on the following link to find out if you have won.’ They were then randomly assigned to one of two treatments. In the Before treatment, we asked ‘Before we reveal if you have won ... how much you would like to donate, if you win the prize...?’ In the After treatment, after they learned their prize, losers (winners) were asked ‘(Now that you have won) ... how much you would like to donate...?’ While 35 percent (38 of 109) of Before participants pledged, only 15 percent (7 of 45) of winners in the After treatment pledged, and none of the 47 losers in the After treatment pledged; these differences are strongly significant in standard parametric and non-­‐parametric tests. Contact restaurants on status of pledge. These results are directly relevant to fundraising and volunteer-­‐solicitation strategies, and suggest that we need to exercise caution in applying expected utility theory to models of pro-­‐social behavior and social preferences. Author(s): David Reinstein; Gerhard Riener; Michael Sanders Measuring the distribution of spitefulness J.Philipp Reiss Maastricht University, Economics Department ABSTRACT: Spiteful, antisocial behavior may undermines the moral and institutional fabric of society, producing disorder, fear and mistrust. Previous research demonstrates the willingness of individuals to harm others, but little is understood about how far people are willing to go in being spiteful (relative to how far they could have gone) or their consistency in spitefulness across repeated trials. Our experiment is the first to provide individuals with repeated opportunities to spitefully harm anonymous others when the decision entails zero cost to the spiter and cannot be observed as such by the object of spite. This method reveals that the majority of individuals exhibit consistent (non-­‐)spitefulness over time and that the distribution of spitefulness is bipolar: when choosing whether to be spiteful, most individuals either avoid spite altogether or impose the maximum possible harm on their unwitting victims. Author(s): Erik O. Kimbrough; J. Philipp Reiss What Drives Over-­‐Expenditure of Effort in Contests? An Experimental Investigation Lucas Rentschler Centro Vernon Smith de Economía Experimental ABSTRACT: We experimentally analyze behavior in perfectly discriminating contests with incomplete information under three valuation structures: (1) common value, (2) private and common value, and (3) independent private value. We find that contestants expend more effort than predicted by Nash, and that contestantsoften choose effort levels that guarantee themselves negative profits in equilibrium. We explore several explanations for this behavior. Author(s): Diego Aycinena, Rimvydas Baltaduonis, Lucas Rentschler On Rational Cooperation Ernesto Reuben Columbia University ABSTRACT: From the perspective of non-­‐cooperative game theory based on ‘selfish’ preferences and perfect information, individuals are solely interested in maximizing their own material payoff and observe the (selfish) preferences of all others individuals should not cooperate, trust or reciprocate in repeated interactions with a known end. Clearly, these are not the facts. Individuals do cooperate, trust and reciprocate even if they know exactly when the interaction will end. Non-­‐cooperative game theory can be adjusted to these facts by assuming people are different some have ‘selfish’ and some have other-­‐regarding preferences (for example, they value reciprocity per se). Additionally, if people are uncertain about each other's preferences, a small degree of reciprocity can be sufficient to induce the selfish to act as if they are reciprocal as well. We investigate in laboratory experiments the importance of such uncertainty in explaining trust and reciprocity in repeated interactions. Author(s): Ernesto Reuben; Sigrid Suetens On the relative performance of contingent and non-­‐contingent incentives Pedro Rey-­‐Biel Universitat Autonoma Barcelona ABSTRACT: In contrast to the simplifying assumption of selfishness, social incentives have been shown to play a role in economic interactions. Before incorporating social incentives into models and policies, however, one needs to know their efficiency relative to standard pay-­‐for-­‐performance incentives. We report evidence from a large field experiment comparing the effectiveness of contingent and non-­‐contingent (‘social’) incentives in eliciting costly effort. The company with which we worked sent 7,250 letters asking customers to complete a survey. Some letters contained cash amounts ranging from $1 to $30, whereas others promised to pay upon compliance. We compare the response rates and cost effectiveness of these contingent and social incentives with each other and with a no-­‐incentives control. In line with previous findings, we find that social incentives generated some effort: small amounts increased the response rate with respect to the control, but the size of the reward had a relatively minor effect. In contrast, the response rate for contingent incentives was low for small amounts but increased rapidly as incentives increased. Importantly, for (almost) any given response rate social incentives were more costly than contingent incentives. Author(s): Uri Gneezy; Pedro Rey-­‐Biel Social Reference Point Shifts and Reciprocity: Behavior and Brain Evidence Arno Riedl Department of Economics 1, Maastricht University ABSTRACT: Positive and negative reciprocity is a widespread behavioral regularity in humans that is thought to be essential in enforcing norms and social order among genetically unrelated individuals. Positive reciprocity refers to kind behavioral responses when treated kindly and negative reciprocity to unkind responses when treated unkindly. Clearly, the evaluation of being treated kind or unkind needs a reference point. Often the status quo is assumed to be the natural exogenous reference point. We argue that in social interaction the reference point evaluating kind and unkind behavior is not necessarily the status quo but endogenously constructed as ones expectations regarding others behavior. For instance, an unexpected kind and unkind act may be perceived as positive and negative surprise, respectively, while an expected kind or unkind act may be taken for granted. To test directly the importance of expectations for the reference point in social interactions in which people interacted in an economic exchange experiment while being scanned (fMRI), we used two conditions: the social condition by using positive and negative money transfers by humans and the non-­‐social condition by using computer generated transfers. Subjects as responders in the scanner were asked to indicate the transfers they expected to receive from other players in the beginning of each round. Thereafter, they were informed about the actual transfers. Then they had to decide how to reciprocate by sending money back or taking money away. Both reciprocal responses were equally costly for the responder. In the social condition, we observe a clear shift in subjects’ reciprocal responses when evaluating it against exogenous and endogenous reference points. No such shift is observed in the non-­‐social condition, suggesting that subject’s expectations indeed influenced their reference point when interacting with humans. Neuroimaging results support this finding. The social condition produced stronger and wider neural responses than non-­‐social condition, implying that interacting with humans require more elaborate inferences. We also find enhanced activity in the anterior rostal medio-­‐prefrontal cortex, which is known to be a social mentalizing area, when evaluating the received transfers with respect to endogenous than to exogenous reference points. Moreover, activation of the ventral striatum (VS) and dorsal anterior cingulate cortex (dACC), areas that have been shown to be involved in the valuation of monetary rewards and a coping function for social stress respectively, correlate with subject’s positive and negative reciprocal responses in the future. The VS activation positively correlates with reward and punish responses in the social condition, suggesting a hedonic value of altruistic reward and punishment. The dACC correlation suggests its function as a neural alarm system which mediates the social reference point shifts. Our findings show that in social interactions reference points are indeed constructed endogenously by our expectation, and also suggest that the brain reward circuitry is functioning as a predictor of human positive and negative reciprocity. Author(s): Arno Riedl; Sanae Okamoto-­‐Barth; Martin Strobel; Hester Breman; Armin Heineke; Rainer Goebel The Rating Effect: Stars and Asset Allocation Thomas Rietz University of Iowa ABSTRACT: Ratings are common in financial markets and have the potential to be either helpful or harmful. We experimentally assess the effect of ratings in a setting where subjects make investment decisions and have full information about the characteristics of the investments. Ratings supply no additional information. Depending on the treatment, subjects make decisions in settings in which assets are unrated, rated as a single group, categorized and rated within categories, in which the ratings are explained, and in which subjects are told the rating rule and calculate their own rating. We find that ratings affect investment decisions, but that the effect largely vanishes when ratings no longer appear. Subjects scoring higher on a test of financial knowledge are less influenced by ratings and less prone to naively diversify, behavior which is harmful in our setting. Our conclusion is that ratings can affect decisions, and that they have less of an effect on more sophisticated subjects. Author(s): Robert McDonald; Thomas Rietz The Role of Social Information in Negotiations. Mary Rigdon Rutgers University ABSTRACT: The gender gap in wages persists (DeNavas-­‐Walt, et al., 2007; Goldin, 1992). In 2011, the ratio of the averages of women’s and men’s median annual earnings in the United States was 0.77 (Hegewisch, et al., 2012). This disparity persists even when industry, firm characteristics, education, tenure, job title, and job performance are controlled for in analyses: Gerhart (1990) reports that about 30% of the gender differences in salaries in a large private firm can be explained by differences in starting salaries. Since the wage gap is not explained by differences in skill or ability, and since it has its roots in differences in starting salaries, a plausible hypothesis is that gender differences in negotiation at the time of hire is a major factor. The wage gap is largely due to a negotiation gap. However, very little empirical research has addressed this issue directly, and what data has been collected is primarily either anecdotal or collected in unincentivized environments (see Barron (2003); Bowles, et al. (2005); Small, et al. (2007)). The purpose of this research is two-­‐fold. First, to investigate whether females ask for less in a controlled bargaining setting that accurately models relevant aspects of negotiating over a starting salary. Second, if females do ask for less, to investigate a simple mechanism — making relevant social information available — that naturally shrinks the negotiation gap. The results are stark: women ask for less, and earn less than their male counterparts. Providing social information—offer demands being made by others in a similar negotiating situation—
directly influences the beliefs females have about the social norms of negotiation. They ask for more, eliminating the negotiation gap. Crucially, their demands are met, eliminating the wage gap. Author(s): Mary Rigdon Altruism, Group Identity and Political Participation: An Experiment Pedro Robalo Tinbergen Institute -­‐ University of Amsterdam ABSTRACT: We study the relationship between political participation, altruism and group identity using an experimental framework. Combining a participation game (Palfrey and Rosenthal, 1983) with an original group identity induction procedure and a measurement of other-­‐regardingness, we find that the individual decision to participate in politics is increasing in identification with one’s group and other-­‐regarding concerns. We manage to implement environments where the group attachment of individuals varies, which allows us to conclude that the impact of group identity on aggregate turnout is not significant. Author(s): Pedro Robalo; Arthur Schram; Joep Sonnemans Firm Competition and the Evolution of Compensating Wage Differentials Andrea Robbett Middlebury College ABSTRACT: Wage differentials arise to compensate workers when their jobs are seen as more arduous or risky than other jobs. Firms need to pay workers more to entice them to do the less attractive jobs, such as working at night or working in industries with greater health risks. We ask whether behavioral characteristics of the workplace can also lead to wage differentials. We report the preliminary results of an experiment in which firm managers compete over a mobile workforce and workers’ observable productivity in certain firms is susceptible to noise or sabotage by co-­‐
workers. Author(s): Jeffrey Carpenter; Peter Hans Matthews; Andrea Robbett Bayesian and Reinforcement Learning: An Experimental Comparison Kathleen Rodenburg University of Guelph ABSTRACT: In this laboratory experiment I test how subjects (180) behave when presented with a binary-­‐
choice decision with the option to purchase additional information before reaching a terminal choice. In addition to testing whether subjects possess the cognitive sophistication, and if so, whether they choose to apply Bayesian Expected Utility (BEU) theory when presented with new corresponding information, I test whether they use reinforcement learning(RL) when attempting to optimize their action choices. I find pervasive evidence of both heuristics influencing decision choice. I find: )subject’s behavior is less divergent from the risk-­‐neutral BEU model than that of the partial RL decision model used in this paper ; 2) improved performance when the message received is compatible with the higher state contingent ex-­‐ante payoffs and the opposite when the reverse is true; 3) RL enhancing decision optimality when the two behaviour types are aligned and diminishing it when the two heuristics clash; 4) the existence of a crossover threshold where BEU is used for tasks that are simple and RL when tasks become more complicated; And, 5) in repeated rounds, the informative power of the message signal influences the ability of the RL to converge on or compromise BEU optimal decision choices. Author(s): Kathleen S. Rodenburg Is It How You Look or Speak That Matters? -­‐ An Experimental Study Exploring the Mechanisms of Ethnic Discrimination Magnus Rodin The Swedish Institute for Social Research ABSTRACT: Using a unique laboratory experiment where subjects are asked to guess the test performance of candidates presented by facial portraits and voice messages, this paper explores the following questions: Are beliefs about performance affected by if a candidate is perceived to have looks that are non-­‐stereotypical for the dominant population and do these beliefs change if the candidate has native-­‐like versus accented speech? The experiment is conducted in Sweden and the results show that candidates not perceived as stereotypically Swedish are considered to be worse performers. These beliefs are found in within-­‐gender but not in cross-­‐gender evaluations and are not eliminated when additional performance-­‐related information about the candidates is provided. When candidates are presented by both looks and speech, differential evaluations based on looks disappear. Instead, we find strong negative beliefs about performance for candidates that speak Swedish with a foreign accent implying that ethnic stereotypes associated with speech override stereotypes associated with appearance. The negative beliefs associated with foreign-­‐accented speech are not supported by corresponding mean differences in the candidate’s actual test performance. Author(s): Magnus Rodin; Gulay Ozcan Carry a big stick or no stick at all: A trust game with endogenous punishment Ismael Rodriguez-­‐Lara Universidad de Valencia ABSTRACT: We consider a trust game with a punishment-­‐phase in which the trustor is giving the possibility to punish (at a cost) the trustees behavior after both subjects play the game in Berg, Dickhaut and McCabe (1995). Our experimental design is based on a within-­‐subject design with trustors and trustees receiving a different endowment in each round. As a result, the trustors capacity to punish is endogenously determined by the proportion of the endowment that they want to keep in that round. We compare the subjects behavior in the game when there is punishment (PUN) and when there is not (NOPUN). Our results suggest that the capacity of punishment (i.e., the difference between the trustor and the trustees initial endowments) crucially affects to the effect of punishment on the level of trust. Overall, we find that punishment is effective to increase the level of trust only if the trustor is at a relative advantage with respect to the trustee. We also find evidence for an endowment effect, meaning that when the trustor has a higher endowment, she transfers an smaller proportion. Finally, our results suggest that there exists an order effect so the moment in which the institution is introduced is important so as to explain the results. Thus, when PUN is introduced after NOPUN, there exists evidence for the crowding-­‐out effect except if the trustor is at relative advantage with respect to the trustee with regard to the initial endowments. In the case in which PUN preceeds NOPUN, we find than the level of trust significantly decreases when the trustor is at relative advantage, while keeping constant in the rest of distributions. The rationale for these results relies on the endogenous capacity of punishment and its effect on the intrinsic and extrinsic motivation, which also affect the level of trustworthiness. Although we find no correlation between the amount sent and the returned amount, there exists evidence that PUN increases trustworthiness except when the trustor is at relative advantage with respect to the initial endowments. Our paper also discusses the efficiency of punishment and the conditions under which the punishment is used by the trustor. Author(s): Vicente Calabuig; Enrique Fatas; Gonzalo Olcina; Ismael Rodriguez-­‐Lara Moral Emotions and Opportunistic Behavior: A Subliminal Priming Approach Daniel Roemer Uni Heidelberg ABSTRACT: In this paper we report an experiment in which we used subliminal priming techniques to study the differential effect of two key negatively valenced moral emotions, shame and guilt, on opportunistic behavior. In particular, we use the cheating task of Fischbacher and Heusi (2008) to analyze cheating opportunities that do not involve a direct effect on any co-­‐player’s payoff. Our subliminal priming approach stems from the literature on social psychology and is similar to Zemack-­‐Rugar et al. (2007) who showed that individuals remained unaware of the priming, and yet behaved consistent with predictions concerning the specific emotion concept primed. We find that priming guilt and shame lead to different levels of opportunistic behavior. In terms of policy implications, we argue that punishments aimed at inducing shame in the target may even lead to counterproductive effects. Our contribution to the economic literature is threefold. First, we identify drivers of opportunistic behavior. Second, we provide one of the first economic studies using subliminal priming techniques. Third, we aim to deepen the integration of research on emotions in economics and psychology by highlighting the importance to draw clear distinctions between superficially similar emotions. Author(s): Brigitte Eisenkolb; Andrew Isaak; Johannes Jarke; Daniel Roemer Predictive Repeated Game Theory: Measures and Experiments Julian Romero California Institute of Technology ABSTRACT: One of the fundamental results in repeated games is the Folk theorem, which predicts a plethora of equilibrium outcomes. Many have argued that this extreme number of equilibria is a virtue, as it can explain a variety of different behaviors. However, this result leaves us with almost no predictive power. This paper provides measures for evaluating the predictive power of a theory given experimental data. After running experiments with human subjects in the experimental laboratory, we use these measures to compare (subgame perfect Nash) equilibrium theory, Mathevet (2012)’s axiomatic approach, and Ioannou and Romero (2012)’s learning model. Author(s): Laurent Mathevet; Julian Romero Promoting Star Networks with Homogeneous Agents Rong Rong ICES, George Mason Univ ABSTRACT: Information dispersion is key to many economic decisions, occurring through market systems as well as networks of agents (Jackson, 2009). Empirical and theoretical findings suggest that an efficient network for information dispersion takes the form of a star: small numbers of agents gather information and distribute it to a large group (Weimann, 1994; Rogers, 1995; Bala and Goyal, 2000; Galeotti and Goyal, 2010). However, testing of this theory in the lab has generally failed to find evidence that these networks emerge. An exception is Goeree et al (2008) that finds frequent star networks under the assumptions of ex ante agent heterogeneity and perfect information. Unfortunately, these conditions do not likely characterize natural environments (Feick and Price, 1987; Conley and Udry, 2010), and thus the conditions that promote star networks remain largely unknown. Drawing from institutions that exist in natural environments, we design experiments to shed light on this issue. We find that investment limits and the ‘right-­‐
of-­‐first-­‐refusal’, both of which are regular features of naturally occurring star networks, have a surprisingly strong ability to promote the formation of star networks even when agents are ex ante homogeneous. Using a cluster analysis, we are able to trace the large positive effects of these institutions to the impact they have on individual decision rules. In particular, we find that these institutions encourage individual rationality as well as positive habits that, it turns out, facilitate the formation and stability of star networks. Our results may have important implications for environments characterized by ex ante homogeneous agents, such as is found in the technology and agricultural sectors. Author(s): Rong Rong; Daniel Houser Learning Under Ambiguity -­‐ An Experiment Yaroslav Rosokha The University of Texas at Austin ABSTRACT: The aim of this paper is to understand the learning process of the agents under risk and ambiguity. We design and conduct an economic experiment where participants are required to choose over pairs of lotteries involving urns of unknown compositions. The composition of risky urns is the result of a known randomization device while the composition process of the ambiguous urn is unknown to the participants. The participants are required to make sequential choices over the lotteries, allowing us to track the beliefs of the agents over time. We gather data for seventy eight subjects divided in two treatments based on the order at which the risky and ambiguous urns were presented to them. First, we estimate the risk aversion parameter under a CRRA utility specification. Then, we estimate the initial beliefs and the corresponding parameters of three different learning models: (1) Bayesian Updating with base rate fallacy; (2) Bayesian Updating with multiple priors; (3) Reinforcement Learning Heuristic. The results here presented are at the aggregate level. For the reinforcement models participants who observed the risky lotteries first anchored their initial beliefs around the prior implied by the reduction of the compound lotteries, while participants who observed the ambiguous lotteries first anchored their initial beliefs at a significantly lower value. The updating behavior, however differs for risky environments and for ambiguous environments. For the multiple priors the results suggest that consideration set varies significantly depending on the order of presentation. Author(s): Othon Moreno Gonzalez; Yaroslav Rosokha Counseling vs. Scoring: How personal interaction affects the assessment of risk preferences Benjamin Roth Heidelberg University ABSTRACT: We investigate the assessment of risk preferences of others under two conditions: either the subjects meet in personally or make their assessment based only on sociodemographic information. Risk preferences are an important criterion when making a decision. Nevertheless, many important decisions are delegated to professionals, such as financial advisors or doctors. Personal interaction is a common element in such counseling interviews, but also costly for both parties. In times of technological change, it is therefore of interest whether a personal meeting helps to improve the quality of advice. By employing a within-­‐subject design, we investigate the prediction accuracy of risk preferences under both conditions. Our results indicate that personal interaction is not a major advantage when assessing the risk preferences of others. Prediction errors remain similar. Author(s): Andrea Leuermann; Benjamin Roth The impact of advice on selection into competition Christina Rott Universitat Autònoma de Barcelona ABSTRACT: We conduct a laboratory experiment, in which subjects decide whether to enter a real-­‐effort tournament. In our treatment group subjects receive advice on whether to participate in the tournament. Advice is given by subjects who have already had some experience with the same decision and have some informational advantage. Consistent with findings in the previous literature, in our control group (without advice) men enter significantly more often than women, that is there is a gender gap in entry. We show that, overall, advice improves the entry decision of subjects, in that the total costs of wrong entry decisions go significantly down. This is mainly driven by increased entry of high ability women and reduced entry of low ability men. The effect of advice on the gender gap in entry is mixed. The gap becomes small and insignificant in the low and high ability groups, whereas it increases in the middle ability group. This increase can be ascribed to gender differences in reacting to advice: Men in the middle ability group follow more the advice to enter the tournament, whereas women in the middle abilitiy group follow more the advice not to enter the tournament. Finally, we also find that women are slightly more hesitant than men to advise entry into tournament. Author(s): Jordi Brandts; Valeska Groenert ; Christina Rott Individual Meta-­‐Cognition and Group Decisions Nicolas Roux Paris School of Economics ABSTRACT: We study the ability of each subject of dyad to combine its private information with his partner’s in a signal detection task. Each subject observes two visual stimuli, and decide which one is stronger based on noisy information provided by its visual system; we also elicit confidence in choices by probabilities matching. The two subjects then freely exchange their opinions and each of them subsequently makes a second decision. Whether the opinions exchanged allows improving the success rate of subjects depends on their ability of combining their private information. Recent research has identified strong failures of group decision making in perceptual task in the sense that groups could be outperformed by the best of their members. It is not clear yet whether the failure comes from some social mechanisms within the group or from an inability of subjects to integrate dispersed informations. Our design allows isolating the informational issue. Preliminary results show that no subject was made worse off by the opinion exchange. However, the assumption of full perceptive information sharing is rejected based on the success rate after opinion exchange. We then relate the lack of performance improvement after discussion to the subject’s metacognitive abilities. We hypothesize that one’s partner’s discrimination ability has a positive impact on one’s performance improvement and that the difference between the two subject’s calibrations has a negative impact. Preliminary results support those assumptions. Author(s): Sebastien Massoni; Nicolas Roux Behavioral Response to Non-­‐price Interventions and Delivery Methods Galib Rustamov Claremont Graduate University ABSTRACT: Climate change and the increase in greenhouse emissions have been two of the most controversial fields in recent history. In response, economist and policy makers have focused on some important tools to change consumer behavior and increase energy savings. I focus on the more recent approach to increase customer energy conservation: non-­‐price interventions. This paper discusses the role of non-­‐price interventions to affect consumer energy conservation in general, and specifically analyzes the customer-­‐specific information and usage data. Empirical analysis of this specific utility information program will be used to draw some conclusions broadly applicable to similar interventions. The research is intended to discover whether the method online or mail by which consumer receives information about own usage behavior affects savings and behavior. The study focuses on whether instant desire to change the energy usage behavior with online feedback is more effective than mail-­‐in where feedback take few days after the program participation. The consumer is free to choose whether he/she will use online version of the survey or mail-­‐in version of the survey, considering the both group shows some ‘unobservable’ factor to participate the survey. Both surveys are offering same questionnaires, and provide similar feedback. The paper also addresses the economic and policy implications of the research. Author(s): Galib Rustamov Using Behavioral Economic to Increase the Efficacy of Teacher Incentives Sally Sadoff UCSD ABSTRACT: Despite growing enthusiasm among education reformers and policy makers for teacher merit pay, no previous experimental study has demonstrated that these incentives can have a positive impact on student performance in the U.S. In this paper, we design teacher incentives that exploit insights from behavioral economics particularly the power of framing. We compare a typical ‘Gain’ frame in which teachers receive payments at the end of the year to a ‘Loss’ frame in which we endow teachers with payments at the beginning of the school year that they must return if their students do not meet performance targets. We find that the loss frame is powerful, increasing student performance by 0.22 standard deviations. Importantly, the difference in achievement gains between the ‘Loss’ and ‘Gain’ groups are statistically significant. This suggests that the divergence in our findings from previous results is due to the addition of framing in the incentive design rather than differences in settings, reward structure or implementation. Author(s): Roland G. Fryer; Jr.; Steven D. Levitt; John List; Sally Sadoff Can Spiteful Behavior Survive?: An Evolutionary Dynamics Approach Tatsuyoshi Saijo Osaka University ABSTRACT: Researchers have been using a few strategies for analyzing social dilemma, for example, cooperation and defection in prisoner’s dilemma games. On the other hand, continuous strategy models with non-­‐linear utility functions have been used for analyzing public good provision problems, for example. Our first question is whether these dilemma models with a few strategies can be obtained from the continuous models. We find that altruistic and spiteful behaviors as well as Nash and Pareto behaviors are embedded in such models. Cason, Saijo and Yamato (2002) found that some Japanese subjects chose spiteful strategies, and American subjects rarely chose them in a continuous model. They also found that spiteful behavior among Japanese subjects led the high provision of the public good in later periods, but American subjects chose an ESS path and hence the level of the public good is lower than that of Japanese subjects. Based upon these observations, we set up a three strategy model with spiteful behavior and using evolutional dynamics, we find that the simplex of the population is divided into two regions where one leads to an ESS outcome, and the other leads to the mixture of spiteful and Nash strategies. The latter attains higher level of the provision than the former does. That is, spiteful behavior can survive and leads to high efficiency of the public good provision. Author(s): Mayuko Nakamaru; Tatsuyoshi Saijo Procurement Auctions with Renegotiation and Wealth Constraints Tim Salmon Southern Methodist University ABSTRACT: In many procurement auctions the costs of providing the good are unknown at the time of the auction. Using a standard low price wins format can therefore lead to substantial losses among the winning sellers who might be suffering from the Winner’s Curse. Such seller losses can be a problem for the buyer if they lead to a seller not being able to complete a project. Renegotiation of a contract price is a standard way that parties attempt to get around this problem. While in principle it might help overcome the problem of seller losses, when sellers know they have the option of renegotiating after an auction it substantially alters the incentives for bidding in the auction stage and could in fact cause any such problems to worsen. We investigate these issues in an environment with sellers who have differing abilities to absorb losses in an attempt to understand whether the ability for a winning seller to renegotiate a sales price can help to overcome the Winner’s Curse and if so, at what cost to the buyer. We also investigate how the presence or absence of the renegotiation option affects the type of seller that wins the auction stage. Author(s): Wei-­‐Shiun Chang; Tim Salmon; Krista Jabs Saral A Model of Memory Load that Explains Important Unexplained Patterns in Search Behavior Adam Sanjurjo UCSD ABSTRACT: Memory limits are a fundamental bottleneck in human decision making. It is no surprise, then, that they may play a prominent role in determining the way people search for information. This paper introduces a simple model of (working) memory load in search, then characterizes the minimum memory load search sequence. This result is then compared to experimental data. Experimental subjects´ behavior in rich search contexts has recently been shown to deviate from optimality in systematic ways. It is shown here that these systematic deviations are consistent with memory load minimization (while inconsistent with the standard candidate explanations). Further, in a new experiment, increased search memory load is found to increase decision error. Together, these results suggest that subjects´ observed systematic deviations from "optimal" search may constitute a rational adaptation; a searcher reduces choice error by reducing her induced memory load (at the expense of acquiring slightly less valuable information). Author(s): Adam Sanjurjo Asymmetric Information in Contests: Theory and Experiments Rajiv Sarin Texas A&M University ABSTRACT: This paper examines common-­‐value contests with incomplete and asymmetric information. In our experimental design one bidder observes an informative signal as to the realized common value of the good. The other bidder holds only public information; she knows only the distribution from which the value of the prize is drawn. This asymmetric information enviroment is compared to a symmetric information environment in which neither bidder observes a signal; both bidders know only the distribution from which the value is drawn. We characterize the equilibrium in a common value all-­‐pay auction with this type of information asymmetry. Consistent with theory, such asymmetric information yields information rents for the informed bidder in both all-­‐pay auctions and lottery contests. Also consistent with theory, asymmetric information reduces the expected revenue in all-­‐pay auctions. In lottery contests, information asymmetry has no effect on revenue. We also observe that bidders who observe a signal are much more prone to bid above a break-­‐even bidding strategy than are bidders who do not observe a signal. Author(s): Brit Grosskopf; Lucas Rentschler; Rajiv Sarin The Dynamics of Social Preferences in Repeated Trust Game Experiments Markus Sass Otto-­‐von-­‐Guericke-­‐University Magdeburg ABSTRACT: Recent experimental evidence from repeated dictator games (Brosig et al. 2010) and public good experiments (Sass & Weimann 2011) suggests a time-­‐dependent instability of social preferences with subjects becoming more selfish the more often they are put in the same decision situation. We conduct repeated trust games in order to confirm this finding in yet another situational context in which social preferences are believed to play an important role. In our experiment, the responders play the trust game four times in one week intervals with weekly changing proposers. We elicit our subject’s preferences by employing Selten's strategy elicitation method. In the NoFeedback-­‐Treatment, the responders neither learn what amounts of money the proposers sent them nor are they paid their earnings before the end of the experiment. This experimental design excludes endowment effects and effects due to learning about other’s behavior since beliefs about social norms cannot be updated. In order to test Plott's hypothesis of discovered preferences (Plott 1996), we also use a control treatment, in which feedback is given about the amounts of money sent, however earnings are again not paid before the end of the experiment. After the last round of trust games, the responders take part in a symmetric solidarity game (Selten & Ockenfels 1998), in which they are matched with inexperienced subjects. This design sheds light on the question whether a time-­‐dependent change in social preferences in the trust game has an impact on behavior in a different game with different social norms. Author(s): Markus Sass; Joachim Weimann White Milk or Chocolate? A Field Experiment on the Impact of Prompts and Incentives on Student Lunchroom Choices Anya Savikhin The University of Chicago ABSTRACT: Almost a third of US children ages 2-­‐19 are deemed overweight or obese. The school lunchroom provides a teachable moment for educators to engage children in making healthful choices. We conduct a field experiment with over 2,500 participants in grades K-­‐8 and systematically evaluate the impacts of prompts and incentives on the selection and consumption of milk in the school lunchroom. At baseline, only 20% of children select the healthier white milk relative to 80% choosing chocolate milk. We find that emphasizing the health benefit of the white milk during choice increases selection and consumption of white milk. We achieve the greatest positive effect when we emphasize that the white milk tastes good's or provide a small non-­‐
monetary incentive for selecting white milk. According to the theory, prompts regarding taste may be more effective than prompts regarding health benefits due to the high time discount rates of children. Post-­‐intervention effects are also discussed. Author(s): Anya Savikhin; John List The Cost of Information and Its Use: An Experiment Rei Sayag Erasmus University Rotterdam ABSTRACT: The irrelevance of historical costs on rational decision making has been the subject of much interest in the economic literature. This paper explores whether individual decision making under uncertainty is affected by the cost of information. Specifically, we investigate if identical information is used differently depending on its cost. The typical explanation for any such difference in behavior is the sunk cost fallacy. Outside of the lab, it is difficult to disentangle the effect of the cost of information itself from the effect of self-­‐selection by individuals who tend to gain the most from this information. To examine this, we create an environment in the lab in which individuals have to make an investment decision under uncertainty. Individuals are offered additional, useful and identical information on the state of the world, which can be used in a Bayesian fashion. Our treatments vary in the process by which information is made available to subjects -­‐ freely, voluntarily at a cost, or imposed at a cost. This design allows us to distinguish between selection, sunk costs, and other behavioral biases. We find that individuals who choose to purchase costly information tend to react more strongly to it than individuals who receive free information. We do not detect any difference between the behavior of individuals who must pay for information and those who receive it freely. The sunk cost fallacy does not seem to explain the observed data. Author(s): Pedro Robalo; Rei Sayag Double or Nothing Carsten Schmidt Universtity of Mannheim ABSTRACT: We report the results of a field experiment conducted in a cocktail bar in Mannheim, Germany. Patrons in the bar have a 50% chance of receiving a 50% reduction in their bar tab. If they receive the reduction, they can opt to bet "double or nothing", a 50% chance of having their bill completely voided or paying the original amount. We study the demographic correlates of safe and risky decisions. We find that smokers are more risk tolerant, there is more risk aversion for larger bill amounts, and groups are more likely to make risk averse decisions than individuals. Author(s): Charles Noussair; Carsten Schmidt One Good Deed a Day? -­‐ Experimental Testing of Prosocial Saturation Jan Schmitz University of Hamburg ABSTRACT: In this paper I run a series of laboratory experiments to estimate and describe to what extent an individual’s social behavior is saturated by charitable contributions. Running a multi session dictator game for an identical group of individuals over different time spans, I investigate whether and to what extend pro-­‐social behavior has diminishing or saturated ‘warm glow’ returns. Individuals are randomly divided into two different groups and participate in a series of experiments facing subsequent decisions to divide 10 Euro between themselves and a charity. A first group runs through two sessions the same day facing their donation declaration. The second group of individuals is retreated in a weekly turn. The saturation of individual’s social concerns is highest for the first group and lowest for the last. On average, subjects tend to give more to charity, the more time has passed since the last good deed. Further, I find support for a licensing effect. Subjects behave more selfish when asked to donate a second time. Author(s): Jan Schmitz Personality Traits and Behavior: The Role of Conscientiousness and Agreeableness in Competitive Social Dilemma Situations Arthur Schneider Trinity College ABSTRACT: In this study we investigate the relationship between personality traits and the level of cooperation in competitive social dilemma situations. We assess whether grouping players according to their Conscientiousness and Agreeableness scores affects the level of contributions and cooperation in a public goods game with punishment. We measure Conscientiousness and Agreeableness scores using the NEO-­‐PI-­‐3 inventory. We conduct four different treatments in which subjects with similar and dissimilar Conscientiousness and Agreeableness scores are grouped together. Overall, we find that grouping subjects with similar Conscientiousness scores produces the highest overall level of contributions, earnings, and surplus preserved. We conjecture that grouping people with similar Conscientiousness scores together will increase overall efficiency of any organization. Author(s): Viktoriya Lineva; Arthur Schneider Long-­‐Term Commitment and Cooperation Frederic Schneider University of Zurich ABSTRACT: In many contracts, cooperative behavior is not directly enforceable. We investigate how an enforceable contract feature, contract duration, can foster mutual cooperation indirectly. In a laboratory experiment, we study cooperation in a repeated Prisoner’s Dilemma game (PD) with varying and endogenous durations of interaction between partners. Subjects play a PD for a total of 150 periods, divided into 15 sets of 10 periods each. We consider two baseline treatments in which subjects are either randomly re-­‐matched every period or remain paired with the same partner for 10 periods before being re-­‐matched. In a third treatment, subjects can choose at the beginning of each set whether they want to play a string of 10 one-­‐shot games (one-­‐shot option) or 10 periods with the same partner (commitment option). We also elicit subjects’ social preferences at the beginning of the session, using a sequential PD. We find that (1) the choice of commitment contracts is initially close to 50 percent but increases over time, (2) subjects who are conditional cooperators are the first to choose the commitment option, (3) endogenously chosen commitment yields higher cooperation rates than exogenously imposed commitment, (4) selfish subjects initially choose the one-­‐shot option, with cooperation rates quickly unraveling to zero, but then follow into the commitment option and (5) with chosen commitment, full cooperation is achieved in all but the very last periods of a set. Author(s): Frederic Schneider; Roberto Weber The Public Loss Game -­‐ An Experimental Study of Public Bads Stephan Schosser Lehrstuhl fur Wirtschaftstheorie ABSTRACT: We analyze cooperative behavior of participants who faced a loss. In particular, we extend the Public Good Game by a fixed loss in the beginning of every period. We show that humans change their behavior compared to corresponding studies with gains only. First, in contrast to literature on gains, we observe significant order effects. When participants first play a treatment with punishment, they cooperate less and face higher punishment costs than when first playing a treatment without punishment. The changes are that drastic that punishment does not pay in the first case, while it does in the later. Second, for participants first playing without punishment the contributions in the very first period of play determine the contributions throughout both treatments of the game, yielding higher contributions in the punishment treatment than when playing with gains. Participants punishing first, show no comparable behavior. Author(s): Stephan Schosser; Bodo Vogt Counterproductive Work Behaviour: An Experimental Study Marina Schroeder University of Magdeburg ABSTRACT: We provide experimental evidence on the effects of incentives on counterproductive work behavior. Subjects are asked to do a task (identify euro coins) for a short period of time and to report the number of coins identified. They can engage in various types of counterproductive work behavior, i.e. erroneous identification, lying with respect to their own productivity or stealing coins. We vary the incentive scheme (competition, piece rate and fixed wage) and the level of incentives (low pay vs. high pay) across treatments. We find that under high incentives the number of errors significantly varies with the nature of incentives. The number of errors is highest for piece rate incentives and lowest under competitive incentives. Further, participants are much more likely to inflate the report under competitive incentives with high payment compared to all other incentive combinations. Author(s): Michele Belot; Marina Schroeder How to stop the disposition effect? The effects of stop loss and take gain orders Simeon Schudy University of Konstanz ABSTRACT: We investigate the role of stop loss and limit orders on the disposition to sell winning assets too early and losers too late in a laboratory experiment. Investors can buy and sell goods over a total of 30 periods. In one treatment, investors can only trade in trading periods whereas in the other treatment investors have the possibility to use limit orders (a stop loss and a take gain option). We measure the individual disposition effect (DE) as the difference between the proportion of winners realized (PWR) and the proportion of losers realized (PLR) using different reference prices (first, last, lowest, highest and weighted average purchase price). According to all reference prices the disposition effect is significantly smaller in the treatment with limits. The use of the lower limit (stop loss) increases PLR and thus reduce DE. The use of the upper limit instead increases PWR and thus DE. However, our results suggest that the mere confrontation with a limit decision reduces subject’s PWR as well. Author(s): Gerson Hoffmann; Urs Fischbacher; Simeon Schudy Do buyer groups facilitate collusion? Luis Manuel Schultz Dusseldorf Institute for Competition Economics ABSTRACT: Legal joint-­‐market institutions like buyer groups, trade associations or industry-­‐wide information systems are a rather common phenomenon. Both EU and US law allow for such exemptions from the general cartel prohibition. At the same time, it is well known that illegal cartels often abuse legal joint-­‐market institutions as a platform for their activity. In this paper we explore the impact of buyer groups in an experimental environment to test if the legal cooperation to purchase inputs jointly leads to collusion (tacit or explicit) on the product market as suggested by the new EU Guidelines on Horizontal Cooperation Agreements. We run five treatments in a three-­‐firm Cournot framework to identify the effects of (i) joining a buyer group, (ii) being able to exclude firms from the group, and (iii) being able to communicate with other firms in the buyer group. Our results identify communication as the main driver for increased collusive outcomes. However, without communication, endogenous buyer groups facilitate tacit collusion, especially when firms have the possibility to punish the deviator. The threat resulting from the possibility not to establish the buyer group or to exclude competitors from it leads to more collusive quantities. We, therefore, confirm the policy makers concern with buyer groups. Author(s): Hans-­‐Theo Normann; Jurgen Rosch; Luis Manuel Schultz Voluntary Disclosure and Perceptions of Fairness Steven Schwartz Binghamton University ABSTRACT: We conduct an experiment on voluntary disclosure within a simple bargaining setting wherein a proposer must choose from a set of two possible offers and a responder makes a binary accept/reject decision. In one treatment the proposer has the option to disclose whether a fairer (more equal) offer was available relative to the one chosen. In theory, lack of disclosure implies the offer is the less fair alternative. Therefore, voluntary disclosure should perform as well as mandatory disclosure in producing fair offers. We find that responders properly infer the meaning of non-­‐disclosure. However, despite the correct inferences made by responders, proposers submit twice as many fair offers with mandatory disclosure as with voluntary disclosure. Our results may speak to the importance of making some disclosures mandatory, even when in theory voluntary disclosure should be as effective. Author(s): Ulrike Denker; Steven Schwartz; Christopher Ward; Richard Young Does moving from war zone change emotions and risk perceptions? A field study of Israeli students Shosh Shahrabani The Max Stern academic college of Emek Yezreel ABSTRACT: The current field study uses data collected one week after the 2009 war between Israel and the Hamas militias in the Gaza Strip ended. The study examines emotions, perceived risks, and comparative optimism-­‐pessimism measure among 198 students exposed to rocket attacks and divided into three groups: (a) students who live permanently in the area, remained during the war in the attacked region, (b) students who live permanently in the southern region of the country (exposed to rockets) but preferred to move temporarily outside the rocket’s target zone during the war; (c) students who left the campus during the war and returned to their homes outside the rocket’s range. The results indicate that those who lived outside the rocket’s range felt the lowest of fear and the lowest perceived terror-­‐risk. However, no significant differences were found in the levels of fear and the perceived risk from terror in those who remained in the rocket’s range and those who decided to temporarily move out from the area under attack. In addition, we found that the intense emotions evoked by the missile attacks during war affected risk perception and decreased comparative optimism not only of those in the war zone, but also for those who live in southern Israel and left the area as a precautionary measure. These findings indicate the need to support not only civilians who live in the target area and remained during the war but also civilians who temporarily moved out of region. Author(s): Shosh Shahrabani ; Uri Benzion; Mosi Rosenboim; Tal Shavit Facing Your Opponents: Social Identification and Information Feedback in Contests Roman Sheremeta Chapman University ABSTRACT: We experimentally investigate the effect of social recognition and information feedback in a rent-­‐seeking contest, systematically varying the level of identification of subjects and the amount of information available about group member’s bids. Experimental results indicate significant over-­‐dissipation in all treatments. Identifying subjects through photo display decreases over-­‐dissipation. Providing information feedback changes the dynamics of individual behavior, but it does not affect aggregate dissipation in contests. We also find that subjects exhibit non-­‐monetary utility of winning, which is not significantly influenced by social recognition or by information feedback. Author(s): Shakun Datta Mago; Anya Savikhin; Roman Sheremeta Inter-­‐Generational Games with Dynamic Externalities Katerina Sherstyuk University of Hawaii, Dept of Economics ABSTRACT: Climate change mitigation, investment in public education and many other long-­‐term economic problems may be viewed as strategic games with dynamic externalities, where the actions of current players affect not only their present payoffs, but also the welfare opportunities for the future generations. We report on a laboratory experiment that investigates whether socially optimal actions may be sustained in such inter-­‐generational dynamic externality games. The results indicate that dynamic efficiency is harder to achieve in inter-­‐generational settings as compared to the traditionally-­‐studied settings with infinitely-­‐lived decision-­‐makers, even when the agent’s incentives are aligned across generations. The analysis of individual decisions and beliefs points at two distinct sources of difficulties brought in by the inter-­‐generational aspect. One is limited caring about the future. The second one is discrepancy between own actions and beliefs about the other’s actions, which becomes more pronounced in the inter-­‐generational setting as compared to the control setting with infinitely-­‐lived agents. The results suggest that it is important to make people aware about the future impact of their present actions, and to maintaining policies that are consistent across generations. We further find that inter-­‐
generational learning through history and advice from previous generations may improve dynamic efficiency, but may also lead to persistent myopia. Author(s): Katerina Sherstyuk; Nori Tarui; Majah-­‐Leah Ravago; Tatsuyoshi Saijo Are Good-­‐Looking People More Employable? Ze'ev Shtudiner Ariel University Center ABSTRACT: We investigate the role of physical attractiveness in the hiring process. We sent 5312 CVs in pairs to 2656 advertised job openings. In each pair, one CV was without a picture while the second, otherwise almost identical CV contained a picture of either an attractive male/female or a plain-­‐looking male/female. Employer callbacks to attractive men are significantly higher than to men with no picture and to plain-­‐looking men, nearly doubling the latter group. Strikingly, attractive women do not enjoy the same beauty premium. In fact, women with no picture have a significantly higher rate of callbacks than attractive or plain-­‐looking women. We explore a number of explanations and provide evidence that female jealousy of attractive women in the workplace and the negative perception of women (but not men) who include pictures of themselves on their CVs are the primary reasons for the punishment of attractive women. Author(s): Ze'ev Shtudiner ; Bradley Ruffle Entrepreneurial choice under ambiguity and the overconfidence bias. Anisa Shyti HEC Paris ABSTRACT: Uncertainty surrounds most entrepreneurial activities, yet there is little evidence on ambiguity attitudes in entrepreneurial decision making. Entrepreneurial decisions are also often related to overconfidence, or individual’s tendency to perceive themselves more favorably than they should. Although ambiguity and overconfidence both influence individual’s likelihood perceptions, there are few studies that examine how these two dimensions jointly influence choice behavior. One important finding in decision making under ambiguity is the ‘competence hypothesis’ (Heath & Tversky, 1991), which shows that individuals who feel skilled in a given uncertain situation are more ambiguity seeking than those who feel less skilled. Thus, we expect overconfident entrepreneurs to exhibit more ambiguity seeking behavior. This paper examines experimentally the impact of overconfidence on attitudes towards ambiguity with Executive MBA students. In the first part of the experiment, subjects were randomly assigned to one of three priming treatments. Each treatment consisted in a general knowledge test, where task difficulty was manipulated: easy, medium, and difficult, respectively for overconfidence, neutral, and underconfidence. In the second part of the experiment, after receiving the test feedback, subjects performed a set of binary choices between an ambiguous prospect, which offers €x with a probability between p_low and p_high, and a risky prospect, which offers the same outcome, €x, with a probability p_r. For a given binary choice set, the probability of the risky prospect, p_r, varies between p_low and p_high. To develop our hypotheses, we build on Budescu's et al. (2002) generalized Prospect Theory with an additional parameter for ambiguity. Our results confirm that overconfidence increases ambiguity seeking behavior. Author(s): Anisa Shyti; PhD Candidate; HEC Paris Complexity and Smart Nudges with Inattentive Consumers Stefania Sitzia UNIVERSITY OF EAST ANGLIA ABSTRACT: Consumer inertia plays a key role in a number of services markets. Typically, consumers do not switch service providers even though the tariffs they are holding are suboptimal. This paper aims to unpack two key psychological determinants tariff complexity and consumer inattention in the context of an experiment closely modelled on the UK electricity and gas markets. By employing an experimental methodology, we are in a position not only to identify the causal role of different psychological dimensions, but we are also able to test the effectiveness of policies designed to improve consumer outcomes. Policies we are able to evaluate are a regulatory bar on complex tariffs as proposed by the UK regulator Ofgem and two nudge manipulations where either advice of the existence of a better tariff or an automatic switch of default tariff takes place. Tariff complexity matters. It is a function not only of the complexity of the tariff for the individual good, which would fall under the remit of Ofgem's regulatory bar, but also of whether tied tariffs are provided for two goods rather than one. However, even with simple tariffs we show that, in the presence of a default tariff and of consumer inattention, markets are still affected by large amounts of consumer inertia. Similarly, providing advice on the existence of a better tariff does not improve outcomes. Intuitively, the reason why reducing complexity of the task or providing advice on the task solves the consumer inertia problem only limitedly is because subjects do not pay enough attention to the task in the first place and instead just stick to the default. A nudge policy of automatically switching default tariffs is a pragmatic and effective solution to obtain better consumer outcomes. The policy automatically changes default tariffs to the optimal one in a given time period, and by doing so it exploits inattention-­‐based consumer inertia to achieve better consumer outcomes while leaving consumers free to choose an alternative tariff if so they wish. Author(s): S. Sitzia; J. Zheng; D.J. Zizzo Registering to donate: A laboratory experiment Robert Slonim University of Sydney ABSTRACT: Blood shortages are common throughout the world and in most developed countries voluntary blood donations do not produce enough blood. However, sufficient pro-­‐sociality in these societies may not be the only (or possibly major) hurdle given massive donation spikes during emergencies such as 9/11 and Hurricane Katrina. Instead, frequent shortages during normal times may be due to difficulties in coordination and noisy demand signals. The current study proposes a donor registry that can solve the coordination and noisy demand signal problems; the registry invites donors to donate only when there is demand. The registry can also organize donors so that those with the highest willingness to donate will be most likely invited to donate, thus further improving social welfare. The current lab study explores five alternative methods to organize the registry not only to test its viability, but also to tease apart and understand the potential different avenues to efficiently run the registry. The treatments vary (1) whether donors can rank their willingness to donate and (2) whether donors can choose to participate in the registry. Allowing donors to indicate their willingness to donate can best coordinate an efficient equilibrium with heterogeneous donors. When the registry becomes more efficient, more potential donors will participate in the registry to coordinate their actions and overall efficiency will increase. Author(s): Robert Slonim; Carmen Wang Strategic and Intrinsic Pro-­‐Social Behavior of Individual Investors Paul Smeets Maastricht University ABSTRACT: We investigate pro-­‐social behavior by linking administrative investor data to a survey and to fully-­‐incentivized artefactual field experiments. We distinguish between strategic (financially motivated) and intrinsic socially responsible investors. We find that intrinsic pro-­‐social investor types also behave more pro-­‐socially in other domains: they donate more to charity, do more voluntary work and are more often registered as organ donors. This effect is particularly strong for domains in which signaling is difficult. Intrinsic pro-­‐social types also behave more pro-­‐socially in a fully incentivized anonymous one-­‐shot trust game. Interestingly, strategic socially responsible investors are significantly less pro-­‐social in the anonymous trust game than conventional investors. Author(s): Arno Riedl; Paul Smeets Optimism on the Honesty Game Alexander Smith Worcester Polytechnic Institute ABSTRACT: We present an experiment in which some subjects were able to take money from others. About half of ‘takers’ took the maximum and the average amount taken was 71%. Takers had accurate expectations about the amount of taking; takes underestimated the amount of taking. There is strong positive correlation between taking and beliefs about the taking of others, but we fail to find evidence that taking is related to gender, age, social networks or religiosity. When compared to other research on the prevalence of honest behavior, our results indicate that context is an important determinant of outcomes. Author(s): Alexander Smith Cognitive Load and Strategic Sophistication John Smith Rutgers University-­‐Camden ABSTRACT: We describe an experiment where we manipulate the cognitive resources available for the subject. We manipulate the availability of these resources by placing subjects under a differential cognitive load. We observe whether differences in the cognitive load treatment are associated with differences in the strategic sophistication of the subject. In cognitive load experiments, subjects are directed to perform a task which occupies cognitive resources, in addition to making a choice in another domain. The greater the cognitive resources required for the task implies that fewer such resources will be available for deliberation on the choice. Although much is known about how subjects make decisions under a cognitive load, little is known about how this affects behavior in strategic games. In addition to the cognitive load manipulation, we also observe choice in a series of one-­‐shot games which are designed to facilitate the measurement of the strategic sophistication of the subject. Author(s): Sean Duffy; John Smith Preference Stability across Games Hector Solaz CESS, Nuffield College, Oxford University ABSTRACT: In a series of eight one-­‐shot two-­‐person games, four simultaneous four sequential games, we analyze whether responses depend on the game or some individual characteristics. All games have a similar framework but different equilibrium predictions and subjects do not receive any feedback till the end of the experiment to avoid learning. The individual features are formally elicited through a battery of tests. In the sequential games we follow the ‘strategy method’ and have the second player decide on a contingent action for each and every possible first player move, without first observing this move. Our analysis centers on the degree to which subjects change their decisions depending on the context. Our results show that decisions in the four simultaneous games are positively and significantly correlated, and decisions in the sequential games are also weakly correlated. However simultaneous and sequential decisions are not related. Demographics seem to play a limited role, as gender and IQ weakly mediate these results, but risk attitudes play no role. Author(s): Hector Solaz; Enrique Fatas; Shaun Hargreaves-­‐Heap Has the Perception of Stock Return Volatility Turned (more) Realistic Past the Crisis? Doron Sonsino The College of Management ABSTRACT: Uncertainty regarding future returns has been invoked as a major explanation to low stock market participation. Panel estimations confirm that investors decrease stock holdings drastically past crisis experience, while experiments show that anxiety and related negative affects generally reduce the willingness to take financial risk. With this motivation, we run a field experiment to test if the extreme-­‐case predictions of potential investors at TASE turned realistic in the highly volatile markets of late 2010-­‐early 2011. The prediction problems were framed as professional consultation tasks and subjects were requested to choose from pairs of (randomly selected) familiar stocks and deliver 95% confidence limits for the 3 months return on selected stocks. The results suggest that the turbulent market conditions did not drive subjects into realistic forecasting. The average hit-­‐rate along 930 intervals was only 27.5% and only few subjects showed calibration rate of 80% or more. Eventual returns were lower than the 5% quantile predictions in 82% of the miscalibration cases, revealing a particularly strong form of unrealistic optimism. Interestingly, the data suggests that confidence lengths significantly increased as predictions turned more extreme. The pattern was replicated in a different experiment where subjects predicted returns from unidentified historical records, suggesting that overconfidence hazards may attenuate when market conditions get relatively extreme. Author(s): Doron Sonsino; Eran Regev "Certain Gains" in Understanding of Prospect Theory: Impact of Cognitive Ability, and How the Problem is Described Gaurav Sood Princeton ABSTRACT: In a series of survey experiments conducted over large heterogeneous sample of respondents using variations of the Asian Disease Problem, prospect theory, effect of making changes to question and response wording, and order of presentation of information, were tested. In the first test of Asian Disease Problem over a nationally representative sample, which indicated that the effect holds in an attenuated form in the national sample, tests of cognitive ability as a moderator of framing revealed the effect is greatest among the most cognitively able, contrary to some prior results. In a step missing from previous work, people were asked to provide reasons for their decisions. Analysis of the responses revealed a significant number of answers weren’t congruent with prospect theory. Author(s): Gaurav Sood; Jon Krosnick Social and Economic Dimensions of Climate Change: Endogenous Risk and Distributive Justice in Threshold Public Goods and Common Pool Resource Games. Barry Sopher Rutgers University ABSTRACT: We study economic and social dimensions of public goods provision and common pool resource situations by considering alternative sharing rules (proportional to wealth or equal sharing), distributions of initial endowments, or harvesting rights (equal or unequal), and risk (low or high risk of losing endowments if the public good is not provided or maintained, or of losing future benefits from a resource if it is over-­‐used). We are motivated to study these problems by actual proposed World Bank projects for addressing climate change in developing countries such as India and Bangladesh. For the case of public goods, we can conceive of a public good as being an instance of insurance, where the nature of the risk requires a communal response, or cooperation, if the insurance is to be provided. The game is a threshold public good game with refunds, meaning that there is a threshold, T, which must be met through voluntary contributions from the participants, in order that the public good may be provided. If the threshold is not met, then contributions are refunded, but there is a risk that all endowments will be lost (think of a flood control project that is only built if there are sufficient contributions). In the case of the common pool, we can think of the game as being a problem of jointly restraining current harvesting in order to ensure a stream of future benefits (think of restraining forest harvesting in order to secure carbon credits). We are particularly interested in exploring the tension between the distribution of initial endowments (or harvesting rights) and the distribution of net benefits, as far as how this affects the relative level of success of groups of participants in providing the public good, or restraining harvesting of a common pool resource. We have conducted a ‘field lab’ study with villagers in West Bengal as subjects. We are currently conducting for in-­‐depth laboratory studies at Rutgers, which will be the main focus of analysis. Author(s): Barry Sopher; George Joseph; Jeff Birchby; Ilhom Abdulloev Age differences in the reaction to incentives do older people avoid competition? Alec Sproten AWI, Heidelberg University ABSTRACT: The ‘aging employee’ has recently become a hot topic in many fields of behavioural research. With the aim to determine the effects of different incentive schemes (competition, social or increased monetary incentives) on performance of young and older subjects, we look at behaviour of a group of younger and older adults on a well established real effort task. We show that older adults differ from younger adults in their performance in all conditions, but not in the improvement between conditions. The age difference in performance is however driven by women. While we replicate the gender difference in competitiveness found in the literature, we do not find a significant age difference in competitiveness. Social incentives have an at least as strong or even stronger effect on performance than increased monetary incentives. This effect is driven by men; women do not show an increase in performance with social incentives. Author(s): Alec N. Sproten; Christiane Schwieren Paying Respect or Paying Money Andreas Staffeldt RWTH Aachen ABSTRACT: People are assumed to care about esteem received by others but little is known about symbolic gestures to express esteem and their ex ante and ex post impact on work performance. In our study we compare the incentive effect of monetary and symbolic rewards. Therefore, we conducted a real effort lab experiment in which agents worked for the benefit of principal’s payoff and principals could response by giving rewards. The symbols used in the experiment are raised thumbs which are costly for the principal but do not raise agent’s payoff. To test for the relevance of costs we designed an additional treatment with costless symbols. The results give evidence that costs are essential for the effectiveness of reward systems and that a low reward can have a detrimental effect on the subsequent work performance compared to no reward. The latter can be interpreted based on the crowding out literature and previous findings of low rewards being perceived as insulting. Surprisingly, this effect is true for both, monetary and symbolic rewards. Author(s): Andrea Hammermann; Andreas Staffeldt What Can We really learn from Public Good Experiments Christian Stahr Otto-­‐von-­‐Guericke-­‐University Magdeburg, FWW ABSTRACT: Real public good problems differ from public good experiments in two respects. First, real public good problems are characterized by large groups and a small MPCR, while public good experiments are conducted with small groups and a large value of the MPCR. Second, in experiments, the public good is simulated by monetary payoffs, while contributions to real public goods do not lead to any monetary payoffs. Using a connected laboratory we conducted experiments with large groups of 60 or 100 subjects and with small MPCR equals to 0.02 and 0.04. We showed that the MPCR has a significant impact, whereas group size does not. Furthermore, it turns out that there is no structural difference to standard public good experiments. In a second step we introduce a real public good without monetary incentives. In groups of 60 and 30 subjects, contributions are used to buy and cancel CO2-­‐Emission allowance provided by the European emission trade system. In this treatment, we find strong structural differences to the standard results in public good experiments. Contributions do not decay and we found a positive final round effect. We use personality traits of the players (Five-­‐Factor-­‐
Inventory (TIPI)) to further investigate the behavioral differences between the contributions to real and artificial public goods. Our dataset contains observations for 3,560 different subjects. Author(s): Weimann J.; Brosig J.; Hennig-­‐Schmidt H.; Keser; C.; Stahr; C. Price Bubbles Are Caused By Diffusion of Behavioral Norms David Stark Columbia University ABSTRACT: As we have witnessed recently, price bubbles can devastate markets, nations and the entire world economy. Bubbles are not limited to certain environments: they can appear in markets of high efficiency, such as in real estate, stock and derivatives. Economists observed bubbles even in experimental ultra-­‐efficient markets, where intrinsic values are known, Knightian uncertainty is eliminated, and interaction between traders is controlled. Yet we know little about the origins of bubbles, despite their frequent emergence in markets and although they are a recurrent theme in policy debates. Here we show that price bubbles are caused by the diffusion of behavioral norms. Such norms cause agents to offer and accept erroneous prices, which diverge from intrinsic values. Norms speared as agents observe each other’s behavior. The longer they interact, the more similar their erroneous behavior becomes. Even if agents are initially adept in pricing, they increasingly commit correlated errors when trading, leading the entire market to bubble. As theory on norms predicts, bubbles are particularly likely when there is little racial diversity among agents and when situational ambiguity is high. By tracing the diffusion of behavioral norms in experimental markets, we identify the mechanisms that create price bubbles. Author(s): Sheen S. Levine; David Stark; Edward J. Zajac Risk taking in a social context Julia Stauf University of Cologne ABSTRACT: We present a lottery choice experiment designed to study how risk taking is affected by social context. We study changes in risk taking if decisions affect not only one’s own payoffs but also those of an otherwise passive, second party. We compare the effect of negatively and positively correlated risky payoffs, and vary the amount of available information about others’ behavior. Our results suggest that social context generally makes risk taking both more conservative and more homogeneous. These phenomena go beyond what simple models of social behavior, developed on the basis of behavior under certainty, would predict. Author(s): Gary Bolton; Axel Ockenfels; Julia Stauf How insiders shape a market -­‐ Evidence from experimental limit order markets. Thomas Stockl University of Innsbruck, Institute for Banking and Finance ABSTRACT: We study how insiders affect the shape of laboratory limit order markets by varying the degree of insider competition and the information about the number of insiders present. We observe that (i) a higher share of insiders in a market generates more efficient prices, lower spreads and higher trading activity. (ii) Manipulation of the information on the number of insiders present does not significantly impact the shape of the market. (iii) independent of the number of insiders, information is reflected gradually in prices rather than instantaneously as suggested by game theoretic models for markets populated by multiple insiders. (iv) The distribution of returns (log of price changes) exhibits excess kurtosis, which increases in the number of insiders. Author(s): Thomas Stoeckl Sharing a Voluntarily Provided Common-­‐Property Resource: An Experimental Examination Brock Stoddard Indiana University ABSTRACT: The growing experimental literature on asymmetric power in social dilemma settings indicates that giving one person informational and/or decision-­‐making advantages significantly impacts a group’s voluntary provision of a public good. In this paper we investigate a sharing mechanism where one person in a group, the allocator, has power to divide a voluntarily provided common-­‐
property resource (CPR) between the other members of the group. We refer to this as the ‘allocator’ mechanism. The initial data analysis reveals the allocator mechanism results in significantly higher provision to a CPR compared to random and egalitarian mechanisms. Detailed group-­‐ and individual-­‐level analysis will be performed in the coming weeks. The allocator mechanism has been implemented by religious communities for nearly 200 years, such as those found in Mormonism, and multi-­‐organization charities in various forms with varying success. This is the initial experimental attempt to directly study this mechanism. The results are important for understanding the economical and behavioral incentives associated with this type of sharing mechanism. The results from this study will be informative in understanding successful and failed attempts to implement such mechanisms in the field. The insights gained from this simplified environment will also guide subsequent experimental and theoretical studies which contrast this mechanism to alternative sharing mechanisms. Author(s): Brock Stoddard; James Walker; Arlington Williams Measuring nonlinear and heterogeneous guilt aversion Sigrid Suetens Tilburg University ABSTRACT: Recent experiments suggest that people’s preferences not only depend on their choices but also on their beliefs of other people’s choices and on their beliefs of other people’s beliefs. Guilt aversion is such belief-­‐dependent preference that has received a lot of attention recently. An individual is considered to be guilt averse if he gets a negative utility from failing to live up to the beliefs of others. Typically, the sensitivity to guilt of a certain individual is taken to be constant, which implies that guilt aversion is linearly increasing in the shortfall between an individual’s second-­‐order belief of another’s payoff and the latter’s actual payoff. We design experiments in order to test whether guilt aversion is indeed a linear function or has another functional shape. Specifically, we characterize the shape of the guilt aversion function at the individual level. In our experiment we let subjects play binary dictator games in which the dictator chooses between a cooperative choice and a selfish choice. The ‘responder’ is asked for his belief about the move of the dictator, and the dictator is elicited his second-­‐order-­‐belief-­‐dependent strategy. Particularly, for all potential ‘responder’ levels of beliefs, the dictator is asked whether to make the cooperative choice or the nice choice. This method allows measuring guilt sensitivity at the individual level and-­‐-­‐-­‐given that each individual makes choices for three decision situations that would result in the same guilt sensitivity under the hypotheses the guilt aversion function is linear-­‐-­‐-­‐allows identifying the guilt aversion function at the individual level. Our experiment also includes control treatments in order to check whether the strategy elicitation method does not lead to different types of behavior and beliefs as the direct-­‐response method. Our experimental results suggest there is a considerable amount of heterogeneity across individuals. Author(s): Charles Bellemare; Alexander Sebald; Sigrid Suetens Cooperation and Assortativity with Endogenous Partner Selection Siddharth Suri Yahoo! Research ABSTRACT: The natural tendency for humans to choose with whom to form new relationships and with whom to end established relationships is thought to facilitate the emergence of cooperation. Helping cooperators to mix assortatively is believed to reinforce the rewards accruing to mutual cooperation while simultaneously excluding defectors. However, the relationship between endogenous partner selection, assortativity, and cooperation has been largely unexplored experimentally. Here we report on a series of human subjects experiments in which groups of 24 participants played a multi-­‐player prisoner’s dilemma game where, critically, they were also allowed to propose and delete links to players of their own choosing at some variable rate. Over a wide variety of parameter settings and initial conditions, we found that endogenous partner selection significantly increased the level of cooperation, the average payoffs to players, and the assortativity between cooperators. Even relatively slow update rates were sufficient to produce large effects resulting in cooperation levels over 80%. Subsequent increases to the update rate still had a positive, although smaller, effect. For standard prisoner’s dilemma payoffs, we also found that assortativity resulted predominantly from cooperators avoiding defectors, not by severing ties with defecting partners, and that cooperation correspondingly suffered. Finally, by modifying the payoffs to satisfy two novel conditions, we found that cooperators did punish defectors by severing ties, leading to levels of cooperation approaching 100% which persisted for longer. Author(s): Jing Wang; Siddharth Suri; Duncan J. Watts Who approves punishment? A comparison of groups and individuals in social dilemmas Fangfang Tan Max Planck Institute for Tax Law and Public Finance ABSTRACT: Peer punishment in civilized societies is often constrained in that the punishment proposed by one party must be evaluated and approved by a third party. In this paper, we study how third parties make decisions when they are empowered with the right to uphold or negate punishments proposed by participants in social dilemmas. We compare how groups differ from individuals when deciding whether to approve punishment proposals as an independent third parties, and how decisions may differ depending on whether punishment is imposed ex post or threatened ex ante. In all cases, third parties must specify the punishment amount when approving the proposals. We find that, in relation to single individuals, groups punish at lower frequency of punishment and with less severity. Interestingly, whether the punishment is threatened ex ante has a significant impact on group decisions but not individual decisions. Author(s): Fangfang Tan; Erte Xiao On attitude towards choice. Some experimental evidence of choice aversion. Benoit Tarroux University Rennes 1 ABSTRACT: This paper investigates how people value the freedom of choice. Do people find an additional satisfaction by having more options as the normative literature on freedom of choice would suggest? Or do they simply consider it as a source of annoyance? Or are completely indifferent to it, as would a purely consequentialist agent assumed generally by economic theory? The experiment consists in asking subjects to state their willingness to accept for different choice sets. This approach allows us to assess whether prior to making their decision, people appreciate a wider set of options or not. The main result is that the value of a choice set is significantly lower than the one of its preferred element, suggesting that choice is perceived as a source of disutility and that on average individuals are choice averse. Author(s): Fabrice Le Lec; Benoit Tarroux Social Pressure and Prosocial Acts: Theory and Experiments Dmitry Taubinsky Harvard ABSTRACT: I propose a model of social preferences that offers simple explanations for why social standards seem to change from one environment to another (e.g., List 2007; Bardsley 2008) or for why recipient’s beliefs impact giving (e.g., Andreoni and Bernheim 2009; Lazear, Malmendier, and Weber 2012). The model also makes a number of new predictions, which I test in a series of three experiments. In experiment 1, Dictators choose how much to give to Receivers in a series of convex choice-­‐sets that allow either a lot of giving or only a little giving. I find systematic WARP violations that are predicted by my proposed model, but are inconsistent with other social preferences models. In experiment 2, I give Dictators the option to opt-­‐out and keep the Receiver uninformed, but again vary whether the sharing environment (conditional on opting in) allows for a lot of giving or only a little giving. I find that Dictators are willing to pay a higher price to opt-­‐out when the sharing environment allows for a lot of giving. In experiment 3, I study a three-­‐person Dictator Game in which one Receiver is more informed than the other. Dictators are systematically biased in favor of the more informed Receiver, even when the informed Receiver can observe the bias. Author(s): Dmitry Taubinsky Risk Sharing and Adverse Selection Franziska Tausch Maastricht University ABSTRACT: Does adverse selection hamper the effectiveness of voluntary risk sharing? We investigate experimentally to which extent individuals of different risk types participate in risk sharing under several forms of uncertainty. Four treatments are implemented that differ with respect to stochastic dominance among individual’s risks. In the benchmark treatment risks are identical whereas in the other treatments risk heterogeneity is implemented. All treatments can be ranked with respect to the likelihood of adverse selection occurring. We find that in the benchmark treatment individuals share risk significantly less than theoretically predicted. In the treatments where individuals differ in their risk types we do observe adverse selection. The type of risk heterogeneity determines the magnitude of the effect in line with the theoretically predicted ranking. However, in all cases adverse selection is less strong than expected utility theory predicts. Notably, when first order stochastic dominance is present mostly high risk individuals share risk, whereas with second order stochastic dominance almost no adverse selection occurs. Author(s): Franziska Tausch; Arno Riedl; Jan Potters Unemployment Insurance, Search Intensity and Match Quality: An Experimental Study Brittany Teahan Purdue University ABSTRACT: This study investigates the effect of changes to the maximum potential unemployment insurance (UI) benefit duration and benefit level on individual job search behavior. We use a laboratory experiment to observe search effort and match quality under four distinct UI policies that differ in benefit level and/or maximum potential benefit duration. Subjects select a type of job to seek and a costly search effort level. Two vertically differentiated jobs differ in daily wage and probability of receiving a job offer given search effort. The better job pays a higher wage but is more difficult to find conditional on effort. Consistent with theoretical predictions, we find more generous UI policies increase the propensity of subjects to search for the better job. We also find evidence that subject search behavior is affected by previous search experience. Consistent with previous studies, search is efficient in terms of total earnings but subjects deviate from the optimal decision sequence. Author(s): Brittany Teahan Choice of Institutions: Experimental Approach in Fishery Management Mihoko Tegawa University of Rhode Island ABSTRACT: Quotas for special groups of students often apply in school or university admission procedures. This paper studies the performance of two mechanisms to implement such quotas in a lab experiment. The first mechanism is a simplified version of the mechanism currently employed by the German central clearinghouse for university admissions, which first allocates seats in the quota for top-­‐grade students before allocating all other seats among remaining applicants. The second is a modified version of the student-­‐proposing deferred acceptance (SDA) algorithm, which simultaneously allocates seats in all quotas. Our main result is that the current procedure, designed to give top-­‐grade students an advantage, actually harms them, as students often fail to grasp the strategic issues involved. The modified SDA algorithm significantly improves the matching for top-­‐grade students and could thus be a valuable tool for redesigning university admissions in Germany. Author(s): Mihoko Tegawa; Chris Anderson On the Decision to Explore New Alternatives: The Coexistence of Under-­‐ and Over-­‐Exploration Kinneret Teodorescu The Technion -­‐ Israel Institute of Technology ABSTRACT: The decision whether to explore new alternatives or to choose from familiar ones is implicit in many of our daily activities. How is this decision made? When will deviation from optimal exploration be observed? The current work examines exploration decisions in the context of a multi-­‐alternative choice task. In each trial, participants could choose a familiar option (the status quo) or a new alternative (risky exploration). The observed exploration rates were more sensitive to the common experience than to the average experience with exploration, causing under-­‐ exploration in ‘rare treasures’ environments and over-­‐exploration in ‘rare mines’ environments. In addition, the results reveal a decrease in exploration over time even when exploration is always optimal, and some exploration even when exploration is never reinforcing. These results can be captured with models that share two assumptions: (1) a distinction between two types of explorations: forward-­‐looking and backward-­‐looking; (2) during backward-­‐looking exploration decision makers tend to rely on small samples of past experiences. The predictive value of a 2-­‐parameter quantification of these assumptions is demonstrated. Author(s): Kinneret Teodorescu; Ido Erev Lying and Public Goods: Can Punishment Still Be Effective? Janna Ter Meer University of Cologne ABSTRACT: Past work has established that punishment in the standard, 4-­‐player, public good game is ineffective when feedback on contributions is noisy. We look at the case where the noise is not exogenously imposed, but generated by the participants themselves. In other words, participants can actively misrepresent their contributions. After making their contribution decisions, participants can report what they have done through an announcement to their fellow group members. We vary the degree by which participants observe the actual or announced contributions and introduce punishment. We find that as long as the actual contributions can still be observed with some accuracy, punishment is effective in maintaining contributions at intermediate levels. However, it is highly inefficient in terms of earnings compared to no-­‐punishment treatments. This is driven, firstly, by the fact that punishment does not phase out over the course of the game. Secondly, high contributors significantly reduce the amount of punishment they allocate, because they are deceived by the high announcements of the others. Surprisingly, we find no detrimental effect of anti-­‐social punishment. Finally, lying, although increasing over time in the high noise treatments, does not seem to be driving any treatment differences. Author(s): Bernd Irlenbusch; Janna Ter Meer The Effect of Negotiations on Bargaining in Legislatures Chloe Tergiman NYU ABSTRACT: In democracies, some bills reach the floor for a vote only after extensive negotiations and back room deals. Indeed, negotiations are an integral part of the bargaining process; members of the legislature spend a considerable amount of time communicating with each other in attempt to reach an agreements before bringing the bill to an actual vote. So far, laboratory experiments have largely ignored the issue of negotiations and communication in a legislative context. In this project we aim to fill this gap and introduce a negotiation stage in multilateral legislative bargaining experiments. We use the theoretical framework of the Baron-­‐Ferejohn (1989) model. Our experimental results show that allowing for explicit negotiations allows subjects to reach decisions that are closer to equilibrium predictions. We also show that type of communication matters and that backroom deals are essential for this. Author(s): Marina Agranov; Chloe Tergiman The 2-­‐person risk-­‐game Timm Teubner Karlsruhe Institute of Technology (KIT) ABSTRACT: We conduct a field experiment in which two subjects at a time participate in a roulette-­‐like dice-­‐
game: the 2-­‐person risk-­‐game. Subjects are recruited ad hoc on campus and are either known to each other (e.g. friends, fellow students, couples) or not. In the first stage of the experiment, participants separately and independently choose between a low and a high risk option, both yielding the same expected monetary outcome. In the second stage, participants are asked to choose among the same options again. Now, however, the prior choice of the respective other is visibly displayed on the game board. Simultaneously, they decide on how to couple own and the other person’s payoff (i.e. collective versus exclusive). From a classical micro-­‐economic perspective, there is no strategic interaction whatsoever: one’s payoff does not depend on the other person’s choice. Payoffs are linked, however. We find that a significant share of the subjects is affected by the other person’s choice and that strangers and friends show different behavioral patterns. We analyze the second mover’s decision problem from a strategic perspective by including weights for payoff differences and the relation towards the other person (envy, gloating, altruism). The observed behavior illustrates that other-­‐regarding preferences and risk attitudes interact in situations where people face both risk and the presence of other people. Our study addresses basic questions in view of private-­‐to-­‐private markets which put economic and risky decision making in a social environment. Author(s): Timm Teubner; Marc Adam; Christof Weinhardt Collusive Relationships and Targeted Punishment: Number Eff ects in Cournot Oligopolies Christian Thoeni University of Lausanne ABSTRACT: It is common wisdom that fewer firms find it easier to collude. Theoretical findings and experimental evidence support this result. In practice, however, cartels in markets with many firms and low concentration are observed. We investigate the option of targeting punishment at breaching firms as a possible explanation for successful collusion within larger groups. In a series of experiments comparing Cournot oligopolies with and without the possibility of targeted punishment, we vary the number of competitors in a market from two to eight as the circumstances under which targeted punishment may help collusion. We find that targeted punishment can stabilize collusion in markets with up to, at least, six firms more than without this option. Author(s): Catherine Roux; Christian Thoni Do Women Have More Shame than Men? An Experiment on Self-­‐
Assessment and the Shame of Overestimating Oneself. Carmen Thoma LMU Munich ABSTRACT: We analyze how subjects’ self-­‐assessment depends on whether its accuracy is observable to others. We find that women downgrade their self-­‐assessment given observability while men do not. Women avoid the shame they may have if others observe that they overestimated themselves. Men, however, do not seem to be similarly shame-­‐averse. This gender difference may be due to different societal expectations: While we find that men are expected to be overconfident, women are not. Shame-­‐aversion may explain recent findings that women shy away from competition, demanding jobs and wage negotiations, as entering these situations is a statement to be confident of one’s ability. Author(s): Sandra Ludwig; Carmen Thoma A simple questionnaire can change everything: are strategy choices in coordination games stable? Lora Todorova Otto-­‐von-­‐Guericke University Magdeburg ABSTRACT: We present results from an experiment designed to study the effect produced on strategy choices when a subject reports risk preferences before engaging in a coordination game. The main finding is that the act of answering a questionnaire about one’s own risk preferences significantly alters strategic behavior. Within a best response correspondence framework, this result can be explained by a change in either risk preferences or beliefs. We find that self-­‐
reporting risk preferences does not induce a change in subject’s beliefs. Our findings raise some questions about the stability of strategy choices in coordination games. Author(s): Siegfried K. Berninghaus; Lora R. Todorova; Bodo Vogt Do people have a preference for increasing or decreasing pain? An experimental comparison of psychological and economic measures in health related decision making Judith Trarbach Otto-­‐von-­‐Guericke-­‐University Magdeburg ABSTRACT: This paper investigates preferences for different health profiles, especially sequences of increasing and decreasing pain. We test conflicting predictions in terms of preferences over two painful sequences. The QALY concept relevant for the determination of different levels of health-­‐related quality of life implies indifference, whereas behavioral theories find preferences related to ordering, following the peak-­‐end-­‐rule. Using an experimental design with real consequences we generate decisions about painful sequences induced by the cold pressor test. The results are compared with hypothetical choice data elicited using standard methods. We find that hypothetical methods reveal decisions in line with the peak-­‐end-­‐rule. However when it comes to real consequences of their decisions, subjects are not willing to pay for that preference on average. Author(s): Judith N. Trarbach; Eike B. Kroll; Bodo Vogt Attention and Revealed Preference in a Portfolio Choice Experiment Stefan Traub University of Bremen ABSTRACT: Experimental economics has meanwhile collected overwhelming evidence that individuals systematically violate basic axioms of normative decision and game theory. A stock taking of replicable experimental designs (``exhibits’’) that produce such results is made, for example, by Bardsley et al. (2010). An important question that arises from carrying over the usually deterministic choice theories to the experimental lab is how to deal with errors made by the participants and noise in the data. In the end, a purely normative test would reject any sensible theory that places some restrictions on observable behavior. Allowing for a little noise or imprecision on the other hand (e.g., by adding an error term) involves the danger of dramatically reducing the power of the respective test to discriminate between intended and random behavior. An alternative way to accommodate theory with empirical evidence is to formulate non-­‐deterministic choice models (such as the Random Preference model). Beyond that several recent economic and psychological process-­‐oriented decision theories address the sequential allocation of attention in choice tasks in order to explain errors, noise, and choice anomalies (examples are Sims’ 2003 Theory of Rational Inattention tested experimentally by Cheremukhin et al. 2011; the Directed Cognition Model by Gabaix, Laibson and coauthors 2000, 2006; Decision Field Theory by Townsend and Busemeyer 1989; see also DePalma et al. 1994; Hensher et al. 2005; and Cameron, DeShazo 2011). In this paper we report on a laboratory experiment in which each of 41 student subjects face a series of 16 successive grouped portfolio selection problems. We classify subjects’ choices into consistent and inconsistent choices as to Varian’s (1982) generalized axiom of revealed preference (GARP) and check whether chosen portfolios are dominated in terms of first-­‐order stochastic dominance. While dealing with their choice tasks, we record the attention paid to each portfolio in terms of time spent on it. We compute the first four central moments of the distribution function of attention. Our preliminary data analysis suggest that the shape of the distribution function of attention has some predictive power for the consistency of choices. This result seems to be in line with some of the predictions made by process-­‐oriented decision theories like Decision Field Theory. Author(s): Peter Hammond; Stefan Traub Financial agency: The role of ex-­‐ante and ex-­‐post accountability Stefan Trautmann Tilburg University ABSTRACT: We implement an experiment in which one subject, the agent, makes risky investment decisions on behalf of another subject, the principal. We find that agents take more risk in the case they invest the principal’s money than in the case they invest their own money. This effect is reduced significantly when the agent can be held accountable by the principal, which we implement by allowing the principal to reward the agent for the investments made. We distinguish between rewards that the principal can base on the realizations of the investments (ex post), and those that she needs to specify on the basis of the level of the investments only (ex ante). We find differences between the ex-­‐ante and ex-­‐post accountability in terms of the investment incentives and beliefs of the agents, and in terms of the rewards they elicit. We discuss implications for financial agency. Author(s): Potters; Jan; Pollmann; Monique; Trautmann; Stefan Is reaction time predictive of choice? An experimental study of threshold strategies Isabel Trevino New York University ABSTRACT: We perform a behavioral estimation of threshold strategies in a 2x2 global coordination game by analyzing the response times associated to signals and actions. In a global game subjects observe noisy private signals about a state variable that, together with the action of the other player, determines payoffs. An optimal threshold is defined as the signal for which a subject is indifferent between taking any one of the two actions. We estimate ‘choice based’ thresholds by using standard methods with the observed choices in the last rounds of the experiment, once behavior is stable. When analyzing response times, we look at the initial rounds when subjects are building a strategy and find that the amount of time spent in each signal is significantly decreasing on the absolute value difference between the estimated threshold for that subject and the value of the signal he is observing. This result suggests that when subjects face signals that are closer to the threshold they will converge to, they take more time to reach a decision than when observing signals that are far away from their threshold. Moreover, we find that the signal associated to the highest or second highest response time is a good predictor of the threshold that subjects exhibit in later periods. Author(s): Andrew Schotter; Isabel Trevino Flexible valuations for consumer goods as measured by the Becker-­‐DeGroot-­‐Marschak mechanism Agnieszka Tymula New York University ABSTRACT: This paper experimentally investigates whether valuations elicited by the commonly used Becker-­‐DeGroot-­‐Marschak (BDM) procedure depend on the distribution of prices that are used in the elicitation mechanism. To answer this question we created a novel within-­‐subject design that allowed us to observe an individual’s bid for a given product repeatedly while varying the price distribution. Our data clearly show that subjects do not bid constantly the same amount for each good on each offer as would be predicted by EU theory, but rather show a mass-­‐
seeking bias. This bias is strongest when the mass of the price distribution is close to the average bid that the subject places on the good. Bids are influenced by the mass to a lesser extent when the mass of the distribution is further away from the mean bid. We characterize preference structures that are consistent with the observed behavior. Author(s): Agnieszka Tymula; Eva Woelbert; Paul Glimcher Commitment aversion Aljaz Ule CREED, University of Amsterdam ABSTRACT: Many economic decisions involve committing to a choice for many future periods. A simple example is a large infrastructure investment. There are various reasons why decision makers may shy away from such commitment even if it is profitable in expected terms. Possible reasons include risk aversion and non-­‐rational discounting, for example. In this paper, we present results from an experiment that provides evidence for another reason why decision makers choose not to commit. This is an innate aversion to committing per se. In our experiment, subjects repeatedly choose between (similar) lotteries. Treatment vary in appointing one of the lotteries as a ‘commitment choice’. After choosing this commitment choice, subjects must choose it for the remaining periods of the experiment. We show that, for some sets of lotteries, designating a commitment option substantially changes subjects’ choices in a way that can only be attributed to an aversion to committing. We also show that this aversion is limited to choices that appear similar. If the costs to committing are high enough, the commitment aversion disappears. Author(s): Arthur Schram; Aljaz Ule Information Acquisition under (Im)perfect Data Privacy Verena Utikal University of Erlangen-­‐Nurnberg ABSTRACT: We investigate the consequences of imperfect data privacy on information acquisition about personal health data. In a simplified game of persuasion players decide on whether or not to acquire information about their health type before searching a match in a market. We contrast three institutional settings: automatic dissemination of certified test results, voluntary dissemination of certified test results and imperfect data privacy about certified test results (i.e. potential involuntary dissemination). We find that the basic intuition, that a higher probability of data loss makes information acquisition less likely, holds only for mixed strategy equilibria. Imperfect data privacy does however not change pure strategy equilibria. We then test the model’s prediction in a laboratory experiment. The behavioral results show that imperfect data privacy does indeed not affect information acquisition. However, participants seem to take imperfect data privacy into account when deciding on a match with an unknown type. Author(s): Simeon Schudy; Verena Utikal Belief Elicitation: A Horse Race among Truth Serums Gijs van de Kullen Tilburg University ABSTRACT: In survey studies, probabilistic expectations about uncertain events are typically elicited by asking respondents for their introspective beliefs. If more complex procedures are feasible, beliefs can be elicited by incentive compatible revealed preference mechanisms (‘truth serums’). Various mechanisms have been proposed in the literature, which differ in the degree to which they account for respondent’s deviations from expected value maximization. In this paper, we pit non-­‐incentivized introspection against five truth serums, to elicit beliefs in a simple two-­‐player game. We test the internal validity (additivity and predictive power for own behavior), and the external validity (predictive power for other player’s behavior, or accuracy) of each method. We find no differences among the truth serums. Beliefs from incentivized methods are better predictors of subject’s own behavior compared to introspection. However, introspection performs equally well as the truth serums in terms of accuracy and additivity. Author(s): S. T. Trautmann; G. van de Kuilen Buying lies: The effects of interaction and context on the ability to detect deceit Jeroen van de Ven university of amsterdam ABSTRACT: We study deception and the ability to detect deceit in a market with asymmetric information. In the experiment, buyers and sellers meet each other in successive brief face-­‐to-­‐face interactions. Sellers are privately informed about which one of two products is most advantageous for the buyer. The sellers make recommendations to the buyers, and have the option to deceive buyers when their incentives are not aligned. We systematically vary two characteristics of the market. First, we vary whether or not the buyer can interrogate the seller. Second, we change the context by varying the complexity of the product. In some treatments the two products are simply labeled as black or red, while in other treatments the products are descriptions of holiday packages showing a range of features. We find that sellers almost always make recommendations that are in their own best interest. Buyers are not able to detect deceit when they have no option to interrogate the seller, but are substantially more likely than chance to detect deceit when the interaction is longer, both when the product is simple and complex. Author(s): Michele Belot; Jeroen van de Ven Punishment and leadership in a voluntary contribution experiment Eline van der Heijden Tilburg University ABSTRACT: A number of studies have shown that the opportunity to punish as well as leading by example increase cooperation levels in a voluntary contribution experiment. In this paper we consider both institutions as a form of leadership, and we examine the separate and the combined effect of the two mechanisms. In a first treatment, leadership is just implemented as leading by example by letting one group member, the leader, contribute before the other group members do. In a second treatment, all group members decide simultaneously. Here, (only) one group member has the opportunity to punish the others group members after he has observed the contributions by all group members. In a third treatment both forms are leadership are put into practice. That is, the leader not only moves first and can thus lead by example, but he can also punish the group members. We find that in all treatments contribution levels are substantially and mostly significantly higher than in a control treatment with a standard, simultaneous, voluntary contribution mechanism and no punishment. However, contributions in the treatment where leaders can lead by example and punish are not significantly different from contributions in the treatments in which only one form of leadership is available. This suggests that different forms of leadership are substitutes rather than complements. Author(s): Eline van der Heijden Stars Need Benefits: An Experiment on Network Formation Boris van Leeuwen University of Amsterdam / CREED ABSTRACT: We experimentally investigate network formation in an environment where players simultaneously decide on their network connections and contributions to a local public good. Our design is based on the theoretical framework developed by Galeotti and Goyal (2010). The main result in their study is that in every strict equilibrium of the game, there is a limited number of players who contribute to the public good. These players form the core of the network. Other (periphery) players link to the core, and contribute nothing themselves. In our 2x2 design, we vary the group and distinguish between an environment where there are benefits from an incoming link and one without such benefits. Assuming self-­‐regarding preferences, the theoretical (equilibrium) predictions are the same across the four treatments. Our experimental results show substantial treatment effects, however. In the situation without benefits, equilibrium networks are rarely observed. However, when we introduce these benefits almost all groups converge to an equilibrium network architecture. We attribute this difference across treatments to the relative payoff differences between the core and periphery players. Benefits reverse the relative payoff difference in favor of the core players. Secondly, we find that doubling the group size has no effect on the observed network architecture but does increase public good contributions to levels close to efficiency when there are benefits for an incoming link. Author(s): Boris van Leeuwen; Theo Offerman; Arthur Schram Peer Effects without Reflection Roel van Veldhuizen University of Amsterdam ABSTRACT: Previous research has shown that peer effects are important in many areas, including education, consumer choice and the workplace. However the existing literature has made little attempt to disentangle possible mechanisms behind these peer effects. In this study, we differentiate between social pressure and social preferences using a laboratory experiment in which participants participate in a real effort team production task. During the production task, participants can either see the results of other participants or know that their results can be seen by other participants or both. Moreover, we investigate the conditions under which peer effects are more likely to occur by notifying some participants about the performance of their team members on a related baseline task. Author(s): Hessel Oosterbeek; Joep Sonnemans; Roel van Veldhuizen Hypothetical Thinking and Information Extraction: Strategic Voting in the Laboratory Emanuel Vespa New York University ABSTRACT: We test for strategic behavior in common-­‐value elections by using an experimental design that distinguishes the mistake of failing to account for the information content of others’ strategies from other mistakes that arise naturally when making inferences. Depending on the treatment, between 50-­‐80% of subjects fail to vote strategically. This mistake is robust to experience and hints about pivotality, and it is mainly driven by the failure to make inferences from hypothetical, rather than actual, events. Finally, the mistake also arises in more general settings where players have no private information. Author(s): Ignacio Esponda; Emanuel Vespa Breaking the glass ceiling with 3no2: Gender differences in doing favors Lise Vesterlund University of Pittsburgh ABSTRACT: We ask if men and women differ in the propensity by which they do favors for others. Using a laboratory study we develop an environment where participants are presented with an implicit request to do a favor. We find systematic and persistent gender differences, as women are 50 percent more likely to do the favor than men. This difference is supported by evidence from a survey of MBAs. When asked to think of a work-­‐related favor that they would rather have declined, women were far more likely to report that they in agreeing to do the favor felt pressured, found it difficult to say no, and that they were worried about the consequences of declining. Women were also more likely than men to say that they agree to do too many things at work. The finding that women find it more difficult to decline a request for a favor is particularly concerning in light of experimental evidence that women also are more likely to be asked for favors. Employees who are more frequently asked and agree to do favors are more likely to become over committed with low skilled tasks, and this may prevent them from undertaking tasks that can lead to promotions. We argue that these gender differences in doing favors can help explain the persistent vertical gender segregation in the labor market. Author(s): Linda Babcock, Brenda Peyser, MJ Tocci, Lise Vesterlund, Amanda Weirup; Laurie Weingart Public-­‐Good Games and the Balinese Robert Veszteg Waseda University ABSTRACT: The Balinese have been successful for centuries in sustaining cooperation among the members of local communities in order to provide public goods through individual contributions. We survey the experimental literature on public goods and highlight those features of the Balinese tradition that have been proven to be both effective in the experimental laboratory and successful in deterring free-­‐riding on the field. The most prominent features discussed are decentralization, democratic decision-­‐making, the use of two currencies, supervision, and the possibility of imposing severe sanctions for free-­‐riding. Our results not only can help to preserve the high level of cooperation among inhabitants in Bali threatened by migration flows and the increasingly intense reliance on the market mechanism, but they also provide general insights both for theoreticians and practitioners on how to create successful communities. In addition, the literature review sheds light on several features of public-­‐good games that have not been satisfactorily explored yet by experimental economists. The novelty of our approach lies in looking at the Balinese tradition through the glasses of mechanism design theory and aligning the related findings of experimental economics in order to understand its success and problems. Author(s): Robert Veszteg; Erita Narhetali To remit, or not to remit: that is the question. A remittance field experiment Angelino Viceisza International Food Policy Research Institute ABSTRACT: We conduct a remittance field experiment among Salvadoran migrants in the metro DC area. Migrants need to decide whether or not to remit funds to a recipient in El Salvador and if so how much. We maintain a (2x2) design in which the remittance budget has a value of $400 or $200, and the remitted funds arrive as cash or grocery vouchers that are non-­‐transferable and applicable to basic necessities that do not include alcohol and cigarettes. Each migrant is randomly allocated to one of the resulting four treatments. We test across these treatments whether control over remittance spending in the form of grocery vouchers affects remittance behavior. We find the following. (1) Migrants send similar proportions across groceries and cash, regardless of the budget size. Qualitative reports suggest that migrants have a preference for a flexible form of control's that can be used for other expenditures in addition to groceries. (2) These effects are not robust to disaggregation by gender. (a) Male migrants have a preference for cash when the remittance budget is $400, but not when it is $200. (b) Female migrants have a similar preference for cash and groceries, regardless of the remittance budget. (3) We develop a conceptual framework that could explain these findings and use survey data to test its implications. The effects are consistent with this framework where migrants exhibit altruism that is conditional on grocery spending, but only up to a certain point. In particular, the gender differences are consistent with previous literature. Author(s): Maximo Torero; Angelino Viceisza Stake Effects On Ambiguity Attitudes For Gains And Losses Ferdinand Vieider University of Munich ABSTRACT: We test the effects of stake size on ambiguity attitudes. Compared to a baseline condition, we find subjects to be more ambiguity seeking for small-­‐probability gains and large-­‐probability losses under high stakes. They are also more ambiguity averse for large-­‐probability gains and small-­‐probability losses. We trace these effects back to stake effects on decisions under risk (known probabilities) and uncertainty (unknown probabilities). For risk we replicate previous findings. For uncertainty, we find an increase in probabilistic insensitivity under high stakes that is driven by increased uncertainty aversion for large-­‐probability gains and for small-­‐probability losses. At the individual level, we find inter alia that subjects display more pessimism for small-­‐
probability losses than optimism for small-­‐probability gains under uncertainty, and that they display more pessimism for large-­‐probability gains than optimism for large-­‐probability losses. These findings, in turn, may have consequences for financial decision making and insurance uptake, which we discuss. Author(s): Vieider; Ferdinand M.; Peter Martinsson; Haileselassie Medhin Moral Hypocrisy, Power, and Social Preferences Marie-­‐Claire Villeval GATE(CNRS-­‐University Lumiere Lyon 2-­‐ ENS) and IZA, Bonn ABSTRACT: We show with a laboratory experiment that individuals adjust their moral principles to their actions and to their situation just as much as they adjust their actions to their principles. We first elicit the individual’s principles regarding the fairness and unfairness of allocations in three different scenarios (a Dictator game, an Ultimatum game, and a Trust game). One week later, the same individuals are invited to play those same games for real. Finally in the same session we elicit again their principles regarding the fairness and unfairness of allocations in the same three scenarios. Our results show that individuals adjust abstract norms to fit the game, their role and the choices they made. First, norms that appear abstract and universal take into account the bargaining power of the two sides. The strong side bends the norm in its favor and the weak side agrees: Stated fairness is a compromise with power. Second, in most situations, individuals adjust the range of fair shares after playing the game for real money compared with their initial statement. Third, the discrepancy between hypothetical and real behavior is larger in games where real choices has no strategic consequence (Dictator game and second mover in Trust game) than those where they do (Ultimatum game). Finally, the adjustment of principles to actions is mainly the fact of individuals who behave more selfishly and who have a stronger bargaining power. The moral hypocrisy displayed (measured by the discrepancy between statements and actions chosen followed by an adjustment of principles to actions) appears produced by the attempt, not necessarily conscious, to strike a balance between self-­‐image and immediate convenience. Author(s): Aldo Rustichini; Marie Claire Villeval Rule Following and Reciprocity in Common Pool Resource Games Alexander Vostroknutov Maastricht University ABSTRACT: We design an experiment to explore the relationship between individual willingness to follow costly rules and cooperation in common pool resource dilemmas. We argue that costly rules function as screening mechanisms by which an observer can identify cooperative types. Subjects are instructed to follow an arbitrary rule at some personal cost, and we group them without their knowledge based on how much cost they incur in service of the rule. These groups then participate in a finitely repeated, dynamic common pool resource (CPR) dilemma. In our CPR game, four players choose ten times how much of a common pool resource to harvest. After each harvesting decision, the resource regenerates in proportion to the remaining stock unless it falls below some critical level. For a sufficiently high growth rate, the CPR game has a unique SPNE in which the resource can be maintained without exhaustion for any number of periods. For lower growth rates, the SPNE generates rapid resource exhaustion. We compare subject behavior under two growth rates, one in which the SPNE maintains the resource and the other in which it is exhausted. We find that rule-­‐following individuals are less likely to exhaust the resource than rule-­‐breaking individuals, and we show that the difference results from increased negative reciprocity among rule-­‐breakers who observe selfish behavior by others. Author(s): Erik Kimbrough; Alexander Vostroknutov Political motivations and electoral competition: Equilibrium analysis and experimental evidence Nicolaas Vriend Queen Mary, University of London ABSTRACT: We study both theoretically and experimentally the set of Nash equilibria of a classical one-­‐
dimensional, majority rule election game with two candidates. These candidates are interested in power and ideology, but their weights on these two motives are not assumed to be the same, i.e., their electoral motivations are not necessarily identical. Apart from obtaining the well known median voter result and the two-­‐sided policy differentiation outcome, the paper uncovers the existence of two new equilibrium configurations, called one-­‐sided's and probabilistic's policy differentiation, respectively. Our analysis shows how these equilibrium configurations depend on the relative interests in power (resp., ideology) and the uncertainty about voter’s preferences. The theoretical predictions are supported by the data collected from a laboratory experiment, as we observe convergence to the Nash equilibrium values at the aggregate as well as the individual levels in all treatments, even in the probabilistic differentiation treatment with a unique mixed strategy equilibrium. The comparative statics effects across treatments (varying the interests in power and ideology as well as the uncertainty about voter’s preferences) are as predicted by the theory. We also find that subject’s learning takes place mainly within the first ten periods (elections), and that most of that learning does not vanish as subjects interchange their roles between candidates of different ideologies. JEL Classification: C72, C90, D72. Keywords: Electoral competition, Power, Ideology, Uncertainty, Nash equilibrium, Experimental evidence. Author(s): Michalis Drouvelis; Alejandro Saporiti; Nicolaas J. Vriend Parental altruism: Evidence from field experiments in Tanzania Jana Vyrastekova Radboud University Nijmegen ABSTRACT: The parental altruism assumption stands central in theoretical papers on population, health and development economics. The gendered version of parental altruism is also invoked, assuming differences in preferences of fathers and mothers, usually conjecturing that mothers are more altruistic towards their children than fathers. Nevertheless, empirical evidence for or against any of these crucial assumptions is largely non-­‐existent, or based on crude macroeconomic data. Although some experimental evidence addresses gender differences in altruism, the parental altruism has not yet been studied in developing countries. We fill this gap by collecting a unique dataset of parental choices in field experiments performed in Tanzania. At eight locations, varying in their extent of urbanization, as measured by the distance to the closest market, we interview female and male participants on their attitudes to family planning methods. As a reward, the interviewed individuals are offered a choice between a good that is only suitable as a consumption good for children (children size slippers), and good(s) that can be directly consumed by the adult decision-­‐maker (cash and/or sugar). Linking the choices from our experiments to a detailed background characteristics of the experiment participants allows us to test the parental altruism hypothesis and address its gendered variant. We find no trivial gender effect, and discuss the urbanization impact on parental altruism. Next to theoretical relevance, understanding parental altruism and its gender (in)dependence is central for implementing effective family planning policies, and for interventions targeting education and population health in developing countries. Author(s): Janine Huisman; Idda Moscha; Jeroen Smits; Jana Vyrastekova Strategies and Beliefs in a Location Game Stephanie Wang University of Pittsburgh ABSTRACT: In the Lemonade Stand Game, three participants must simultaneously choose to locate on one of twelve spots around a circle. Their payoffs are equal to the distance (measured in spots) of the clockwise participant plus the counterclockwise participant. If two of the three participants choose the same location, they both get 6 while the third participant gets 12. If all three participants choose the same location, then they all get 8. The sum of the three participants’ payoffs is always 24. Participants will make location choices in this game for a number periods. Before they make each decision, they will be incentivized to state their beliefs about the location choices of everyone in their group. We will estimate best-­‐fitting strategy for each participant out of the list of algorithms submitted in the Lemonade Stand Game tournament. One salient example is cooperating with one player of the group to exploit the third player. Furthermore, we test if the players can be categorized into different reasoning levels of the level-­‐k/Cognitive Hierarchy models. Finally, we estimate the parameters of a general learning model to characterize the changes in behavior over repeated rounds of this game. Author(s): David Klinowski; Stephanie W. Wang Neural Random Utility Ryan Webb New York University ABSTRACT: A central premise of neuroeconomic theory is that there exists an observable object in the human brain, termed subjective value, which is the neural correlate of utility. Unlike utility, however, subjective value is cardinal, maps to a unique point on the real number line, can exist under non-­‐choice circumstances, and is causally related to choice. We develop the basic structure of a new class of neuroeconomic model, the neural random-­‐utility model, which operationalizes each of these features in a choice model, and we demonstrate the serviceability of this model by making direct measurements of subjective value and relating these measurements to choice prediction. We analyze data from an experiment in which 12 subjects underwent fMRI scans while they were shown pictures of 20 desirable items such as DVD's, books, and CD's. After the scanning session, subjects were then asked to perform a second experiment in which they made a series of choices over pairs of the items. Our results suggest that the magnitude of measured neural activity predicts the likelihood that an item will be chosen and can be well-­‐represented by our neural random-­‐utility model when we account for the degree of measurement error present in fMRI imaging. Author(s): Ryan Webb; Ifat Levy; Stephanie Lazzaro; Robb Rutledge; Paul Glimcher The Non-­‐Equivalence of Labor Market Taxes: An Experiment Matthias Weber University of Amsterdam, CREED, Tinbergen Institute ABSTRACT: Under full rationality a labor income tax and a corresponding employer’s contribution are equivalent. Without full rationality this equivalence is no longer obvious, however. If people react differently to the two this has direct impact on policy making and consequences for optimal taxation theory and political economics. This paper examines how people react to these duties in a real-­‐effort laboratory experiment. We study their differential effects on labor supply, subjective well-­‐being, exerted effort and preferences concerning the size of the public sector. To measure the latter in the laboratory we develop a novel approach to elicit these preferences in an incentive compatible manner. Our results show that people react differently to the distinct duties. Our findings suggest that labor supply is higher under an income tax while subjective well-­‐being is higher under an employer’s contribution. Our findings also suggest that an employer’s contribution induces preferences for a larger public sector. Author(s): Matthias Weber; Arthur Schram Restoring Fairness in Hierarchies Roberto Weber University of Zurich ABSTRACT: There is much evidence in the economics literature that leadership can facilitate improved group coordination and performance in a variety of ways. We add to this literature by examining the effect of leadership when managers and their employees have imperfectly aligned incentives. Managers often have better information about the quality of a project than their employees. They can use their superior information to generate buy-­‐in, but must maintain credibility in the face of strong incentives to misrepresent their information. We study an experimental environment designed to capture this tradeoff between short-­‐term gains from deception and long-­‐term gains from credibility. Managers use their noisy information to advice investment or abstention to their group members who have no information of their own. Managers have a binary payoff structure that rewards full group ‘buy-­‐in’ in some states and no buy-­‐in in others. Group members’ payoffs include a complementarity in that they depend on both the state of the world and the number of other group members choosing to invest. We find that improving the quality of the manager’s information, the fineness of messages, or the ex-­‐post ability to verify the manager’s statement all improve performance. Managers frequently send deceptive messages. Giving bad advice does not appear to affect an manager’s credibility in the short run, but has a significant negative effect in the long run. Author(s): David Cooper; John Hamman; Roberto Weber Experiments on Public Debt Joachim Weimann Otto-­‐von-­‐Guericke Universit?t ABSTRACT: All over the world, we observe countries which are heavily indebted. However, experimental studies on public debt crisis and the analysis of mechanisms that can help in escaping from the curse of growing national debts are still pending. In this paper, we first analyze the driving forces of the inherent tendency of states to borrow more and more. We conjecture that it is the fact that public debt allows for the shifting of burdens on future generations that causes this behavior. Secondly, we analyze different debt rules. In our experiment, subjects decide on both the size and the funding of a public good in an economy. As we allow lower group contributions than public good sizes, economies can run into public debt. In the baseline treatment, one generation ‘lives’ for a known number of rounds. If the debt becomes higher than a (not perfectly known) exogenous given value, the economy has to spend all their income to reduce the debt. In a second treatment, we introduce a number of generations, following one after the other. Thus each generation is able to shift the burden of the debt to the following generation. To test sustainability, we use three mechanisms in both treatments: (1) no special debt rule, (2) a deficit cap, and (3) a penalty on too high public debt levels. Author(s): Martin Fochmann; Abdolkarim Sadrieh; Joachim Weimann Price competition between teams with heterogeneous incentive schemes Ori Weisel Hebrew University ABSTRACT: We experimentally study a Bertrand duopoly market where each competitor is modeled as a three-­‐player team. Each player states a bid and the team with the lower total bid wins the competition. There are two incentive structures by which profits are divided among members of the winning team: each member receives (1) her own asking price or (2) the average asking price in the team. In different experimental conditions the market is either homogeneous (both teams have the same incentive structure) or heterogeneous, and information about the incentive structure of the opposing team is either public or private. Our results show that heterogeneity notably reduces the ability of market participants to tacitly collude, contradicting predictions derived from game theory and reinforcement learning, but in line with the legal intuition of the current EU horizontal merger guidelines. Additionally, we find that when information about the incentive structure of the opposing team is public (rather than private) collusion is more likely in homogeneous markets, but less likely in heterogeneous markets. Author(s): Michael Kurschilgen; Alexander Morell; Ori Weisel Learning to Use Private Information: Reexamining the Asymmetric-­‐Information Game in the Laboratory Yuanji Wen Bocconi University ABSTRACT: This paper examines the two-­‐stage asymmetric information game studied by Feltovich (1997, 1999 and 2000) and compares it to a modified game in the laboratory. By tracking subject’s information acquisition, we find indirect evidence in support of a belief-­‐based learning model. The maximum-­‐likelihood estimations of subject’s learning behaviors using the EWA model (Camerer and Ho 1999) show that two informational roles learn at different speeds and the informed group is more active in learning than its counterpart. Besides, the hybrid feature of EWA estimations suggests that the reduced belief-­‐based model characterizes player’s behaviors better than the reduced reinforcement model but slightly underperforms in prediction. Finally, our cluster and individual estimations combined with information acquisition data show that the majority of seemingly fictitious plays coincides with the dominance of a belief-­‐based updating pattern defined by information acquisition measures. Tracking subject’s information acquisition can be a complementary tool to study learning. Author(s): Yuanji Wen Endowment heterogeneity, identity, punishment and cooperation: Evidence from a public good experiment Qian Weng University of Gothenburg ABSTRACT: This paper explores how three possible mechanisms to deter free-­‐riding and foster cooperation function interactively in a team production setting. In a repeated public good experiment, subjects are entitled with either homogenous or heterogeneous endowment, either strong or weak identity, and either existence or nonexistence of decentralized punishment opportunity. At the team level, we observe that when punishment is not possible strong identity can counteract the negative impact of endowment heterogeneity on cooperation and on earnings, whereas it has no effect on the homogenous teams. The introduction of punishment successfully raises cooperation in all endowment distribution and identity strength combinations. With punishment strong identity significantly enhances cooperation in homogenous but not in heterogeneous teams. Punishment occurs more vehemently in homogenous than in heterogeneous teams under both identity strengths. Within the heterogeneous teams, we find that lowest endowment individuals always show the greatest degree of cooperation, although the gap between them and the higher endowment individuals narrows with the introduction of punishment. Strong identity increases cooperation in all endowment levels without punishment, but not so with punishment. The punishment inflicted in general does not differ significantly between individuals with lowest and with higher endowment, and it is also unaffected by the strength of identity. The overall inequality in ultimate earnings is reduced through repeated interactions since lower endowment individuals receive greater net gains in all treatments and in particularly so in the treatments with punishment. Author(s): Qian Weng Are efficiency wages equality wages? Exogenously induced fairness norms in working environments Peter Werner University of Cologne ABSTRACT: A number of studies have demonstrated the importance of fairness concerns and relative pay for employee motivation both in the lab and in the field, corroborating the influential fair wage-­‐
effort hypothesis by Akerlof and Yellen (1990). According to this notion, an employee withdraws effort if her actual wage falls short of a ‘fair’ level. However, a concise definition of the term ‘fair wage’ is hardly possible. While wage equality is one potential reference point for fairness considerations, there are often alternative fairness norms that conflict with equal pay. We investigate how subjective entitlements influence wage offers and effort exertion in a gift exchange experiment. Our design allows us to examine the effect of competing fairness norms: if claims arising from a priming task conducted before the actual decision situation alter fairness perceptions of the employees, equality should not be the predominant norm for wage offers, and wage schemes that differentiate between employees should be more acceptable. Indeed, we find that the competitive assignment of advantage leads to heterogeneous entitlements among experimental employees. A norm of differentiation clearly dominates a norm for equal pay, resulting in substantial wage discrimination. At the same time, we do not find a detrimental impact of disadvantageous relative positions, as effort levels of high-­‐wage and low-­‐wage subjects are similar despite the realized wage differences. Finally, communication significantly influences effort exertion. Author(s): Gary Bolton; Peter Werner Are People Risk-­‐Vulnerable? Marc Willinger Faculty de Sciences Economiques, LAMETA, university de Montpellier 1 ABSTRACT: We report on a within-­‐subject experiment, with substantial monetary incentives, designed to test whether or not people are risk-­‐vulnerable. In the experiment, subjects face a simple portfolio choice problem in which they have to allocate part of their wealth between a safe and a risky asset. We elicit risk vulnerability by observing each subject’s portfolio choice in two different contexts that only differ by the presence or absence of an actuarially neutral background risk. Our main result is that most subjects -­‐-­‐ above 80% -­‐-­‐ are risk-­‐vulnerable, i.e. choose a less risky portfolio when exposed to background risk. Contrasting our experimental data with predictions provided by alternative decision-­‐theoretic models (EU, CPT, RDU), leads us to the conclusion that EU theory is closest to our data . Author(s): Mickael Beaud; Marc Willinger Communication With Multiple Senders and Multiple Dimensions: An Experiment Alistair Wilson University of Pittsburgh ABSTRACT: We implement the Battaglini (2002) model of multi-­‐sender-­‐multi-­‐dimension cheap talk in the laboratory, analyzing the effects of sender competition on information transmission. Our results indicate that competing senders provide enough information for close to full revelation, but receiver’s ability to use this information crucially depends on senders’ biases. Receivers are close to full extraction when biases identify an ex-­‐ante trustworthy sender. When there is no ex-­‐
ante trustworthy source, fully exploiting sent messages requires information to be inferred across dimensions. However, receivers seem to treat dimensions as separate choices, and are much closer to best-­‐responding within each separate dimension. Author(s): Emanuel Vespa Selling Prices of 'sSelf-­‐Made's Ambiguous Assets Frederik Witte EM LYON Business School ABSTRACT: Ambiguous assets are characterized by uncertain probabilities (Sarin and Weber 1993). Individual-­‐level experiments show that selling prices for ambiguous assets are sensitive to the center and range of probability intervals (Curley and Yates 1985, Einhorn and Hogarth 1985, Kahn and Sarin 1988). Experimenters typically ask participants to choose between an ambiguous asset — that is given — and a certainty equivalent (Eisenberger and Weber 1995). What happens when we ask them to choose between an ambiguous asset — they created — and a certainty equivalent? The choice is common in business. Consider an entrepreneur choosing between developing his venture or selling it. I present an experiment designed to observe participant’s certainty equivalent for an ambiguous asset they generated by real effort. The objective was to operationalize ambiguity without specifying the center and range of the probability interval. 73 participants were randomly assigned to two conditions. In the treatment condition, participants performed an idea-­‐generation task. It’s payoff was the ambiguous asset. In the baseline condition, participants simply chose an ambiguous asset from a list. All participants knew the rules of payment, but not the value of the ambiguous asset. They revealed their certainty equivalent for the ambiguous asset after completing a real-­‐effort task. Two hypotheses were confirmed. Asset prices are larger in the treatment condition, than in the baseline condition (H1). Task confidence positively affects asset prices (H2). Overall, participants demand larger certainty equivalents when they created the ambiguous asset. Author(s): Frederik C. Witte The effect of effectiveness: Donor response to aid effectiveness in a direct mail fundraising experiment Daniel Wood Clemson University ABSTRACT: We test how donors respond to new information about a charity’s effectiveness. Using two waves of direct marketing mail solicitations to prior donors, Freedom from Hunger (FFH) tested whether donations increased in response to the provision of information about their impact as measured by scientific research. The base script, used for both treatment and control, included a more standard identifiable victim appeal, highlighting a qualitative story of a particular beneficiary of FFH services. We find no effect on the full sample of prior donors from adding information on the scientific impact research. However, we find an important heterogeneity: large prior donors and more recent prior donors respond to the information on aid effectiveness by increasing their giving, whereas small prior donors and less recent prior donors respond with decreased giving. These responses are consistent with differing motives, where large donations are altruistic while small donations provide ‘warm-­‐glow’ utility. Author(s): Dean Karlan; Daniel H. Wood Justification and Cooperation Erte Xiao Carnegie Mellon University ABSTRACT: The need for justification is a widely observed social phenomenon. This paper develops a theoretical framework and reports laboratory evidence to show how pure justification pressure affects cooperative behavior in economic exchange environments. In a one-­‐shot anonymous interaction, compared with the case when the behavior is simply observed by the audience, individuals are more likely to act on what they believe the audience thinks they should do when they also have to explain the decisions to the audience. When it is salient that the audience thinks one should cooperate, justification pressure significantly promotes cooperation even absent negative consequences for non-­‐cooperative behavior. We discuss the implications of our findings for shaping institutional design to promote cooperation. Author(s): Erte Xiao With or without siblings: sorting by information feedback mechanism in experimental Chinese labor market Fanzheng Yang Department of Economics, Iowa State University ABSTRACT: In China, one child policy was in effect in the late 1970s. Only child and child with siblings grow up in quite different social and economic environments, so they have considerably different personalities. Some argue that only children might have lower tendency in competition because they are more vulnerable than children with siblings. However, others argue that only child also bares more pressure from parents, and this may reinforce only child’s attitudes toward competition. Therefore, we study how differently only child and child with siblings make the decision of choosing competition in an experimental labor market. We carried out the lab experiments at Central University of Finance and Economics in Beijing, in March and December 2011. We distinguish participant’s behaviors by designing a two dimensional incentive structure: pay by piece rate or tournament and with or without cohort’s relative performance feedback. Our findings show that only child is more reluctant to compete than child with siblings. Self-­‐
performance evaluation reveals that only child is less confident than child with siblings. Without knowing the relative performance of others, only child with higher ability is not significantly more likely to choose tournament than the counterpart. Furthermore, when receiving feedback about their relative performance, only child, now presumably to be more likely to choose tournament, does not significantly shift their interest toward more competition, which is in contrast with those overconfident children with siblings who wisely avoid tournament when they know their inferiority. Author(s): Fanzheng Yang; Li Yu Does Gradualism Build Trust? Evidence from A Multi-­‐round Trust Game Maoliang Ye Harvard University ABSTRACT: This paper examines the effect of gradualism -­‐-­‐ increasing the stake of investment slowly over time rather than requiring a high stake of investment immediately -­‐-­‐ in trust building using a multi-­‐round trust (investment) experiment. I randomly assign participants to three treatments: starting and continuing at a high stake (the ‘Big Bang’ treatment), starting at a low stake but jumping to a high stake after a few rounds (the ‘Semi-­‐gradualism’ treatment), and starting at a low stake and gradually increasing the stake over time (the ‘Gradualism’ treatment). In each round, two players of each group make their decisions sequentially. The first (‘trustor’) makes a binary decision whether to pass the stake to the second one (‘trustee’), which measures her trust. If she passes, the stake triples and the trustee then makes a binary decision: to return a low unfair amount or a high fair amount to the trustor, which measures her trustworthiness. I find that at the beginning the degree of trust is the same for all treatments, but trustworthiness is lower for the ‘High Start’ treatment; as a result, the ‘High Start’ treatment has a lower level of subsequent trust and a lower rate of successful cooperation (here cooperation is defined as ‘successful’ if and only if both of the following conditions hold: the trustor is trusting and the trustee is trustworthy.). However, trustworthiness and trust for all treatments sharply decrease in the end (‘end-­‐of-­‐game’ effect). Author(s): Maoliang Ye Gender Discrimination and Social Identity: Experimental Evidence from Urban Pakistan Basit Zafar Federal Reserve Bank of New York ABSTRACT: Gender discrimination in South Asia is a well-­‐documented fact. However, gender is only one of an individual’s many identities. This paper investigates how gender discrimination depends on the social identities of interacting parties. We use an experimental approach to identify gender discrimination by randomly matching 2,836 male and female students pursuing bachelors-­‐
equivalent degrees in three different types of institutions. “Madrassas (religious seminaries), Islamic Universities, and Liberal Universities” that represent distinct identities within the Pakistani society. We find that gender discrimination is not uniform in intensity and nature across the educated Pakistani society and varies as a function of the social identity of both individuals who interact. Even within these elite groups, we find that women “but only low-­‐
status women” are discriminated against. While we don't find evidence of high-­‐status men discriminating against women, men of lower socio-­‐economic status and higher religiosity tend to discriminate against women, particularly those who are closest to them in social distance. Moreover, this discrimination is entirely taste-­‐based. Our findings suggest that social policies aimed at empowering women need to account for the intersectionality of gender with social identity. Author(s): Adeline Delavande; Basit Zafar Using Clock Auction to Reallocate Round-­‐trip Routes: An Experimental Study Fangzhou Zhang Shanghai Jiaotong University ABSTRACT: This paper studies a clock auction mechanism to reallocate four round-­‐trip flights that are perfect complements between two airlines who own different shares (2 versus 2 or 1 versus 3). The airline which first activates the bidding button sells his share in the ascending clock auction or purchases the other airline’s share in the descending clock auction. We conduct a series of lab experiments and test the result against the predicted Nash equilibrium. We find that the equilibrium price is consistent with prediction in ascending clock auction but lower in descending clock auction. We also observe dramatic deviation from equilibrium in relatively ‘unlucky’ airline whose private valuation is quite low. To explain the inconsistency of the bidding behavior against the theory, we elicit subject’s risk preference by a lottery game and calculate his equilibrium bidding function based on his risk attitudes, rather than assuming risk neutrality. Forgone expected payoff is estimated to measure the opportunity cost of deviation. Subject’s bidding behavior and the social efficiency are further compared with first-­‐price and second-­‐price sealed-­‐bid auctions. Author(s): Fangzhou Zhang Test the principle of maximum entropy in constant sum 2x2 game: Evidence in experimental economics Wang Zhijian Experimental Social Science Laboratory, Zhejiang University ABSTRACT: By using laboratory experimental data, we test the uncertainty of strategy type in various competing environments with two-­‐person constant sum 2x2 game in the social system. It firstly shows that, in these competing game environments, the outcome of human’s decision-­‐making obeys the principle of the maximum entropy. Author(s): Bin Xu; Hongen Zhang; Zhijian Wang; Jianbo Zhang Separating Intertemporal Substitution and Risk Aversion: an Experimental Study Songfa Zhong National University of Singapore ABSTRACT: This study aims to separate intertemporal substitution and risk aversion as proposed in Epstein and Zin (1989) in an experimental setting, where subjects make intertemporal allocation decision with risks across time either uncorrelated, or perfectly positively correlated, or perfectly negatively correlated. We find evidence for separation between intertemporal substitution and risk aversion in support of Epstein and Zin (1989), when comparing allocation decisions across treatments. Subsequently, we show that the estimated parameter for intertemporal substitution is significantly bigger than that for risk aversion. Our study sheds light on the understanding of the interplay between risk and time preferences. Author(s): Miao Bin; Zhong Songfa Walking through the Fog: experimental insights Wenting Zhou The University of York ABSTRACT: We use an analogy walking in the fog to explain the problem under study. Fog has the property that vision becomes less and less clear the further ahead that you try to look, and after a certain distance you simply cannot see anything (except the fog). Suppose you want to get to a certain point, the location of which you know, and you have a compass, a map, and a pedometer (so that you know at all times where, and how far away, is the destination). The only problem is that the map does not show heights, and you know that walking up and down hill consumes energy (the former more than the latter) and that you want to get to your destination with the least consumption of energy. Because of the fog you cannot, without unjustifiable assumptions, plan a route to the destination by backward induction; you must therefore rely on some forward looking way of proceeding to your destination. Rather obviously, we have no optimising theory to explain what people should do and hence we are running this experiment to see what they do do, and hence discover some satisficing heuristics. We report on the results from a pilot experiment. This problem, although set in a non-­‐economic context, obviously has relevance to many economic problems, not least the life-­‐cycle savings/consumption problem, for which backward induction still does not seem to be behaviourally valid. Author(s): Wenting Zhou; John D Hey Hierarchy Reversal Aversion: Evidence from Developmental and Cross-­‐Cultural Research Xinyue Zhou Sun Yat-­‐Sen University ABSTRACT: People prefer resource allocations that remove inequality, yet at the same time they have a strong tendency to maintain the status quo. These two motivations often conflict with each other. Here we designed an economic game (i.e., the redistribution game) to address a particular form of “ status quo motivation”, the motivation to prevent reversing the current socioeconomic hierarchy. In the redistribution game, participants were shown an unequal distribution of resources between two other people. Participants were given an option to transfer a fixed amount of resources from the advantageous side to the disadvantageous side. Choosing to transfer resources would invariably lead to the reduction of inequality. However, in some of trials, transferring resources would also lead to the reversal of the hierarchy. For example, the initial distribution is (4,1), that is, 4 for one person, and 1 for another person. The redistributor need to decide whether they want to transfer 2 from the advantageous side to the disadvantageous side or not to make the transfer. Thus, if participants just want to reduce inequality, they should choose to transfer, making the final distribution (2,3). In contrast, if participants are driven to maintain the hierarchy, they should choose not to transfer because making the transfer would lead to the reversal of the hierarchy. In Study 1, adult participants showed a strong tendency to maintain the current socioeconomic hierarchy, especially when reducing inequality led to hierarchy reversal. In Study 2, we examined the developmental origins behind this tendency by having young children between the ages of 3 and 10 playing the redistribution game. The 3-­‐6-­‐
yr-­‐old children were not afraid to reverse the hierarchy whereas 7-­‐8-­‐yr-­‐old children started to show reluctance to reverse the hierarchy. In Study 3, we used the same redistribution game to study a group of Nomadic Tibetan Herders in order to examine cultural differences. Tibetan Herders were more motivated to prevent hierarchy reversal than people living in industrialized societies. Author(s): Xinyue Zhou; Wenwen Xie; Qing Yang Harm on an Innocent Outsider as a Lubricant of Cooperation An Experiment Lilia Zhurakhovska Max Planck Institute for Research on Collective Goods, Bonn ABSTRACT: When testing the power of the morally grounded reticence to impose harm on an innocent outsider in a lab experiment, we were very surprised that widespread moral intuitions get it wrong. We tested participants on a one-­‐shot symmetric prisoner’s dilemma. In the treatment, insiders impose harm on a third, passive participant whenever at least one of them cooperates. Using the strategy method, we vary the degree of harm. However severe the harm, we do not find a significant difference between the baseline and the treatment. The results are even more striking if we control for beliefs. Insiders are more skeptical about the cooperativeness of other insiders if cooperation imposes harm on the victim. Yet, conditional on their more skeptical beliefs, they cooperate significantly more, as long as the harm is not particularly severe. Knowing that a bystander will suffer is a lubricant of cooperation. This result carries a second surprise: As has been shown, some people just cooperate because of altruism, but most of the cooperation can be explained by conditional cooperation. This explanation implies that there is a positive correlation between being optimistic about cooperativeness of others and own cooperativeness. The very implication is violated in our data. Apparently, beliefs about cooperativeness are not the only condition that determines cooperative behavior. Our data suggests that conditional cooperators are also motivated by the desire to distance themselves from outsiders. Apparently a cognitive component (beliefs) and a motivational component (increasing relative payoff) interact. Author(s): Christoph Engel; Lilia Zhurakhovska How and when can economic skills enhance cooperation? Johannes Ziesecke University of Marburg ABSTRACT: General wisdom has it that economic training and education tends to produce less cooperative people. This issue has attracted particular attention in the discussion of the current financial crisis where it was asked if an increasing marketization of societies has prepared the ground for amorally selfish behavior of financial intermediaries. We provide some evidence against this claim by help of an experiment. We use an investment game with a public-­‐goods character. In 4 different settings we test the influence of economic guidance and the impact of different designs like a partner-­‐ in comparison to a random-­‐stranger-­‐design. We show that a moderate guidance of strategic abilities increases the degree of cooperation if the institutional setting allows for reputation building. Thus, we conclude economic practice can enhance cooperation in a socially stable environment. Author(s): Evelyn Korn; Stephan Meisenzahl; Johannes Ziesecke Priming and Asset Markets Arthur Zillante University of North Carolina at Charlotte ABSTRACT: Trading in financial markets is believed to be influenced by more than market fundamentals. We use a laboratory experiment to examine the link between behavior in experimental asset markets and the psychological literature on priming moods and emotional states. Subjects in these experimental markets participate in three tasks. The first is a standard Holt-­‐Laury risk elicitation task used to bound the subject’s innate risk attitude. The second is a scrambled sentence test, where subjects are asked to form complete sentences from a set of words. Some subjects see sets of words intended to prime risk-­‐seeking and risk-­‐averse behavior. The third task is an asset market experiment with two assets. One asset is relatively safer than the other as the dividend distribution for one asset second order stochastically dominates the other. Our results show that even with two assets the bubble found in traditional experimental asset markets exists. Additionally, there is some evidence that subjects in the treatment with the risk-­‐
seeking prime trade at higher levels and more frequently than those in other treatments. Author(s): Arthur Zillante; Jim Bartkus Coordination with Communication under Oath Adam Zylbersztejn University Paris 1, Paris School of Economics ABSTRACT: Herein we explore whether the social psychology theory of commitment via a truth-­‐telling oath can reduce coordination failure. Using a classic sequential coordination game, we ask all players to sign voluntarily a truth-­‐telling oath before playing the game with cheap-­‐talk communication. Three results emerge with commitment-­‐via-­‐the-­‐oath: (1) coordination increased by nearly 50 percent; (2) senders’ messages were significantly more truthful and actions more efficient, and (3) receivers’ trust of messages increased. Keywords: Coordination game; Cheap talk communication; Oath. JEL Classification: C72; D83. Author(s): Nicolas Jacquemet; Stephane Luchini; Jason Shogren; Adam Zylbersztejn