### Séminaires

# Lunch séminaire Théorie, Organisation et Marchés (TOM)

L’objectif du Lunch séminaire TOM (Théorie, Organisation et Marchés) est de fournir un cadre convivial et régulier dans lequel des recherches en cours peuvent être discutées de façon informelle.Les travaux seront à contenu principal théorique autour des thèmes marchés / jeux / finance / contrats / enchères / rationalité / organisations. Il est ouvert à tous les doctorants et chercheurs du site. Il accueille des présentations de doctorants, ainsi que des présentations de travaux en cours de chercheurs, du site ou de passage, jeunes ou moins jeunes.

Ce séminaire est organisé par Antonin Macé

Contact opérationnel : No Rakotovao

Inscrivez-vous ici (avant lundi 18h) pour commander un sandwich ou une salade

Ce séminaire bénéficie d’une aide de l’État gérée par l’Agence Nationale de la Recherche au titre du programme d’Investissement d’avenir portant la référence ANR-17-EURE-0001.

## Prochainement

Aucun événement à venir.

## Archives

**Jeudi 27 juin 2024 12:30-13:30**- R2-20
**WEBER Giacomo**:__Coarse Agents and Intergroup Phenomena__- RésuméThis paper proposes a framework for analyzing intergroup phenomena. We consider heterogeneous agents divided into two groups. Agents are paired to play normal-form two-player games under complete information. When opponents belong to the same group (in-group), players form correct expectations about their behavior. Conversely, players form coarse expectations when their opponent is from the out-group. In equilibrium, these coarse expectations must coincide with the average behavior observed in the out-group. We apply this framework to an organizational setting where the groups represent subdivisions, and each game corresponds to a team task. These tasks are assumed to exhibit strategic complementarities. An omniscient designer assigns agents to teams in a way that maximizes the overall probability of task success. Our study focuses on the optimal assignment strategy, emphasizing the role of coarse expectations. We also discuss further economic applications

**Jeudi 20 juin 2024 12:30-13:30**- R2-20
**SERRANO ROBERTO**(Brown University) :__Mediated (Anti)Persuasive Communication__**SERRANO Roberto**(Brown University)**Zeky Murra-Anton**- RésuméCan private information or mediation change a sender’s behavior and im- prove the receiver’s expected utility in persuasive communication games? In a mediated Bayesian persuasion model, private information cannot improve the receiver’s expected utility when the sender communicates it. When the inter- mediary communicates the private information, the receiver’s expected utility improves only with a positive autarky value of the intermediary’s private infor- mation (AVIPI), a novel information accuracy measure we propose. Finally, the sender’s strategic behavior is generally affected by the intermediary’s presence as he tries to persuade the intermediary to, in turn, persuade the receiver

**Jeudi 13 juin 2024 12:30-13:30**- R2-20
**BARDIER Pierre**(PSE) :__Unanimity of two-selves in decision making__**Pierre BARDIER, Bach DONG-XUAN, Van-Quy NGUYEN.**- RésuméWe propose a new model of incomplete preferences under uncertainty, which we call unanimous dual-self preferences. Act f is considered more desirable than act g when, and only when, both the evaluation of an optimistic self, computed as the welfare level attained in a best-case scenario, and that of a pessimistic self, computed as the welfare level attained in a worst-case scenario, rank f above g. Our comparison criterion involves multiple priors, as best and worst cases are determined among sets of probability distributions, and is, generically, less conservative than Bewley preferences and twofold multi-prior preferences, the two ambiguity models that are closest to ours.

**Jeudi 6 juin 2024 12:30-13:30**- R1-14
**ELMSAUHER Béla**(PSE) :__Level 0 specification in 3x3 games: an experiment__- RésuméOne regular critic of the level-k model is the level 0 specification, which is often chosen arbitrarily. In this experiment, I present multiple 3x3 games (2 players having 3 possible strategies) to understand what is the most realistic level 0 specification. The most commonly used level 0 specification is the uniform one (i.e. the level 0 plays randomly) and it can be compared to a maxmax level 0 (i.e. the behavior with the highest potential) and to a maxmin level 0. In my experiment these level 0 specifications are also compared to a cooperation strategy. In a pilot session with 62 participants, I found that subjects played quite heterogeneously with the cooperation strategy being played the most often

**Jeudi 30 mai 2024 12:30-13:30**- R1-16
**MOREIRA Humberto**(Fundação Getulio Vargas’ Brazilian School of Economics and Finance) :__Market Power and Insurance Coverage__**GOTTLIEB Daniel (London School of Economics)**- RésuméWe study how market power affects insurance policies in a general class of models. We show that exclusion is a robust feature of insurance with a monopolistic firm (or when firms have enough market power), but not with perfect competition. However, the reduction in coverage relative to competitive markets is not uniform. While monopolists under-provide coverage for individuals with low willingness to pay, competitive markets under-provide coverage for those with a high willingness to pay to avoid cream skimming by competitors. The welfare comparison between perfectly competitive and monopolistic markets depends on whether the distortion at the extensive margin (higher under monopoly) exceeds the distortion at the intensive margin (higher under competition for those with a high willingness to pay). Using simulations based on an empirical model of preferences, we find that those both of these effects are quantitatively important.

**Jeudi 23 mai 2024 12:30-13:45**- R1-09
**MARIOTTI Thomas**(TSE) :__The war of attrition under uncertainty: Theory and robust testable implications__- RésuméWe study the war of attrition with symmetric information when players’ payoffs depend on a homogeneous linear diffusion. We first show that a player’s mixed Markov strategy can be represented by an intensity measure over the state space along with a subset of the state space over which the player concedes with probability 1. We then show that, if players are asymmetric, then, in all mixed-strategy Markov-perfect equilibria, these intensity measures must be discrete, and characterize any such equilibrium through a variational system for the players’ value functions. We illustrate these findings by revisiting the standard model of exit in a duopoly under uncertainty and construct a mixed-strategy Markov-perfect equilibrium in which attrition takes place on path despite firms having different liquidation values. We show that firms’ stock prices comove negatively over the attrition zone and exhibit patterns documented by technical analysis

**Jeudi 16 mai 2024 12:30-13:30**- R1-14
**BLUMENTHAL Benjamin**(ETH Zurich) :__Informational Lobbying and Implementation Standards__- RésuméPolicymakers are often uncertain about the right course of action. To inform their decisions, policymakers might rely on information provided by interest groups (IGs). Given that their interests are often misaligned, IGs might under-provide information to policymakers. This paper explores the possibility for policymakers of committing to implementation standards prior to IGs' lobbying, to induce more information transmission. I show that setting implementation standards ex-ante can benefit policymakers, despite possible ex-post inefficiencies, by inducing informational lobbying in cases in which IGs would not have lobbied with implementation standards set ex-post. I discuss implications of these results for constitutional design, legislative politics, and bureaucratic politics.

**Mardi 7 mai 2024 12:30-13:30**- R1-09
**LLEONART ANGUIX Manuel**(Universitat Autònoma de Barcelona) :__Firm Organization Under Cover-Ups__- RésuméThis paper explores the dynamics of optimal organizational hierarchies in environments where supervisors may engage in cover-ups upon detecting mistakes. Employing a theoretical model, I investigate how firm size, the cost of mistakes, and their probability influence the wage structure within firms. I discuss the potential implications of organizational culture and the likelihood of cover-ups on the optimal firm hierarchy. This study contributes to the understanding of optimal organizational design and provides insights for firms seeking to mitigate the risks associated with errors and cover-ups in hierarchical structures

**Jeudi 25 avril 2024 12:30-13:30**- R1-13
**SATPATHY Aviman**(PSE) :__Navigating Complexity in Choice under Uncertainty: Coarse Payoff-Assessment Learning Model__**Philippe Jehiel**- RésuméWe introduce the novel Coarse Payoff-Assessment Learning (CPAL) model that describes reinforcement learning in agents who tend to simplify the choice problems they face by focusing on the aggregate consequences of choosing over different clusters of alternatives (pre-arranged into exogenously defined similarity classes) instead of assessing each alternative individually. This technique is especially relevant in choice problems with uncertainty where the overall set of alternatives is too large to be considered extensively. We consider a smooth version of such a learning model and apply it to families of decision problems that differ in the set of available similarity classes. We first note that the steady-states of the learning dynamics correspond to a smooth (quantal) version of the Valuation Equilibrium [Jehiel and Samet, 2007]. We present examples of choice problems that reveal the possibility of multiple equilibria and verify the local asymptotic stability of pure equilibria. In contrast, we also identify conditions under which a unique, fully mixed equilibrium emerges - characterized by identical valuations across all similarity classes involved in the mixing, although the agent almost surely selects the alternative with the highest valuation. In particular, when the trivial decision problems involving a single similarity class lead to sufficiently high payoffs, we show that the latter arises, and we prove that the unique steady-state (that involves uniform randomisation) is globally asymptotically stable in the learning dynamics.

**Jeudi 4 avril 2024 12:30-13:30**- R1-14
**XEFTERIS Dimitrios**(University of Cyprus) :__Information aggregation with delegation of votes__**Amrita Dhillon, Grammateia Kotsialou, Dilip Ravindran**- RésuméLiquid democracy is a system that combines aspects of direct democracy and representative democracy by allowing voters to either vote directly themselves, or delegate their votes to others. In this paper we study the information aggregation properties of liquid democracy in a setting with heterogeneously informed truth-seeking voters -- who want the election outcome to match an underlying state of the world -- and partisan voters. We establish that liquid democracy admits equilibria which improve welfare and information aggregation over direct and representative democracy when voters' preferences and information precisions are publicly or privately known. Liquid democracy also admits equilibria which do worse than the other two systems. We discuss features of efficient and inefficient equilibria and provide conditions under which voters can more easily coordinate on the efficient equilibria in liquid democracy than the other two systems.

**Jeudi 28 mars 2024 12:30-13:30**- R1-14
**FRANKEL David**(Melbourne Business School) :__Equilibrium Selection in Participation Games, with Applications to Security Issuance__- RésuméIn many applied settings, an activity requires a critical mass of participants to be worthwhile. This can give rise to multiple equilibria. We study seven well-known equilibrium selection theories: two heuristic arguments, two models with rational players, and three from the evolutionary literature. With one exception, each relies on strategic complementarities. We weaken this to a mild single crossing property and show that the theories' predictions have a common form: an agent plays a best response to some fictional distribution of the participation rate of her opponents. We then use this robust framework to study security design in a setting in which issuance revenue is used to fund investments that are, in turn, used to pay distributions on the securities. We show that all monotone securities are underpriced and that debt is optimal as it is the least underpriced. Moreover, underpricing in equity offerings can be mitigated by share rationing and a minimum sales requirement.
- Texte intégral [pdf]

**Jeudi 21 mars 2024 12:30-13:30**- R1-14
**SI Shuhua**(Columbia) :__Modes of Thinking in a Complex Game__- RésuméWe study a complex Blotto-like resource allocation game over a continuum of items of different values. Strategies a priori lie in a (high-dimensional) space of functions, each defining how much to bid on each item as a function of its value, given a total resource constraint. We propose to analyze the game in a simplified strategy space: we focus on a three-dimensional space of curve which includes power shapes and S-shapes and analyze the game assuming that a player's strategy set consists of a random selection of visually distinct curves that paves the space of curves. By randomly generating relatively small strategy set profiles, we obtain a distribution of most frequently played strategies for each solution concept examined (e.g., maxmin, best-response dynamics, Nash) which we compare to curve-fitted experimental data. Predictions using restricted sets point towards a significant use of S-shaped curves, as indeed overwhelmingly observed in the data, in contrast to the power curve prediction that obtains when maxmin or equilibrium analysis is performed over rich strategy sets. We also examine comparative statics with respect to asymmetries in resource endowments across players.

**Jeudi 14 mars 2024 12:30-13:30**- R2-21
**BORISOVA Polina**(PSE) :__Hyperbolic Experimentation__**Nikhil Vellodi**

**Jeudi 7 mars 2024 12:30-13:30**- R2-21
**BI Wei**(European University Institute) :__Selling Incremental Products Optimally__- RésuméThis paper studies the intricacies of optimal selling mechanisms for incremental products, with a particular emphasis on scenarios where agents' private information profoundly influences their preferences across periods. To mitigate their intertemporal information rent, the manufacturer encourages upgrades through refunds rather than payments, introducing inefficiencies but maximizing profits. Our findings reveal that even without commitments, the manufacturer prefers innovative upgrades in the iterated products. Importantly, consumers bear all efficiency loss due to innovative upgrades. Additionally, we reveal that per-consumer information disclosure allows the manufacturer to strategically tailor disclosures about innovation, enhancing incentives for truthful reporting and extracting more surplus from specific consumer segments.

**Jeudi 29 février 2024 12:30-13:30**- R2-01
**USHCHEV Philipp**(ULB) :__The Network Origins of Entry__- RésuméWe develop a model of the process of entry under social learning via word-of-mouth (WOM). An incumbent's product is known to the consumers, while the success of a potential entrant hinges on creating consumers' awareness of the entrant's product through WOM. We model WOM as a percolation process on a random graph. We show that whether an entrant can gain a non-negligible level of awareness depends on the social network structure via two sufficient statistics, which are the ratios of different factorial moments of the degree distribution. We categorize the different pricing equilibria into the classical blockaded, deterred, and accommodated entry taxonomy. Under deterred entry, our model produces a model of limit pricing by an incumbent to prevent an entrant gaining a non-negligible level of awareness. When we focus on a multinomial logit demand, we show that increasing the network density shifts the pricing equilibrium from blockaded to deterred, and finally to accommodated entry. We also show that the aggregate consumer surplus may be non-monotonic with respect to network density. Finally, if the incumbent has knowledge about the consumer's number of friends and can charge personalised prices, we find that it is optimal for the incumbent to charge lower prices to more connected consumers.

**Jeudi 21 décembre 2023 12:30-13:30**- R1- 15
**ANGELUCCI Charles**(MIT Sloan School of Management) :__Searching for Collaboration: The Dynamics of Relationship Building__**ORZACH Roi**(MIT Sloan)- RésuméBuilding a successful collaboration is often a time-intensive and gradual process. We model collaborative dynamics with self-enforcing incentives. Two players are presented with infinitely many ex-ante identical projects, each yielding asymmetric benefits. Every period, they collectively explore or exploit multiple projects and make voluntary transfers to each other. After exploring a project, players learn its benefits and choose whether to exploit the project in future periods. We show that lengthy exploration occurs, and that the way the collaboration evolves exhibits significant path dependence. Players temporarily exploit projects, return to previously abandoned projects, and initially explore a limited number of projects

**Jeudi 7 décembre 2023 12:30-13:30**- R1-15
**KIRNEVA Margarita**(CREST) :__Informing to Divert Attention__- RésuméI study a multidimensional Sender-Receiver game in which Receiver can acquire limited information after observing the Sender's signal. Depending on the parameters describing the conflict of interest between Sender and Receiver, I characterise optimal information disclosure and the information acquired by Receiver as a response. I show that in the case of partial conflict of interests (aligned on some dimensions and misaligned on others) Sender uses the multidimensionality of the environment to divert Receiver's attention away from the dimensions of misalignment of interests. Moreover, there is negative value of information in the sense that Receiver would be better off if she could commit not to extract private information or to have access to information of lower quality. I present applications to informational lobbying and consumer's choice.

**Jeudi 23 novembre 2023 12:30-13:30**- R1-15
**LIQUI LUNG Caroline**(Cambridge University) :__Intersectionality in Individual Choice Behavior: The Pitfalls and Opportunities__- RésuméI show how intersectionality, the interconnections of social organizations that create interdependent systems of disadvantage, plays a role in individual choice behavior. I analyze the model in Liqui Lung (2022) with multi-dimensional social types and show how agents endogenously determine which dimensions of social identity affect belief formation as a function of their exogenously specified social type, their social context and their underlying ability. I discuss what the implications of this behavior are for both individual and aggregate choice behavior, and show how social constraints, such as stigmatization or stereotypes, and endogenizing the social type, affect these patterns. I then illustrate how these insights could help explain the pitfalls we encounter in the evaluation of one-dimensional policy measures, and discuss how we could harness intersectionality to develop more effective approaches.

**Jeudi 16 novembre 2023 12:00-13:00**- R1-15
**CONJEAUD Ivan**(PSE) :__*Spontaneous coupling of Q-learning algorithms in equilibrium__- RésuméMost contributions in the algorithmic collusion literature only consider symmetric algorithms interacting with each other. We study a simple model of algorithmic collusion in which Q-learning algorithms repeatedly play a prisoner's dilemma and allow players to choose different exploration policies. We characterize behavior of such algorithms with asymmetric policies for extreme values and prove that any Nash equilibrium features some cooperative behavior. We further investigate the dynamics for general profiles of exploration policy by running extensive numerical simulations which indicate symmetry of equilibria, and give insight for their distribution.

**Jeudi 9 novembre 2023 12:30-13:30**- R1-15
- *

**Jeudi 19 octobre 2023 12:30-13:30**- R1-15
**MACKENZIE Andrew**(Maastricht University) :__Tract housing, the core, and pendulum auctions__**Andrew Mackenzie and Yu Zhou**- RésuméWe consider a model of tract housing where buyers and sellers have (i) wealth constraints, and (ii) unit demand over identical indivisible objects represented by a valuation. First, we characterize the strong core. Second, we characterize the bilateral weak core, or the weak core allocations with no side-payments. Finally, when buyer wealth constraints and valuations are private information and when transfers are discrete, we introduce two families of pendulum auctions, both of which consist of obviously strategy-proof implementations of the bilateral weak core. The buyer-optimal pendulum auctions are preferred by the buyers but are inefficient when side-payments are possible, while the efficient pendulum auctions are efficient

**Jeudi 5 octobre 2023 12:30-13:30**- R1-15
**KRANTON Rachel**(PSE) :__Social Connectedness and Information Markets__- RésuméThis paper investigates information quality in a simple model of socially- connected information markets. Suppliers’ payoffs derive from the fraction of consumers who see their stories. Consumers prefer to share and act only on high-quality informa- tion. Quality is endogenous and highest when social connectedness is neither too high nor too low. In highly-connected markets, low-quality stories are widely seen, giving suppliers little incentive to invest in quality. Increasing the volume of misinformation and increasing consumers’ cost of tuning in to suppliers’ broadcasts can each increase equilibrium information quality

**Jeudi 28 septembre 2023 12:30-13:30**- R1-15
**ZERBINI Antoine**(LSE) :__Segment and Rule: Modern Censorship in Authoritarian Regimes (with Kun Heo)__- RésuméThe internet grants citizens with the ability to choose which media outlets to consume. This access to foreign independent outlets risks exposing citizens to negative information about the authoritarian regime. In response to this novel threat, authoritarian regimes introduced censorship firewalls which seem to fail at their task: millions bypass these restrictions everyday. We contend that the regime deliberately aims for a specific segment of the population to self-select into bypassing the firewall: regime opponents. By bypassing the firewall, opponents are occasionally convinced to comply after seeing positive and credible reporting about the regime by banned foreign outlets. The firewall ensures that regime supporters exclusively consume content from domestic outlets and so their compliance is secured via regime propaganda. We label this strategy one of segment-and-rule and show how it maximizes compliance. We also explain how authoritarian regimes can engineer segment-and-rule by making local outlets parrot the party line, investing in domestic entertainment or strategically banning foreign entertainment. By providing citizens with more choice, the internet may have entrenched authoritarian regimes.

**Jeudi 14 septembre 2023 12:30-13:30**- R2- 21
**MOISSON Paul-Henri**(TSE) :__Meritocracy and Inequality__- RésuméThe paper develops a model of career concerns in which agents publicly choose between several activities in which to exert effort, yet differ along a privately observable characteristic (headstart) that affects their performance. The agents' audience values their talent, effort and headstart. We highlight the race between two effects: a displacement effect by which the poor (headstart-wise) try to avoid a lower talent image and thus avoid the activity chosen by the rich, and a distinction effect by which the rich try to reap a higher headstart image and thus avoid the activity chosen by the poor. While displacement drags the poor towards activities with lower incentives on effort, distinction pulls the rich towards activities with higher incentives. We interpret the model in terms of meritocracy, characterize optimal activity design and discuss several policy implications, emphasizing the unintended consequences of common interventions
- Texte intégral [pdf]

**Jeudi 22 juin 2023 12:30-13:30**- Salle R1-16, Campus Jourdan, 75014 Paris
**VANSTEENBERGHE Eric**(PSE) :__*Insurance Supervision under Climate Change: A Leader Detection Method__- RésuméThis paper introduces a novel supervisory tool to enhance resilience in insurance markets faced with climate change uncertainties. I develop a theoretical model simulating an insurance market with non-cooperative insurance experts who set premiums based on subjective risk modeling. I present a supervisory tool that pools expert opinions to identify fast-learning leaders who can accurately classify extreme climate-induced events. This tool addresses the challenge of insurability with a changing climate and helps supervisors anticipate shifts in risk parameters. The robustness and utility of the proposed tool are validated through a series of simulations where it demonstrates superior performance compared to conventional pooling methods.

**Jeudi 15 juin 2023 12:00-13:15**- Salle R1-14, Campus Jourdan, 75014 Paris
**LETROUIT Lucie**(Université Gustave Eiffel) :__*"Innate rewards and decision making"__- RésuméIn the present paper, I propose a dual-self model in which decision utility relies on two fundamental sources of value, namely innate rewards (which humans experience innately, because they used to be evolutionarily useful, e.g. rewards linked with eating sugary or fatty food that signal a high calorie content...) and constructed values (which rely on a complex cognitive model of the world and of the individual’s personal objectives). Typically, humans rely more on their innate rewards when their needs are depleted or their emotions high. The specificity of this model with respect to classical self-control models resides in the fact that both innate rewards and the weight of innate rewards in the decision utility are considered endogenous. I argue that it (1) allows to shed new light on various economic behaviors and (2) to identify new sources of economic inefficiency, linked with the nowadays wide discrepancy between the innate rewards elicited by actions and these actions' constructed values (e.g. too sugary food that turns out to be bad for health…). (3) I argue that it calls for new types of policy measures aimed at (3.a) shaping the innate rewards associated with goods and purchase situations, so that behaviors maximizing the decision utility also maximize the (real) underlying utility and (3.b) reducing the role of innate rewards with respect to constructed values in typical day-to-day economic decisions. (4) I briefly discuss how such policy measures could be operationalized in practice.

**Jeudi 8 juin 2023 12:30-13:30**- Salle R1-10, Campus Jourdan, 75014 Paris
**GHOSH Rajarshi**(ESSEC) :__*Quadratically Normalized Utilitarian Voting__**Marcus Pivato**- RésuméMost efficient voting mechanisms assume quasi-linear preferences of the voters and require them to express the intensity of their preferences using money (e.g. by bidding or buying votes). Moreover, mechanisms in the current literature that use an artificial currency instead can only be applied to multiple simultaneous binary decisions. We propose a new mechanism named "Quadratically Normalized Utilitarian Voting" which does not use money to buy votes, does not assume quasi-linearity of the voters' utilities, and can be used for single propositions. We show that voters vote in proportion to their true utility for an alternative in all Nash Equilibria and that the mechanism maximizes a weighted utilitarian social welfare function at each such equilibrium.

**Jeudi 1er juin 2023 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**DEL MERCATO Elena**(Paris 1) :__*Sufficient conditions for a "simple" decentralization with consumption externalities__**Van-Quy Nguyen**- RésuméWe consider a pure exchange economy with consumption externalities in preferences. We use the notion of competitive equilibrium à la Nash. We provide the Social Redistribution assumption to restore the Second Welfare Theorem. We then introduce the differentiable characterizations of Social and Strong Redistribution. We show that all these conditions are weaker than other relevant assumptions studied in the literature. Our conditions entail interesting results on the decentralized implementation of Pareto optima, that link together the competitive supporting price and the shadow price of the utilitarian social planner. Finally, we provide a specific condition for Bergson-Samuelson utility functions, which has a nice interpretation in terms of positive or negative externalities.

**Mercredi 24 mai 2023 14:00-15:15**- Salle R2-21, Campus Jourdan, 75014 Paris
**BERNHEIM Douglas**(Stanford) :__*The Challenges of Behavioral Welfare Economics__- RésuméONLINE - joint with seminar Behaviour

**Jeudi 11 mai 2023 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**COUTINHO Rafaël**(Universidade Federal de Pernambuco) :__*Infinitely repeated cheap-talk: From information to influence__**Paulo Melo-Filho**- RésuméWe solve an infinitely repeated version of Crawford and Sobel (1982) Strategic information transmission model. We find that repeated interaction allows for perfect information transmission for patient players, in the spirit of folk theorems. For impatient players, however, we also find repeated partition equilibria that are more informative than in the static game. Additionally, if the decision maker offers a biased policy, the partition equilibrium becomes even more informative. For small policy bias, more information is Pareto improving, whereas for large bias only the sender benefits. This suggests that, through repeated interaction, the sender can use his superior information to obtain a favorable policy. Applying this model to lobbying shows that interest groups can influence policy makers without monetary contributions. Plus, the influence can have positive or negative welfare impacts, depending on the level of capture.

**Jeudi 20 avril 2023 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**AGRANOV Marina**(Caltech) :__*"Information Aggregation on Networks: an Experimental Study"__**joint work with Ben Gillen and Dotan Persitz**- RésuméWe study the impact of network architecture on the efficiency of information transmission and dynamics of learning in large networks using laboratory experiments. While the theoretical literature has recently made progress in identifying the geometric features of networks that enable the flow of information and those that impede it, the empirical literature lags behind. Our project will fill this gap by designing a novel interface that allows studying the interplay between network architecture and information diffusion in large networks in a controlled laboratory environment. In particular, we address the following questions: How do network structures affect the likelihood of reaching a consensus? Conditional on reaching the consensus, how likely agents are to choose the correct action? How fast convergence occurs? Is it possible at all to observe consensus in large networks? Is it possible to observe connected networks in which agents agree to disagree?

**Jeudi 6 avril 2023 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**WEBER Giacomo**(PSE) :__*On Coarseness in Organizations__- RésuméI investigate a team-task assignment problem for a fully informed designer who wants to maximize overall effort. The designer faces two problems: (i) how to sort heterogeneous (on their cost of effort) agents into pairs; (ii) how to assign a task from a set of heterogeneous tasks to each pair. Tasks are characterized by different degrees of strategic complementarity. The complete information case serves as a benchmark: the designer should pair agents assortatively and assign tasks with higher degree of complementarity to more efficient pairs. I assume that agents from different departments who work together on some task will form "coarse beliefs" on the actions of their opponents. The designer can exploit coarseness to induce an overall higher level of effort, by pairing the least efficient workers across departments and the most efficient within. I compare these findings with the case of incomplete information.

**Jeudi 23 mars 2023 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**BALLESTER Miguel**(Oxford University) :__*The Rationalizability of Survey Responses (with Jose Apesteguia)__- RésuméIn this paper we propose and study the notion of survey rationalizability. For the simplest case of dichotomous surveys, rationalizabilitymeans that both questions and ideal views of individuals can be located in the real line in a way that agreed questions correspond to those providing higher utility than a threshold. We show how the relative location of questions can be learnt using a revelation mechanism that involves pairs of individuals and triplets of questions, and that the acyclicity of these revelations suffices for rationalizability. We then show that our analysis readily extends to the cases of non-dichotomous Likert-scales and of probabilistic responses. We further study the identification of the main parameters of the model and show that, in an exponential version of the probabilistic model, even the cardinal locations can be fully determined. We conclude by studying an alternative model of survey responses known as Guttman-scales.

**Jeudi 16 mars 2023 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**VOHRA Rakesh**(University of Pennsylvania) :__*Contagion and Bailouts in Financial Networks__- RésuméDiversified cross-shareholding networks are thought to be more resilient to shocks, but diversification also increases the channels by which a shock can spread. To resolve these competing intuitions we introduce a stochastic model of a diversified equity holding network in which a firm's valuation depends on its cash endowment and the shares it owns in other firms. Our stochastic model of holding networks is based on random matrices rather than Erdos-Renyi type random graphs. We show that a concentration of measure phenomenon emerges: almost all realized holding networks instances drawn from any probability distribution in a wide class are resilient to contagion if endowments are sufficiently large. Furthermore, the size of a shock needed to trigger widespread default increases with the exposure of firms to each other. Distributions in this class are characterized by the property that a firm's equity shares owned by others are weakly dependent yet lack ``dominant'' shareholders. As widespread default involves substantial deadweight costs, to counter them, a regulator can inject capital into failing firms. These injections have positive spillovers that can trigger a repayment cascade. But which firms should the regulator bailout so as to minimize the total injection of capital while ensuring solvency of all firms? While the problem is, in general, NP-hard, for a wide range of networks that arise from a stochastic model of the kind just described, we show that the optimal bailout can be implemented by a simple index policy in which a firm index depends only on its characteristics and its position in the network. Specific examples of the setting include core-periphery networks. This talk is based on two papers. One joint with Victor Amelkin and Santosh Venkatesh and the other joint with Krishna Dasaratha and Santosh S. Venkatesh.

**Jeudi 9 mars 2023 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**TINE Vinicius**(UFPE) :__*Campaign Spending, Media Capture, and Political Accountability (co-author : Rafael Coutinho Costa Lima)__- RésuméCampaign spending and media influence are well-known instruments which are intensely utilized by special interest groups to affect political decisions, and the joint utilization of those instruments has been documented and discussed. However, both literatures on campaign spending and on media capture in political economy have developed independently, possibly missing important interconnections and therefore a broader understanding of the behavior of interest groups and their influence on democracy. In that context, we develop a theoretical analysis of the strategical relationship between campaign spending and media influence, as a step towards connecting those literatures. We develop a political agency framework which describes how special interest groups utilize campaign finance and the media to influence voters' information on the political process, in order to extract political rents. We show that when the marginal cost of political campaigns increase, campaign effort and media influence are strategical substitutes, and higher voter welfare is obtained. On the other hand, when the marginal cost of media influence increases, those instruments exhibit complementarity. In that case, higher voter welfare is not necessarily attained, and it might decrease if information levels are already too high. Finally, we analyse the effects of tighter campaign spending caps in the model, showing that higher voter welfare is obtained even when illegal campaign funds are used.

**Jeudi 15 décembre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**FRIEDMAN Evan**(PSE) :__*QRE with a continuum of types.__- RésuméQuantal response equilibrium (QRE), a statistical generalization of Nash equilibrium, is a standard benchmark in the anlaysis of experimental data. A QRE is a solution to a non-polynomial system of equations, and so does not have a closed-form expression. For this reason, applications rely on numerical methods, and theoretical results are limited. An insight of recent work on finite games is that, whereas it is difficult to solve for any given QRE, it is often easy to characterize the set of all QRE that satisfy a regularity condition. This non-parametric QRE analysis implies novel bounds on the common parametric models and opens up the possibility for novel theoretical applications. This analysis, however, simply does not apply to infinite games. In this paper, we identify a class of infinite games–those with binary actions and a continuum of types–for which non-parametric QRE analysis is feasible. We derive characterization results and apply them to several classic games.

**Jeudi 8 décembre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**ROESLER Anne-Kathrin**(University of Toronto) :__*How informed do you want your principal to be? joint with Rahul Deb and Mallesh Pai__- Résumépresentation joint with Roy Seminar A buyer's value for a good depends on his private type and the quality that is unknown to him. The seller of the good learns about the quality via a signal; the signal realization is the seller's private information. We pose and answer the following question: how much (if any) private information would the buyer want the seller to have? Formally, we characterize the buyer-optimal outcome: this is the signal and the corresponding seller-optimal equilibrium of the informed principal game that yield the highest consumer surplus. We show that there are conditions under which private information for the seller can lead both to greater profits and higher consumer surplus; that is, compared to being uninformed, seller private information can lead to Pareto gains. The seller's signal in the buyer optimal outcome is typically not fully informative.

**Jeudi 1er décembre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**TALLON Jean-Marc**(PSE) :__*Efficient Allocations under Ambiguous Model Uncertainty - Chiaki Hara (Kyoto University), Sujoy Mukerji (Queen Mary University of London), Frank Riedel (University of Bielefeld), Jean Marc Tallon (Paris School of Economics)__- RésuméWe study efficient allocations in an economy of consumers who have heterogeneous smooth ambiguity preferences but a common identifiable set of relevant probability measures in the sense of Denti and Pommato (2022). When the efficient allocations would be represented by linear risk sharing rules in the absence of ambiguity, we identify the systematic nonlinearity caused by its presence, and explore asset pricing implications thereof, by constructing a representative consumer. We show, among other things, that the Hansen-Jagnnathan bound is higher than in the ambiguity-neutral case and moves countercyclically, a behavior that has not been explained by any representative-consumer model.

**Jeudi 17 novembre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**PETERS Dominik**(Dauphine) :__*Voting Rules in Participatory Budgeting__- RésuméUsing participatory budgeting (PB), cities give their residents an opportunity to influence how the city's budget is spent. This is often done by collecting project proposals, and then voting on those proposals. This talk discusses work in computational social choice on this voting problem. First, we discuss the voting systems in use in major cities, which mostly involve a naive greedy algorithm based on approval scores. Then we formalize a model of voting under a knapsack constraint with additive valuations, and discuss how standard approval votes fit in that model. We then discuss a new voting rule, the Method of Equal Shares, that we have recently proposed and that is planned to be first used by a small city in Poland next year. This rule provides proportional representation, formalized by an axiomatic guarantee, which I'll argue is a desirable goal in PB. Finally, I will mention PB work on the core, strategyproofness, computational complexity, and several extensions which could form directions for future research: allowing negative votes, allowing for constraints, as well as analyzing input formats and the process of agenda setting, among others.

**Jeudi 10 novembre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**SINGH Manpreet**(PSE) :__*A "soft" war of attrition__- RésuméI analyse a “soft” war of attrition in a procurement auctions framework inspired by renewable energy auctions in India. Before the auction, the auctioneer declares the quantity of a product it wants, and each bidder declares the amount they can provide. The bidders compete on prices in an open descending price auction. Winners provide the quantity they declared, a partial loser provides the residual demand, and full losers provide zero. I characterise perfect Bayesian equilibrium in these games under incomplete information and common prior in diverse cases. The payoff structure leads to bunching at the reserve price in the PBE by the bidder with the highest quantity. I conjecture the existence of an equilibrium, and prove that if it exists, it is unique. The presence of bunching and asymmetric payoff can lead to inefficient selection in such auctions. While developed as an auction, the framework can be easily extended to any game of sequential exit.

**Jeudi 20 octobre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**DEKEL Amit**(PSE) :__*Set-Valued Rational Expectations and Farsighted Stability__- RésuméAn abstract game consists of a set of states, preferences over states, and an effectivity correspondence specifying which coalitions are allowed to move from one state to another. Traditional solution concepts used to analyze abstract games under the assumption that players are farsighted (i.e. take into account the entire chain of moves that may follow their own), such as the Farsighted Stable Set or the Largest Consistent Set, do not fully respect the principle of optimizing behavior. This is commonly referred to as the maximality critique. Dutta and Vohra (2017) propose a fix to this problem that comes at the expense of restricting the protocol of play. Ray and Vohra (2019) propose a fix that comes at the expense of scaring existence away. We build on ideas in these papers to propose a fix which is free of such shortcomings. Unlike existing solution concepts, we specify expectations which are set-valued, meaning that several states may be expected to follow any given state. We relate the proposed solution concept to existing ones, discuss its properties, provide necessary conditions for existence, and study its application to strategic and extensive form games.

**Jeudi 6 octobre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**LACLAU Marie**(HEC) :__Communication on networks and strong reliability (joint with L. Renou and X. Venel)__- RésuméWe consider sender-receiver games, where the sender and the receiver are two distant nodes on a communication network. We show that if the network has two disjoint paths of communication between the sender and the receiver, then we can replicate not only all equilibrium outcomes of the direct communication game (i.e., when the sender and the receiver communicate directly with each other), but also of the mediated game (i.e., when the sender and the receiver communicate with the help of a mediator).

**Jeudi 22 septembre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**JANTSCHGI Simon**(UZH/Dauphine) :__Markets and Transaction Costs__- RésuméTransaction costs are omnipresent in markets but are often omitted in economic models. In markets in which the price is set to equate revealed supply and demand, we show that the presence of transaction costs can fundamentally alter incentive and welfare properties. We categorize transaction costs into two types. Asymptotically uninfluenceable transaction costs---such as fixed and price fees---preserve the key properties of the model without transaction costs, namely asymptotic strategyproofness, efficiency, and robustness to misspecified beliefs and to aggregate uncertainty. In contrast, influenceable transaction costs---such as spread fees---lead to complex strategic behavior (price guessing) that may result in severe market failure. When transaction costs are collected as fees by an intermediary, we show that the same categorization determines the intermediary's optimal behavior. Any uninfluenceable fee can maximize the intermediaries utility (e.g. revenue) if optimally scaled, while purely influenceable fees lead to zero revenue. Furthermore, all uninfluenceable optimally-scaled fees lead to the same welfare. Our insights extend beyond markets equalizing demand and supply.

**Jeudi 15 septembre 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**BLUMENTHAL Benjamin**(ETH Zurich) :__Policymaking under Influence__- RésuméPolicymaking is a fraught process: politicians often fail to change the status quo despite their best efforts. Interest groups (IGs) can, for better or for worse, affect the fate of politicians' proposals. I consider a model of policymaking with an imperfectly effective politician and an IG who, through costly lobbying, can affect the likelihood that the politician-proposed policy successfully replaces the status quo. I show that policymaking can be affected through threats of lobbying or active lobbying, depending on the cost of lobbying, the strength of the IG's influence, and the alignment of preferences between the IG and the politician. In particular, I show that the relationship between the IG's ability to shape policymaking and the cost-effectiveness of its lobbying can be non-monotonic. I also discuss empirical implications of the model and highlight the importance of taking into account status quo policies to study the welfare implications of IGs' activities.

**Jeudi 16 juin 2022 12:30-13:30**- Salle R1-15, Campus Jourdan 75014 Paris
**NG Robin**(UC Louvain) :__Ratings and Reciprocity__- RésuméRecent evidence suggests online ratings and reviews are intrinsically motivated and reviewers reciprocate offers of sufficient value-for-money with good ratings. We incorporate reciprocity into a model of ratings, where consumers rate positively if they receive a high value-for-money and perceive firms to be sufficiently kind. Our key mechanism is that to benefit from reciprocity, firms offer lower prices in exchange for good ratings, increasing future prices and profits. Firms reinforce this mechanism and are more likely to induce good ratings when i) consumers exhibit a stronger sense of reciprocity or when ii) it is easier for consumers to leave ratings. This way, the presence of reciprocity can explain the documented phenomenon of rating inflation, and less informative ratings, especially as it becomes easier to leave ratings. We discuss how further marketplace features affect the reciprocity mechanism and the informativeness of ratings: First, a two-sided platform might indeed choose to make it easier for consumers to rate, thereby undermining the informativeness of ratings. Second, many platforms screen the quality of sellers and kick out low-quality sellers. We characterize when such quality screening reduces or reinforces the informativeness of ratings. Third, competition between sellers leads to lower informativeness of ratings. Finally, we show that more-informative ratings do not necessarily translate to larger consumer surplus.

**Jeudi 9 juin 2022 12:30-13:30**- Salle R1.14, Campus Jourdan 75014 Paris
**LOJKINE Ulysse**(PSE, Université de Nanterre) :__A general power index__- RésuméI propose a general power index in games. The power of an agent over an outcome is understood as the equilibrium effect on the outcome of variations in the agent’s preferences. I show that the new index, ?, has the following properties: (i) classic measures of freedom of choice are a special case of the ? index in the context of individual choice; (ii) the Banzhaf and Shapley-Shubik indices are special cases of the ? index in the context of binary voting games; (iii) in bargaining games, the ? index of a player combines his outside option and other parameters; (iv) the ? index allows a generalization of the Barry decomposition of success between power and luck to all unidimensional spatial games; (v) the ? power of a firm on its price is related to its Lerner index, but also to its technology and to the potential entry of competitors; (vi) in a simple competitive exchange setting, each agent has zero ? power on the equilibrium price but a ? power density can be defined, which, under some assumptions on preferences, is proportional to the agents’ wealth.

**Mercredi 1er juin 2022 12:30-13:30**- Salle R1.14, Campus Jourdan 75014 Paris
**ZENOU Yves**(Monash University) :__Perceived Competition in Networks__**Olivier Bochet, Mathieu Faure, Yan Long**- RésuméWe consider an aggregative game in which agents have an imperfect knowledge about the set of agents they are in competition with. We model this lack of knowledge through a directed graph that we call the perception network. In this framework, a natural equilibrium concept emerges, the Perception-Consistent Equilibrium (PCE). At a PCE, each agent chooses an action level that maximizes her subjective perceived utility while the action levels of all individuals must be consistent. We prove the existence of PCEs in a large class of aggregative games. We also show that, at any PCE, the efforts are always ordered accordingly to some centrality measure in the perception network. For a specific subclass of aggregative games, we decompose the network into communities and completely characterize the PCEs by identifying which sets of agents are active, as well as their effort level. We prove that, at the unique stable PCE, the agents’ action levels are proportional to their eigenvector centrality in the perception network. We illustrate our results with two well-known models: Tullock contest and Cournot competition.
- Texte intégral [pdf]

**Jeudi 12 mai 2022 12:30-13:30**- Salle R1.14, Campus Jourdan 75014 Paris
**GHERSENGORIN Alexis**(PSE) :__Personalized Pricing with Inequality Concerns__**Victor Augias (Sciences-Po) ; Daniel M.A. Barreto (Sciences-Po).**- RésuméWe analyze the redistributive consequences of a monopolist who can obtain additional information about consumers' tastes beyond the prior distribution, and thus implement personalized pricing. We study the problem of a social planner, who can commit to an information policy and has redistributive preferences. The latter is captured by a social welfare function that put higher weights on poorer consumers. We show that for sufficiently strong redistributive preferences, the social planner will sacrifice richer consumers' surplus and gives additional profits to the monopolist.

**Jeudi 21 avril 2022 12:30-13:30**- Salle R1-14, Campus Jourdan, 75014 Paris
**BENABOU Roland**(Princeton) :__*__

**Jeudi 14 avril 2022 12:30-13:30**- Salle R1.14, Campus Jourdan 75014 Paris
**KLEIN Paul**(SU) :__Strategic Investment and Learning with Private Information__**Peter Wagner**- RésuméWe study a two-player game of strategic experimentation with private information in which agents choose the timing of risky investments. Agents learn about future returns through privately observed signals, others' investment decisions and from public experimentation outcomes when returns are realized. We characterize symmetric equilibria, and relate the extent of strategic delay of investments in equilibrium to the primitives of the information structure. Agents invest without delay in equilibrium when the most optimistic interim belief exceeds a threshold. Otherwise, delay in investments induces a learning feedback that may either raise or depress beliefs and investment choices. We show that private information in strategic experimentation can increase ex-ante welfare.

**Jeudi 7 avril 2022 12:30-13:30**- Salle R1.14, Campus Jourdan 75014 Paris
**PAHLKE Marieke**(PSE) :__Partial Feedbacks and Ambiguity Aversion in Normal-Form Games__- RésuméWe investigate players' behavior in normal form games when receiving partial feedback about the strategies of others. In our setting, we allow for strategic uncertainty and assume players are ambiguity averse. We characterize conditions for the existence of maxmin self-confirming equilibria under full and partial feedback. These insights can be used to investigate how to optimally design players' information feedback. In this talk, I will present some general results and two applications, one to games with negative externalities (e.g., Cournot competition) and one to public good games.

**Jeudi 31 mars 2022 12:30-13:30**- Salle R1-14 Campus Jourdan 75014 Paris
**CASELLA Alessandra**(Columbia) :__Delegation under Liquid Democracy: One Experiment and a Half__- RésuméUnder Liquid Democracy (LD), decisions are taken by referendum, but voters are allowed to delegate their votes to other voters. LD has been heralded as the golden medium between direct and representative democracy: better than the former because experts can be delegated votes and thus weigh more; better than the latter because different experts can be chosen for each decision, according to their specific competences. When experts are correctly identified, LD shifts voting weight from a larger number of less informed voters towards fewer better-informed voters. Theory shows that the outcome can be superior to simple majority voting. However, by reducing the number of voters, delegation reduces the variety of independent information sources. Optimal delegation is counter-intuitively rare. We report the results of a lab experiment where by design experts are correctly identified and yet systematic over-delegation leads to LD underperforming relative to simple majority. A second experiment, run on a large electorate online, will have a different design, making it possible for some information to be worse than random. We hypothesize that in such an environment, closer to the confused conditions in which political decisions are made, LD may perform better, relative to majority voting.

**Jeudi 17 mars 2022 12:30-13:30**- Salle R1.15, Campus Jourdan 75014 Paris
**GOURSAT Laure**(PSE) :__How Do I Know How Much I Like You? Valuation and v-stability for matching markets with incomplete information__- RésuméIn this project, I set a one-to-one matching market with no transfer and incomplete information - only the current state of the market (matching and realized match utilities) is observed. I wonder: (1) How do agents estimate counterfactual match utilities? (2) How does this estimation affect the market outcome? To address those questions, (1) I define a heuristic termed “valuation” - an agent who considers blocking with a target extrapolates from his current utility and the current utility of the target’s current partner. (2) I characterize v-stable matchings – unblocked matchings when blocking decisions are based on comparisons of valuations. The necessary and sufficient condition for v-stability is “happiness sorting” - any two partners must hold the same within-side rank according to realized match utilities. Working with various standard classes of preferences illustrates that the size and relation to the u-stable set are very changeable. In the general case, the lack of existence of a v-stable matching and the lack of convergence of a dynamic decentralized v-blocking pair process predict persistent moves on the market. I consider several extensions of the valuation heuristic: accounting for the possibility of unmatched agents, allowing extrapolation on the utility’s arguments rather than on the utilities themselves, generalizing to mixed matchings.

**Jeudi 10 mars 2022 12:30-13:30**- Salle R1.15, Campus Jourdan 75014 Paris
**HAGENBACH Jeanne**(Sciences Po) :__Motivated vs. Skeptical Beliefs__**Charlotte Saucet (University Paris 1)**- RésuméThis experiment aims at studying how individuals reconcile the formation of equilibrium beliefs and the formation of motivated beliefs in strategic settings of asymmetric information. Subjects play several rounds of Sender-Receiver disclosure games, in which the Receiver's equilibrium beliefs are skeptical. The experimental treatments affect whether or not the Receiver has intrinsic preferences over what he/she believes. We then vary whether the skeptical beliefs are aligned or not with the Receiver's preferred beliefs. We show that the Receivers' level of skepticism is lower when the skeptical beliefs are self-threatening than when Receivers have no preferences over beliefs. When the skeptical beliefs are self-serving, skepticism is not significantly enhanced compared to the case where Receivers have no preferences over beliefs. We investigate the role of Receivers' IQ and priors on the beliefs they form.

**Jeudi 9 décembre 2021 12:30-13:30**- Salle R1-14 - Campus Jourdan - 75014 PARIS
**PIVATO Marcus**( CY Cergy) :__*__

**Jeudi 2 décembre 2021 12:30-13:30**- Salle R1-14 - Campus Jourdan 75014 PARIS
**TUMENNASAN Norovsambuu**(Dalhousie University) :__One Truth and a Thousand Lies: Defaults and Benchmarks in Mechanism Design__- RésuméWe study a behavioral robust implementation problem in which agents are endowed with benchmark strategies. Benchmarks may arise, for instance, as a consequence of recommendations made by the planner, or through social norms. We introduce (Interim) Behavioral Benchmark Equilibria (BBE) and prove a revelation principle: any social choice function (SCF) implementable in BBE is also fully implementable by a direct revelation mechanism in which agents' benchmark action is truthtelling. Specializing to direct mechanisms, we provide the following results. First, behavioral robust implementation is equivalent to interim rationalizability with iterative elimination of strategies that are dominated by truthtelling. When preferences are interdependent, an SCF is behavioral robust implementable if and only if it satisfies a strengthening of the contraction property, Bergemann and Morris (2009) (BM). We document the restrictions that behavioral agents impose: in a leading example of BM, the maximum possible interdependence in preferences is half the bound identified in BM. Behavioral robust implementation is therefore possible. In addition we show that in any private value environments, behavioral agents impose no further restrictions than those already imposed by robust implementation.

**Jeudi 18 novembre 2021 12:30-13:30**- Room R1-14 - Campus Jourdan and On line
**KORIYAMA Yukio**(CREST) :__The Winner-Take-All dilemma__- RésuméWe consider collective decision making when the society consists of groups endowed with voting weights. Each group chooses an internal rule that specifies the allocation of its weight to the alternatives as a function of its members' preferences. Under fairly general conditions, we show that the winner-take-all rule is a dominant strategy, while the equilibrium is Pareto dominated, highlighting the dilemma structure between optimality for each group and for the whole society. We also develop a technique for asymptotic analysis and show Pareto dominance of the proportional rule. Our numerical computation for the US Electoral College verifies its sensibility.
- Texte intégral [pdf]

**Jeudi 14 octobre 2021 12:30-13:30**- Salle R1-14 - Campus Jourdan 75014 PARIS
**LEBRETON Mael**(PSE) :__Generosity Biases the Learning of Cultural Conventions__- RésuméHuman groups can markedly differ in fairness and cooperation norms, and these differences can create intergroup misunderstandings and conflict. At the same time, humans also trade and travel across cultural divides, suggesting that they can learn and adapt to new culture-specific conventions and rules of engagement. While such adaptions avoid intergroup conflict and benefit intergroup exchange, how humans learn group-specific rules that are often implicit and distinct from already learned values and norms remains poorly understood. Here we examine this fundamental learning process underlying social rule acquisition. We created three populations with different yet unobservable rules of engagement and varied whether or not decisions affected interaction partner outcomes. Participants made bargaining offers to responders from these different populations and could observe whether their offer was accepted or rejected. Participants quickly adapted to group-specific rules in learning environments without social consequences, but were overly generous and ended up misrepresenting what would be acceptable when decisions affected their partner’s outcomes. We propose a computational model, combining Bayesian principles and social preferences, that mechanistically explains how generosity leads to biased sampling, impeded learning, and false beliefs about what offers are deemed acceptable. Results suggest that generosity can induce self-fulfilling beliefs in pro-sociality norms that may help to increase cooperation and reduce conflict between distinct groups but also create inaccurate stereotypes and economic inefficiencies.

**Jeudi 7 octobre 2021 12:30-13:30**- Salle R1.14 - Campus Jourdan 75014 PARIS
**MACÉ Antonin**(PSE) :__*__

**Jeudi 30 septembre 2021 12:30-13:30**- Salle R1-14 - Campus Jourdan - 75014 PARIS
**HERRERA Helios**(Warwick) :__Echo chamber elections__**Ravideep SETHI**- RésuméElections aggregate information of citizens who gather information, in part from mainstream news, but possibly to a large extent from their different echo chambers which filter news in particular ways. We study information aggregation and electoral outcomes in the presence of endogenous echo chambers. Citizens choose their echo chambers in a self-serving way, in part to shield them from the outside world (mainstream news) and thus preserve their political faith as often as possible. The key asymmetry is how isolated citizens are, namely how much they trust/distrust mainstream news to begin with. We study in several contexts the electoral advantage that the side with more isolated followers may have. This advantage can turn into a sure electoral victory regardless of the true state of the world, thereby failing information aggregation, in some instances. Information aggregation still fails if voters can abstain, don’t fully trust their echo chamber sources either, or gain utility from consuming favourable news.

**Jeudi 23 septembre 2021 12:30-13:30**- Salle R1-14 - Campus Jourdan 75014 PARIS
**BLUMENTHAL Benjamin**(ETH Zurich) :__Is More Information Good for Voters?__- RésuméRecent work in the political agency literature shows how being less informed about policy-making can improve voters' welfare. Using a model of targeted spending with homogeneously informed voters, I show how diverse papers analysing a large range of issues in policy-making -- the role of interest groups, the influence of media, fiscal transparency, the effect of non-binding law, the impact of ideology -- rely on the possibility of partial control, partial screening, or both, when voters benefit from less information. Building on this mechanism, I subsequently ask: if voters are heterogeneously informed, is it better to be part of a more informed elite or the less informed masses? How is the masses' welfare affected by the existence of a more informed elite? I show that the answers depend on the elite's ability to transmit verifiable information and the nature of its additional information.

**Jeudi 17 juin 2021 12:30-13:30**- Online
**GLEYZE Simon**(PSE) :__Ideology and Communication__**Co-author: Agathe Pernoud**

**Jeudi 10 juin 2021 12:30-13:30**- Online
**KOESSLER Frédéric**(PSE) :__Mediation and Information Design in Large Games__**Co-authors: Marco Scarsini and Tristan Tomala**- RésuméWe define the notion of Bayes correlated Wardrop equilibrium (BCWE) for general non-atomic games with anonymous players and incomplete information. BCWE describes the set of all outcomes that could arise when a mediator (such as a social planner or a traffic information system) provides information and helps players coordinate their decisions. We show that, whatever the information structure, a (Bayes Wardrop) equilibrium of the corresponding game induces a BCWE. Conversely, a revelation principle holds: every BCWE can be implemented with a direct recommendation system and obedient strategies. If the game admits a convex potential (e.g., traffic routing games or adoption games with negative externalities) and there is complete information, every BCWE is a convex combination of Wardrop equilibria: mediation is essentially irrelevant. If the game does not admit a convex potential, correlated signals can expand the set of equilibrium outcomes and improve welfare, even under complete information. Under incomplete information, every BCWE outcome of a game with a convex potential can be fully implemented: there exists an information structure such that every Bayes Wardrop equilibrium induces that outcome. Finally, we show that BCWE is the appropriate solution concept to study information design in games with a finite set of players when the set of players tends to infinity.

**Jeudi 3 juin 2021 12:30-13:30**- Online
**MINARDI Stefania**(HEC Paris) :__Consumption of Values__**Co-authors: Itzhak Gilboa and Fan Wang**- RésuméConsumption decisions are partly influenced by values and ideologies. Consumers care about global warming, child labor, fair trade, etc. Incorporating values into the consumer's utility function will often violate monotonicity, if consumption hurts values in a way that isn't offset by hedonic benefits. We distinguish between intrinsic and instrumental values, and argue that the former tend to introduce discontinuities near zero. For example, a vegetarian's preferences would be discontinuous near zero amount of animal meat. We axiomatize a utility representation that captures such preferences and discuss the measurability of the degree to which consumers care about such values.
- Texte intégral [pdf]

**Jeudi 27 mai 2021 12:30-13:30**- online
**PATTY Morgan**(LEDa, PSL) :__Top dominance__- RésuméTo deal with crucial issues of iterated elimination of weakly or strictly dominated strategies (IEWDS or IESDS), we propose a new elimination procedure. Our procedure, named iterated elimination of top dominated strategies (IETDS), is based on the new notion of top dominance. It is more consistent than IESDS in a certain sense. Top dominance is more restrictive than weak dominance (and may be more restrictive than strict dominance): it requires weak dominance and strict domination of the strategy on a specific profiles set. Contrary to IESDS, IETDS may reduce the set of Nash equilibria (whilst never expanding it and never eliminating strict Nash equilibria) without the problem of inconsistencies and order dependence faced by IEWDS and IESDS.

**Jeudi 20 mai 2021 12:30-13:30**- Online
**CHARROIN Liza**(Université Paris 1) :__Logrolling affects the relative performance of alternative q-majority rules__- RésuméIt has been argued that simple majority rule is the best decision rule for a committee taking a large number of binary decisions. In addition to a general symmetry condition, the underlying argument assumes that members vote sincerely on each proposal. We argue that the conclusion changes if members engage in logrolling agreements. We propose simple algorithms to predict agreements and voting outcomes under different q-majority rules, and apply them to a large number of randomly generated preference profiles. We find that on average, logrolling improves the performance of unanimity rule, and worsens the performance of majority rule. If the number of decisions to be taken is large enough, unanimity rule outperforms majority rule. We conduct a laboratory experiment to verify whether subjects engage in the predicted agreements, and whether the relative performance of unanimity rule improves as predicted. Predicted agreements occur more often under unanimity rule and when they increase the aggregate payoff, while they are less likely if they are more complex (involve bigger coalitions and bundles of project). On aggregate, logrolling always has a positive impact under unanimity rule while it has a mixed impact under majority rule. We conclude that in the presence of logrolling, greater majority requirements may be desirable.

**Jeudi 6 mai 2021 12:30-13:30**- online
**COMPTE Olivier**(PSE) :__On belief formation and the persistence of superstitions__- RésuméWe propose a belief-formation model where agents attempt to discriminate between two theories, and where the asymmetry in strength between confirming and disconfirming evidence tilts beliefs in favor of theories that generate strong (and possibly rare) confirming evidence and weak (and frequent) disconfirming evidence. In our model, limitations on information processing provide incentives to censor weak evidence, with the consequence that for some discrimination problems, evidence may become mostly one-sided. Sophisticated agents who know the characteristics of the censored data-generating process are not lured by this accumulation of evidence, but less sophisticated ones end up with incorrect beliefs.

**Jeudi 8 avril 2021 12:30-13:30**- online
**BLOCH Francis**(PSE) :__Attack and interception in networks__**Kalyan Chatterjee & Bhaskar Dutta**- RésuméThis paper studies a game of attack and interception in a network, where a single attacker chooses a target and a path, and each node chooses a level of protection. We show that the Nash equilibrium of the game exists and is unique. It involves a mixed strategy of the attacker except when one target has a very high value relative to others. We characterize equilibrium attack paths and attack distributions as a function of the underlying network and target values. We also show that adding a link or increasing the value of a target may harm the attacker - a comparative statics effect which is reminiscent of Braess’s paradox in transportation economics. Finally, we contrast the Nash equilibrium with the equilibria of two variations of the model: one where nodes make sequential protection decisions upon observing the arrival of a suspicious object, and one where all nodes cooperate in defense.

**Jeudi 1er avril 2021 12:30-13:30**- online
**ISPANO Alessandro**(CY Cergy) :__Designing Interrogations__- RésuméWe provide an equilibrium model of interrogations with two-sided asymmetric information. The suspect knows his status as guilty or innocent and the likely strength of law enforcers' evidence, which is informative about the suspect's status and may also disprove lies. We study the evidence strength standards for interrogating and for drawing adverse inferences from silence that minimize prosecution errors. We consider scenarios where interrogations can be delegated. We describe the optimal mechanism under full commitment and a dynamic interrogation with two-sided information revelation implementing the optimum in equilibrium.
- Texte intégral [pdf]

**Jeudi 18 mars 2021 12:30-13:30**- online
**BI Wei**(European University Institute) :__Adverse Selection with Dynamic Learning__- RésuméMost of the current literature about adverse selection in the dynamic game form assumes that the information asymmetry exists ex ante while I build a search model (the seller samples the buyers and the buyers make the offer) to characterize the case where the information asymmetry evolves as time goes on and study the agents' behavior. I find the condition such that the market is efficient in which case all sellers could make a trade whatever their types. Furthermore, I study how the agents' behavior changes when they are bounded rational. I first consider a situation where the coarse buyer forms an expectation of the value from the trade, which is unconditional on whether the seller will accept his offer. I find that he tends to overbid if the information asymmetry exacerbates. Secondly, I consider the coarse buyer expects the seller's behavior as if it was stationary conditional on the state. I find that when the adverse selection effect is significant, he will not make a higher offer compared to the rational one.

**Jeudi 11 mars 2021 12:00-13:00**- online
**MOULIN Hervé**(University of Glasgow) :__Worst Case Voting and Bargaining__**with Anna Bogomolnaia & Ron Holzman**- RésuméThe guarantee of a given anonymous mechanism is the worst case welfare an agent can secure even against unanimously adversarial other agents. The higher the guarantee, the easier the participation in the mechanism. The worst case design question is how high can such guarantee be, and what type of mechanism achieves it? We adress this question in the n-person probabilistic voting/bargaining model. Feasible guarantees include the Uniform lottery over of the p deterministic outcomes, and those implemented by the Random Dictator mechanism or Voting by Veto. Finding the set M(n; p) of maximal (unimprovable) guarantees is simple only if n > p ñit reduces to the Uniform one ñ, and when n = 2 ñit is a polytope where each vertex combines a round of veto with one of random dictatorship ñ. If 3 6 n < p the set M(n; p) is a simplicial complex of dimension d = d p n e, that we describe in detail only when d = 1. In general the Uniform, Veto and Random Dictator guarantees are still the building blocks of 2 d simplices in M(n; p); each of dimension d. The corresponding guarantees are easy to interpret and implement.

**Jeudi 4 mars 2021 12:30-13:30**- online
**QU Xiangyu**(Paris 1) :__Perfect Altruism Breeds Time Consistency__**Co-author : Antoine Billot**- RésuméPublic policies are supposed to be determined by maximization of the social lifetime util- ity. This paper focuses on the general process, aggregation rules and unanimity conditions, which makes these policies socially eligible by individuals through their own discount factors and instantaneous utilities. We show that perfect altruism via an adapted form of unanimity is the key condition helping to characterize a time consistent society concerned with intergen- erational fairness in the presence of individuals who are heterogeneous in discount factors as well as in instantaneous utilities. In addition, different intensity levels of altruism are proved to provide different forms of aggregated social discounting and instantaneous utility, these forms giving birth to several lifetime utilities, from the standard exponential discounted function to the quasi-hyperbolic and the k-hyperbolic ones. Moreover, by showing that the degree of so- cial present-bias can be regulated by the choice of the number of periods involving altruism through unanimity, new insights emerge and potentially overturn some of the most standard economic policy recommendations.

**Jeudi 11 février 2021 12:30-13:30**- online
**VELLODI Nikhil**(PSE) :__Optimal Vaccine Allocation: Spillovers and Incentives__**Co-author: Joshua Weiss**- RésuméWe offer a novel theoretical framework to study optimal vaccination policies. The key features of the model are that agents: 1) differ both in their potential exposure (x) to others and vulnerability (y) to severe illness, 2) exert negative externalities through interaction, and 3) can take voluntary preventative measures, for instance self-isolation. Our main result is a complete characterization of the second-best policy. Three striking features emerge. First, it is non-monotone -- people with intermediate y are vaccinated more than those with either low or high y. Second, it exhibits an exposure premium among those who do not self-isolate -- people with higher x require lower overall risk, xy, to be vaccinated. Third, for those who voluntarily self-isolate, it is invariant to y, depending only upon x. Numerical results demonstrate that policies vaccinating only the most vulnerable perform significantly worse than other simple heuristics, especially when supplies are limited.
- Texte intégral [pdf]

**Jeudi 4 février 2021 12:30-13:30**- online
**GOURSAT Laure**(PSE) :__On why affirmative action may never end and why it should__**Co-author : Philippe Jehiel**- RésuméSuccessive welfare-maximizing governments must decide whether to implement an affirmative action policy. This affirmative action policy is purported to improve the skills distribution of a minority group, although this improvement becomes trivial over time. Employers pay workers according to their expected productivity, but without observing whether they benefited from affirmative action or not. This has a depressing effect on the wages of non-beneficiaries (from either the majority or the minority group), leading in turn to a feeling of injustice. We find that governments perpetually choose to implement an affirmative action policy, even though the overall feeling of injustice is worse than the purported beneficial effect on the skills of the minority group (which becomes marginal over time). This is because each government’s actual policy decision is not observed by market participants and thus has no direct effect on wages. This outcome is in sharp contrast with a first-best, welfare-maximizing policy plan, in which governments choose to end affirmative action after a certain number of periods, when the purported improvement in the skills distribution of the minority group becomes small enough.

**Jeudi 17 décembre 2020 14:00-15:00**- online
**BLUMENTHAL Benjamin**(ETH Zurich) :__Political Agency and Legislative Subsidies with Imperfect Monitoring__- RésuméVoters are frequently ill-equipped to evaluate politicians' actions and can often only imperfectly monitor them: to wit, that a road was built is easy to see, whether shady practices were involved is harder to tell. I consider a stylised two-period political agency model with moral hazard and adverse selection to study the consequences of imperfect monitoring on politicians' behaviour and voters' welfare. I show that imperfect monitoring has an ambiguous effect on voters' welfare and that being unable to monitor politicians perfectly can be welfare improving. I also study how an interest group might subsidise policy making and show that legislative subsidies from an interest group with interests opposed to those of voters can improve voters' welfare. Finally, I endogenize the monitoring process and show that voters' will rationally not monitor politicians perfectly.

**Jeudi 10 décembre 2020 12:30-13:30**- online
**GOURSAT Laure**(PSE) :__Whether and Where to Apply: Competition and coordination on a matching market with common preferences and priorities and Single and Scored Application mechanism.__- RésuméThis paper defines a class of Bayesian games, termed “Application Game”, to model strategic interactions on a two-sided one-to-one agent-object matching market, where preferences and priorities are common and the allocation occurs through the “Single and Scored Application” (SSA) mechanism, equivalently 1-truncated serial dictatorship. The analysis derives the Bayes-Nash equilibria, which amounts to answering the question: “Whether and where to apply?”. It elicits the trade-off between ambition (accepting competition on high-value objects), and pragmatism (seeking coordination on under-demanded objects). A surprising common feature of the multiple equilibria is the block structure, namely agents sort into a finite number of blocks of neighboring priority where they adopt exactly the same application strategy.

**Jeudi 3 décembre 2020 14:00-15:00**- online
**GHERSENGORIN Alexis**(PSE) :__Reactance: a Freedom-Based Theory of Choice__- RésuméProhibitions create the desire they were intended to cure'' (Lawrence Durell). This phenomenon -- often referred to as reactance -- has been recognized for decades by psychologists. Despite numerous economic implications, no theoretical model has been proposed yet. This paper develops axiomatically a revealed preference approach of menu-dependent choices triggered by reactance. The choice model relies on four components: a complete and transitive utilitarian preference; a partition of alternatives into distinct types; a reference set, composed of the alternatives the decision-maker feels entitled to have access to; and a menu-dependent offset function that triggers reactance. The Weak Axiom of Revealed Preference is suitably relaxed. Building on four additional axioms, our representation theorems obtain endogenously the four components along with a uniqueness result. We derive from our representation a preference over menus and argue that our model can be seen as a choice theory counterpart of the freedom of choice literature. Finally, we discuss several applications.

**Jeudi 26 novembre 2020 14:00-15:00****PAHLKE Marieke**(PSE) :__Dynamic Consistency in Ambiguous Persuasion__- RésuméBeauchêne, Li, and Li (2019) show that ambiguous persuasion leads to new interim equilibria with higher expected ex-ante value for the Sender compared to the standard Bayesian persuasion. Ambiguity arises due to a set of communication devices chosen by the Sender. However, in their setting, the equilibrium strategy of the Receiver is in general not ex-ante optimal. This paper shows that we can restrict without loss of generality to messages that produce recommended actions or synonyms of recommended actions. Using this result we can define rectangular beliefs on a general state space. We show that given these rectangular beliefs the Receiver behaves dynamically consistent. Hence, the interim equilibrium of Beauchêne et al. (2019) can be generalized to a Perfect Bayesian Equilibrium.

**Jeudi 19 novembre 2020 12:30-13:30**- online
**AUGIAS Victor**(SciencesPo) :__Wishful Thinking: Persuasion and Polarization__- RésuméHow can wishful thinking impact strategic information disclosure? To investigate this question we model a game of persuasion in which Receiver has a tendency to distort beliefs optimistically in the direction of more pleasant states. In the spirit of Caplin and Leahy (2019), we model wishful thinking as a choice of beliefs: following the reception of information, Receiver optimally chooses its posterior belief by trading off the anticipatory utility she derives from being optimistic as well as the cost of distorting beliefs. We characterize the optimal correspondence between Bayesian and motivated belief and analyze how a rational persuader would choose to disclose information to such an agent, characterizing conditions on the preferences of Receiver that define whether the persuader is better or worse off when facing a wishful thinker vis-à-vis the Bayesian canonical model of Kamenica and Gentzkow (2011). Finally, we extend the model to a setting in which wishful voters with heterogeneous partisan preferences choose through majority voting whether or not to adopt a proposal and show that, if voters preferences are symmetrically distributed around the median voter, optimal public persuasion induces maximum belief polarization. This formalizes a new mechanism for the emergence of polarization which comes as a byproduct of strategic information disclosure to wishful agents.

**Jeudi 12 novembre 2020 12:30-13:30**- salle R1-09, campus Jourdan - 75014 Paris
**SALAMANCA Andrés**(PSE) :__Biased Mediators in Conflict Resolution__- RésuméWhat is the role of biased mediators for reaching negotiated settlements in social conflicts? Previous empirical research in Political Sciences has suggested that mediators are often more effective if they are unbiased (or impartial). This research contributes to the previous debate following a game theoretic analysis. We study a model of cheap-talk in which an agent possesses private information about a binary state of the world. This information is required by an uninformed principal in order to take an action in the real line. Individuals have quadratic preferences, with a difference in their bliss point parameterized by a state-dependent bias parameter. Therefore, a conflict of interests between both parties arises because of a discrepancy in their bliss point. Provided that mediation is beneficial for at least one party, we show that whenever the variation of the bias across states is large enough, the agent will refuse to participate in a mediation process that is biased towards the principal. Otherwise, the mediator’s bias is inconsequential for reaching an agreement, hence a biased mediator is as effective as an unbiased one.

**Jeudi 5 novembre 2020 12:30-13:30**- salle R1-09, campus Jourdan - 75014 Paris
**BLOCH Francis**(PSE) :__Myopic and Farsighted Stable Sets in 2-player Strategic-Form Games__- RésuméThis paper revisits the analysis of stable sets in two-player strategic- form games using recent advances in farsighted stability. Our two main contributions are (i) to establish a connection between myopic stable sets and the stable matchings of an auxiliary two-sided matching problem and (ii) to provide a characterization of singleton farsighted stable sets based on a structural property of 2-player games called “the block partition property.” Our analysis also generalizes and unifies existing results on myopic and farsighted stable sets in 2-player games.

**Jeudi 8 octobre 2020 12:30-13:30**- salle R1-14, campus Jourdan - 75014 Paris
**VELLODI Nikhil**(PSE) :__Competition and Data Privacy on Platforms__**Erik Madsen (NYU)**

**Jeudi 1er octobre 2020 12:30-13:30**- salle R1-14, campus Jourdan - 75014 Paris
**MUN Sofiia**(CES, Université Paris 1 Panthéon-Sorbonne) :__An Econometric Estimation of Prospect Theory for Natural Uncertainty__**Emmanuel Kemel (HEC-Paris)**- RésuméProspect Theory (PT) has been one of the most experimentally studied models for describing behavior under risk. This model also applies to decisions involving natural sources of uncertainty, the context in which probabilities are unknown, and that concerns the large majority of real-life decisions. Surprisingly however, the literature does not propose any elicitation of PT in such context. The paper reports a lab experiment allowing to estimate all the PT parameters in decisions involving a natural source of uncertainty: the participation rate in the 2019 European Parliament elections in the UK. This source is a genuine example of uncertainty inasmuch as is has no objective probability distribution, nor even past frequencies related to similar (Brexit) context. We analyzed the data using structural econometric modeling, that allows to estimate subjective probabilities, jointly with PT components. We also elicited PT parameters for risk, considered as a benchmark, in a within-subject design. The main features of CPT apply under natural uncertainty: attitudes follow the fourfold pattern, and evidence for loss aversion is captured. Additionally, the estimated subjective probabilities give plausible results.

**Jeudi 17 septembre 2020 12:30-13:30**- salle R1-14, campus Jourdan - 75014 Paris
**WEBER Giacomo**(PSE) :__Clustering and Analogy-based Expectation Equilibrium with an Application to Product-differentiated Oligopoly__**Philippe Jehiel**- RésuméWe combine together the notion of Analogy-based Expectation Equilibrium (ABEE) and the K-means clustering algorithm. ABEE is a bounded rationality equilibrium notion that tackles cognitive rationality: players form beliefs bundling games together, the collection of these bundles for a player is called analogy partition. Bundles are primitives in the ABEE original framework; behavior is endogenous. K-means is a machine learning problem of unsupervised learning. K-means algorithm is typically used to identify underlying structure in data, forming classes (clusters). Therefore, data is a primitive; while bundling is endogenous. Both concepts aggregate data. In ABEE, data is about players’ behavior. Our objective is to endogenize the way players bundle games in the ABEE in the spirit of K-means. Thus, both bundles and behavior are endogenous in our model. We provide a mixed extension of a strategic environment to ensure existence of stable analogy partitions. We apply this concept of stability to oligopolistic markets with differentiated products. When firms compete à la Bertrand, we find multiple stable ABEEs. Instead, when firms compete à la Corunot, sometimes stable ABBE can be found only in the mixed extension.

**Jeudi 10 septembre 2020 12:30-13:30**- salle R2-01, campus Jourdan - 75014 Paris
**GLEYZE Simon**(PSE) :__Informationally Simple Incentives__- RésuméWe consider a mechanism design setting in which agents can acquire costly information on their preferences as well as others’. The choice of the mechanism generates informational incentives as it affects what information is acquired before play begins. A mechanism is informationally simple if agents have no incentive to learn about others’ preferences. This property is of interest for two reasons: First, we show that a mechanism is ex-ante strategy-proof only if it is informationally simple. Second, this endogenizes an “independent private value” property of the interim information structure. Our main result is that a mechanism is informationally simple only if it is dictatorial. This holds for generic environments and any smooth cost function which satisfies an Inada condition. Hence even (interim) strategy-proof mechanisms incentivize agents to learn about others and are not ex-ante strategy-proof. Moreover, the Independent Private Value assumption is unlikely to arise endogenously, though we show that endogenous information acquisition resolves the discontinuity associated with full surplus extraction.

**Jeudi 18 juin 2020 14:00-15:00**- online
**RACHIDI Tobias**(University Bonn) :__Sequential Voting and the Weights of Nations__- RésuméThis paper studies the design of voting mechanisms for institutions of representative democracy where representatives participate in the collective decision-making process on behalf of groups of citizens having different sizes. Most of the previous literature focused on the case of two alternatives, whereas I allow for more than two alternatives while assuming that preferences are single-peaked. First, I establish that strategy-proof and surjective social choice functions can be implemented dynamically via the successive voting procedure that is frequently used in practice. Second, I consider preference distributions such that it is optimal for all groups to aggregate preferences within groups according to simple majority voting and characterize the welfare-maximizing mechanism among all strategy-proof and surjective social choice functions for the collective decision-making process involving the representatives. The optimal mechanism takes the form of a weighted successive voting procedure where the majority requirement is monotonically decreasing along the sequence of ballots and the weights of nations exhibit degressive proportionality, that is, the weights are increasing, but the weights per citizen are decreasing in the group size. However, for large group sizes, the degree of overweighting smaller groups relative to the proportional benchmark is lower compared to a square root rule, challenging previous results from the literature. Moreover, the overweighting of smaller groups is larger for more moderate alternatives compared to more extreme alternatives.

**Jeudi 11 juin 2020 14:00-15:00**- Online
**KOESSLER Frédéric**(PSE) :__The Informed Information Designer Problem__**Vasiliki Skreta, UT Austin and UCL**- RésuméAbstract (VERY PRELIMINARY) This paper analyses interim information design as an informed principal problem. A designer is privately informed about the state and chooses an information disclosure mechanism to influence the decisions of multiple interdependent agents. We define a simple solution concept, the strong core mechanism, which is an incentive-compatible mechanism refining the notion of core mechanism and containing the neutral optimum defined in Myerson (1983). A strong core mechanism is not blocked by an alternative mechanism which is incentive compatible for some belief that assigns zero probability to states in which the designer does not benefit from the alternative mechanism. We show that a strong core mechanism always exists, and that it is an equilibrium outcome of the interim information design game. If there is a single agent with binary actions, then an ex-ante optimal allocation mechanism is a strong core mechanism. In more general settings, the solutions of the ex-ante and interim information design problems are different.

**Jeudi 4 juin 2020 14:00-15:00**- online
**SHABAYEK Shaden**(PSE) :__Hidden Opinions__- RésuméI provide a behavioral foundation for opinion polarization, by formalizing an opinion sharing dynamic game. I consider a centrality measure which is inversely related to the number of influence sources of acquaintances. I call this centrality measure local popularity. Facing a social cost, e.g peer pressure or cognitive dissonance or even the possibility of social isolation, individuals with low local popularity optimally choose to hide their opinions and share opinions heard within their social circles. They choose to be consensual. Other individuals, who face a lower social cost due to their high local popularity or expertise, optimally choose to express their opinion. Such individuals who choose to express can update their viewpoint based on the opinions of neighbors who also express; expressing allows players to debate over an issue. Expressers can undergo repulsive influence from a neighbor with ideologically opposed views. When choosing to express or hide, they evaluate the cost related to each action as a function of their network position summarized by their local popularity. Individuals fail to internalize the effect of their chosen binary action on the overall dynamics of opinions in the long-run. In particular, they are myopic. First, I show that opinion updating with those two types of individuals yields a reducible influence system, and I show that opinions converge in the long-run. Second, I characterize two specific patterns of long-run opinions: consensus and polarization. I provide conditions on the network for which the interaction between both types of individuals leads to local consensus and a global polarization of opinions in society. The main contribution of this paper is to provide a behavioral framework to assess long-run opinion polarization, which departs from models of social influence with non-strategic agents.

**Jeudi 14 mai 2020 14:00-15:00**- online
**TERCIEUX Olivier**(PSE) :__Unpaired Kidney Exchange: Overcoming Double Coincidence of Wants without Money__**M. Akbarpour, J. Combe, Y. He, V. Hiller and R. Shimer**- RésuméWe propose a new matching algorithm—Unpaired kidney exchange—to tackle the problem of double coincidence of wants without using money. The fundamental idea is that “memory” can serve as a medium of exchange. In a dynamic matching model with heterogeneous agents, we prove that average waiting time under the Unpaired algorithm is close-to optimal, and substantially less than the standard pairwise and chain exchange algorithms. We evaluate this algorithm using a rich dataset of the kidney patients in France. Counterfactual simulations show that the Unpaired algorithm can match nearly 57% of the patients, with an average waiting time of 424 days (state-of-the-art algorithms match about 31% with an average waiting time of 675 days or more). The optimal algorithm performs only slightly better: it matches 58% of the patients and leads to an average waiting time of 410 days. The Unpaired algorithm confronts two incentive-related practical challenges. We address those challenges via a practical version of the Unpaired algorithm that employs kidneys from the deceased donors waiting list. The practical version can match nearly 87% of patient-donor pairs, while reducing the average waiting time to about 141 days.
- Texte intégral [pdf]

**Jeudi 7 mai 2020 14:00-15:00**- online
**GHOSH Rajarshi**(ESSEC) :__ZOOM. Optimal Identification: An Equilibrium Model of Decision Making under Uncertainty in a Social Context__- RésuméIn the economic literature, there is little behavioral foundation for why and how people use their social identity in decision making. This paper presents an equilibrium model that studies the use of social identity in decision making under uncertainty through the interaction between the agent and her social context. The uncertainty in the decision-making process stems from the fact that agents only have access to noisy private data about their ability type. Agents learn from experience whether they will consider data related to their social identity when they process this private data into an estimate of the probability of success of a task. With the use of their social identity, agent's aim to maximize the likelihood that they will make the welfare-maximizing choice, given the possible realizations of their private data. The model shows therefore that agents use their social identity to reduce the uncertainty in decision making, which shows that choice behavior can be driven by the agents' observable characteristics, even when these observable characteristics have no direct effect on the agents' utility. I derive the conditions under which a social identity becomes salient and show that agents choose the identity from the set of salient identities that best reduces uncertainty given the cost of adoption of this identity. Through a selection, population and minority effect, the agents' perception of the social context is self-reinforcing and both asymmetric and symmetric population equilibria co-exist. Finally, I analyze the implications for labor market outcomes and I derive the necessary demand-side conditions for effective affirmative action policy.

**Jeudi 23 avril 2020 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**MUN Soffia**(PSE) :__ANNULE__

**Jeudi 16 avril 2020 14:00-15:00**- online
**HAGENBACH Jeanne**(Sciences Po) :__Selective Memory of a Psychological Agent__**Frédéric Koessler**- RésuméWe consider a single psychological agent whose utility depends on his action, the state of the world but also the belief that he holds about that state. The agent is initially informed about the state. Before action, he decides which states to remember and which ones to forget. We model the memory selection process by a multi-self game in which the informed first self discloses information to the uninformed second self with identical preferences. While it can be that perfect recall does not occur in equilibrium, we identify broad categories of psychological utility functions in which it does. We next add the possibility of an exogenous probability of forgetting and examine how it changes selective memory.

**Jeudi 9 avril 2020 16:00-17:00**- online
**CASELLA Alessandra**(Columbia) :__Mediating Conflict in the Lab__**Evan Friedman and Manuel Perez**- RésuméAccording to the theory of mechanism design, the presence of a mediator can strictly improve the chances for peace between two contestants. What is striking is that the result follows even when the mediator is less informed than the two parties and has no enforcement power. We test the theory in a lab experiment where two subjects negotiate how to share a resource. In case of conflict, the subjects' privately known strength determines their payoff. The subjects send cheap talk messages about their strength to one another or to the mediator, before making their demands or receiving the mediator's recommendations. We find that, in line with the theory, messages are significantly more sincere when sent to the mediator. However, contrary to the theory, peaceful resolution is not more frequent, even when the mediator is a computer implementing the optimal mediation program. While the theoretical result refers to the best (i.e. most peaceful) equilibrium under mediation, multiple equilibria exist, and the best equilibrium is particularly vulnerable to small deviations from full truthfulness.

**Jeudi 2 avril 2020 14:00-15:00**- en ligne
**COLO Philippe**(PSE) :__Expert-based scientific knowledge__- RésuméThis paper studies the strategic foundation of scientific knowledge when it is entirely expert-based. Scientific theories are models of reality which can be formalised as probability distributions over possible scenarios. An informed sender is assumed to know the most likely model, but cannot prove it to be true. A receiver is in a situation of model-uncertainty and has smooth ambiguity preferences. Communication is a cheap-talk game over models. The receiver’s knowledge is built on the sender’s advice and is thus vulnerable to strategic misalignment. I show that when the ext-post optimal action of the sender is lower (higher) than the one of the receiver, the larger ambiguity aversion, the easier (harder) for the sender to be credible. Under the assumption that the receiver has maxmin expected utility preferences, I show that incentives for information withholding disappear. All equilibria of this game can be ranked by informativeness, making the most informative one interim Pareto dominant and the most credible equilibrium prediction. These results shed light on the decisive role of ambiguity aversion in the foundation of expert-based scientific knowledge.

**Jeudi 19 mars 2020 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**CASELLA Alessandra**(Columbia) :__ANNULE. Mediating Conflict in the Lab__**Evan Friedman and Manuel Perez**- RésuméAccording to the theory of mechanism design, the presence of a mediator can strictly improve the chances for peace between two contestants. What is striking is that the result follows even when the mediator is less informed than the two parties and has no enforcement power. We test the theory in a lab experiment where two subjects negotiate how to share a resource. In case of conflict, the subjects' privately known strength determines their payoff. The subjects send cheap talk messages about their strength to one another or to the mediator, before making their demands or receiving the mediator's recommendations. We find that, in line with the theory, messages are significantly more sincere when sent to the mediator. However, contrary to the theory, peaceful resolution is not more frequent, even when the mediator is a computer implementing the optimal mediation program. While the theoretical result refers to the best (i.e. most peaceful) equilibrium under mediation, multiple equilibria exist, and the best equilibrium is particularly vulnerable to small deviations from full truthfulness and from full rationality.

**Jeudi 12 mars 2020 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**LETROUIT Lucie**(Université Gustave Eiffel) :__Why can't we be friends? An evolutionary approach to the emergence of ethno-cultural hierarchies__- RésuméEconomic behaviors are deeply infl uenced by social norms and, in particular, ethnocultural hierarchies. That's why understanding how such hierarchies emerge and persist through individual interactions in multi-cultural societies is of crucial importance for economists. This paper proposes a three-group (one majority and two ethno-cultural minorities) evolutionary game model of the emergence of ethno-cultural hierarchies in the context of a typical western country where one social group is much larger than the others. From a methodological point of view, this model's originality stems from the multiple groups and strategies implied, which renders the use of graph theory necessary. The model's main results are the following: (1) in accordance with the empirical literature on the subject, ethno-cultural hierarchy views tend to be shared by most individuals within and across groups, (2) minority-di fferentiating ethno-cultural hierarchies, where di fferent minorities are associated with di fferent social statuses, are evolutionarily very resilient and often persist in the long run, instead of more egalitarian ones, because they play a divide-and-rule" role that prevents minorities from allying and jointly demanding more equality, (3) old minorities in a country benefi t from the arrival of a new minority in the long run, if this new minority is culturally close enough to the majority, either directly through status improvement or indirectly through the social recognition of this new minority.

**Jeudi 5 mars 2020 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**STERN Lennart**(PSE) :__Funding Global Public Good Institutions via taxes on aviation emissions- a comparison of three proposed mechanisms__- RésuméThe EU is considering taking unilateral action to achieve higher carbon pricing for international aviation. It has been proposed that the EU could start a club mechanism whereby participating countries would have to tax the emissions of all outgoing flights as well as all flights coming in from non-participants at a fixed rate. They would then have to allocate a cer- tain proportion of the tax revenue to some agreed-upon global institution for climate change mitigation. I propose an alternative club mechanism in which participating countries could allocate their tax revenue to Global Public Good Institutions (GPGIs) of their choice from a set of GPGIs recognised as eligible in the treaty establishing the mechanism. Of the tax revenue that participants decide to not allocate, they could retain a fixed proportion. The remaining revenue would be added to the GPGIs, in proportion to the aggregate allocations made to them. I consider an example where the set of eligible GPGIs consists of 2 GPGIs for global health and 4 for climate change mitigation. I estimate the payoffs that countries would derive from money being added to the different GPGIs, taking into account the heterogeneous benefits generated by each GPGI, the rents that some GPGIs create and the redistributional effects that some GPGIs induce by changing e.g. the oil price. Using an iterated best response algorithm and assuming that initially only the EU and China participate, I predict that this new mechanism can achieve 22% more wel- fare gains than the previously proposed one, raising 50% more revenue for GPGIs. I then enrich the mechanism by allowing countries to also give to so-called “Proportional Matching Funds”, which distribute their budget to subsets of GPGIs, in proportion to the direct allocations made to these GPGIs. This enriched mechanism can achieve 67% more welfare gains than the previously proposed one, raising 100% more revenue for GPGIs.
- Texte intégral [pdf]

**Jeudi 19 décembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**GLEYZE Simon**(PSE) :__Segregation and Expectation Formation on the Returns to Schooling__**Philippe Jehiel PSE/UCL**

**Jeudi 5 décembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**POMMEY Guillaume**(PSE) :__Partnership Dissolution with Cash-Constrained Agents__- RésuméWhen partnerships come to an end, partners must find a way to efficiently reallocate the commonly owned assets to those who value them the most. This requires that the aforementioned members possess enough financial resources to buy out others' shares. I investigate ex post efficient partnership dissolution when agents are ex post cash constrained. I derive necessary and sufficient conditions for ex post efficient partnership dissolution with Bayesian (resp. dominant strategy) incentive compatible, interim individually rational, ex post (resp. ex ante) budget balanced and ex post cash-constrained mechanisms. Ex post efficient dissolution is more likely to be feasible when agents with low (resp. large) cash resources own more (resp. less) initial ownership rights. Furthermore, I propose a simple auction to implement the optimal mechanism. Finally, I investigate second-best mechanisms when cash constraints are such that ex post efficient dissolution is not attainable.
- Texte intégral [pdf]

**Jeudi 21 novembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**NUNEZ Mathias**(CREST - Ecole Polytechnique) :__A solution to the two-person implementation problem__**Jean-François Laslier, Remzi Sanver**- RésuméWe propose a solution to the classical problem of Hurwicz and Schmeidler [1978] and Maskin [1999] according to which, in two-person societies, no Pareto efficient rule is Nash-implementable. To this end, we consider implementation through mechanisms that are deterministic-in-equilibrium while lotteries are allowed off-equilibrium. For strict preferences over alternatives and under a very weak condition for extending preferences over lotteries, we build simple veto mechanisms that Nash implement a class of Pareto efficient social choice rules called Pareto-and-veto rules. Moreover, under mild richness conditions on the domain of preferences over lotteries, any Pareto efficient Nashimplementable rule is a Pareto-and-veto rule and hence is implementable through one of our simple veto mechanisms.

**Jeudi 14 novembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**OZKES Ali**(WU Vienna) :__The effects of strategic environment and communication on cooperation__**Nobuyuki Hanaki (ISER, Osaka)**- RésuméWe study the interaction of the effects of the strategic environment and communication on the levels of cooperation observed in lab experiments on two-person finitely repeated games with a Pareto-inefficient Nash equilibrium. We replicate previous findings that point to higher levels of cooperation under complementarity compared to substitution. We demonstrate that this effect is not due to differences in levels of reciprocity or strategic thinking. Through a simple reinforcement learning model, we find that slow or insensitive learning might drive this effect. We then show that when subjects are allowed to communicate in free-form before making choices, cooperation levels increase significantly but the strategic environment effect disappears. We investigate how communication works towards cooperation with a novel content analysis. In particular, we employ a machine-learning based natural language processing approach to identify how the effect of communication is dependent on the strategic environment.

**Jeudi 7 novembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**PEI Harry**(Northwestern University) :__Trust and Betrayals: Reputational Payoffs and Behaviors without Commitment__- RésuméI introduce a reputation model in which all types of the reputation-building player are facing lack-of-commitment problems. I study a repeatedtrust gamein which a patient player (e.g., a seller) wants to win thetrust of some myopic opponents (e.g., buyers) but can strictly benefit from betraying them. Her benefit frombetrayal is her persistent private information. I provide a tractable formula for the highest equilibrium payoff forevery type of that patient player. Interestingly, incomplete information affects this payoff only through the lowestbenefit in the support of the prior belief. In every equilibrium that attains this highest payoff, the patient player’sbehavior depends nontrivially on past play. I establish bounds on her long-run action frequencies that apply toall of her equilibrium best replies. These features of her behavior are essential for her to extract information rentwhile preserving her informational advantage. I construct a class of such high-payoff equilibria in which thepatient player’s reputation depends only on the number of times she has betrayed as well as the number of timesshe has been trustworthy in the past. The method I developed is useful for studying other repeated incompleteinformation games between a patient informed player and a sequence of myopic uninformed players.
- Texte intégral [pdf]

**Jeudi 24 octobre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**GHERSENGORIN Alexis**(PSE) :__Revealed Preference Formation__**N. Boissonnet and S. Gleyze**- RésuméWe propose a theory of preference formation together with conditions on choice data that identify our model. We introduce non-behavioral objects which causes DM’s preferences and we describe how DM can rationally act on these causes. Our contribution is to provide rigorous and testable foundations for building theories of endogenous preferences, evolving attention, etc. Specifically, we consider a model of reason-based choice in which DM’s behavior is motivated by some properties she thinks are relevant for decision making. Alternatives are compared with respect to how well their properties satisfy DM’s motivations. Preferences are evolving because whenever she becomes aware of a set of properties, DM can decide to make them relevant or irrelevant for her future choices. We identify when this decision is based on the maximization of a meta-preference, implying that preference formation is deliberate. We resolve the underdetermination associated with (deterministic) theories of reason-based choice by exploiting the intertemporal structure of preference changes.

**Jeudi 10 octobre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**WANG Olivier**(NYU Stern) :__Decision Making under Time Pressure__- RésuméThis paper studies individual decision making when deadlines are random. Since the quality of any decision relies on information and it takes time to gather information, a decision maker should have a preference over deadlines as well as over menus. Past research studies rationally inattentive decision making without deadlines. It is established that a decision maker’s informational constraint is revealed by her distaste for contingent planning. This paper extends the analysis to random deadlines and establishes the relationship between a decision maker’s preference over timed choice problems and the set of information acquisition paths she has available. It is demonstrated that if a decision maker’s preference over random deadlines satisfies the von Neumann-Morgenstern independence axiom, then it is as if the decision maker’s optimal way to acquire information depends only on the menu she is presented with and is independent of the deadline. Moreover, if no information is lost along any path then the decision maker’s distaste for contingent planning becomes weaker as she is allowed more time to decide.
- Texte intégral [pdf]

**Jeudi 3 octobre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**COLO Philippe**(PSE) :__Advising against inefficiency__- RésuméI examine a setting where two imperfectly informed senders use cheap talk communication to sequentially advise receivers taking part to a game of contribution to a public bad. The first sender aims at maximising social welfare, the second at increasing inefficiency. Because the receivers' game is inefficient, the first sender is biased towards the outcome of the receivers' game. I show that no truthful revelation is possible and that the second sender always lowers social welfare. In particular, because of the second sender, the first sender can be better off not maximising information transmission.

**Jeudi 19 septembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**PAHLKE Marieke**(PSE) :__Dynamic Consistency in Incomplete Information Games with Multiple Priors__- RésuméThis paper explores multi-stage incomplete information games with common ambiguous information about state or types and ambiguity averse players. We characterize a belief formation process that allows players to take their knowledge about the structure of the game into account. This process leads to subjective rectangular ex-ante belief sets for all players. We show that under these belief sets players behave dynamically consistent. Therefore, using our belief formation process, we can generalize the concept of Sequential Equilibria to multi-stage ambiguous incomplete information games and show existence. Furthermore, we show that ambiguity can introduce Sequential Equilibria, that do not exist without ambiguity.

**Jeudi 12 septembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**COPIC Jernej**:__Constrained Efficiency and Applications__- RésuméThe presentation will be about the concept of constrained efficiency and applications to market settings, such as bilateral trade and auctions under incomplete information. In bilateral trade, where a buyer and a seller have private cost and valuation over an indivisible good, I show that there exist constrained efficient market equilibrium outcomes that exhibit money burning: a wedge between the receipt of the seller and the price to the buyer where the remaining surplus must be disposed of. Previously, it has been presumed that no ex ante or ex post constrained efficient allocation exhibits such waste and that the traditional parametrization due to Myerson and Satterthwaite (1983) by probability of trade and expected payment from the buyer to the seller is sufficient to capture all market equilibrium outcomes. Such a wedge between prices cannot be represented by a lower probability of trade and the general (canonical) representation of trading outcomes is by probability of trade and personalized prices. Consequently, characterizations of market equilibrium mechanisms by posted prices as exhaustive, Hagerty and Rogerson (1987), and as constrained efficient, Copic and Ponsatí Obiols (2016), are not true without the additional assumption that there is no such wedge. Canonical representation has implications for mechanism design, bargaining, and markets. In the presentation, I will provide specific examples where a market equilibrium mechanism with a wedge in prices is optimal for ex ante welfare, and where it solves the problem of a non-benevolent broker. Time permitting, the presentation will include an application to auctions.
- Texte intégral [pdf]

**Jeudi 5 septembre 2019 12:30-13:30**- salle R2-20, campus Jourdan - 75014 Paris
**RACHIDI Tobias**(PSE) :__Optimal Voting Mechanisms on Generalized Single-Peaked Domains__- RésuméI study a private-values voting environment with arbitrarily many voters and alternatives. Voters have generalized single-peaked preferences that give rise to median spaces as introduced in Nehring and Puppe 2007. These preferences allow to go beyond preferences which are single-peaked on a line. For generalized single-peaked preference domains that give rise to median spaces, I characterize the voting rules that maximize ex-ante utilitarian welfare among all mechanisms satisfying strategy-proofness, anonymity and surjectivity. Typically, the welfare-maximizing voting mechanisms involve qualified majority requirements. Moreover, a trade-off between welfare maximization and ex-post Pareto efficiency arises; this tension is not present when preferences are single-peaked on a line. I illustrate my optimality results by means of several applications including collective choice when preferences are single-peaked on a tree as well as voting over multiple public goods under constraints.

**Lundi 12 août 2019 12:30-13:30**- *

**Jeudi 13 juin 2019 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan, 75014 Paris
**BLOCH Francis**(PSE) :__Hiding in networks__**Bhaskar Dutta and Marcin Dziubinsky**

**Mardi 28 mai 2019 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan, 75014 Paris
**KE Shaowei**(China Europe International Business School (CEIBS).) :__Behavioral neural networks__- RésuméWe provide an axiomatic foundation for a class of neural network models applied to decision making under risk, called the neural-network expected utility (NNEU) model. We weaken the independence axiom from expected utility theory in a new way consistent with many experimental findings. We show how to use simple neurons, referred to as behavioral neurons, in the NNEU model to capture behavioral biases such as reference dependence, certainty effect, overweighting of small probabilities, etc. Empirically, we find that using standard estimation approaches, the NNEU model usually has small training error but large testing error due to overfitting. However, by requiring the neural network of the NNEU model to carry behavioral neurons, its performance can be significantly improved.

**Jeudi 23 mai 2019 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan, 75014 Paris
**BIZET Romain**(Mines ParisTech) :__Keep calm and carry on: A theory of governmental crisis communication__**Pierre Fleckinger (Mines ParisTech and PSE)**- RésuméThis paper develops a model of strategic communication which questions how informed governments can credibly and effectively disclose warning information to exposed populations in the wake of major disasters. Credible public communication strategies take the form of severity scales whose optimal design hinges on the divergence between individual and social preferences, typically embodied by the existence of external effects of individual responses. Private and committed modes of communication are also investigated and shown to increase expected welfare. The paper also considers combinations of crisis communication with disaster management strategies. In particular, when ex post transfers are available - e.g. through public relief funds or emergency response plans - it is optimal for the government to subsidize pro-social actions and to provide partial public insurance when responses entail negative externalities. These transfers facilitate communication and improve response. Several policy aspects and applications of the model are discussed.
- Texte intégral [pdf]

**Jeudi 9 mai 2019 12:30-13:30**- salle R2-01, campus Jourdan - 48 bd Jourdan, 75014 Paris
**BOBTCHEFF Catherine**(PSE) :__On the social value of (not) disclosing negative results__**Raphaël Levy (HEC) et Thomas Mariotti (TSE)**

**Jeudi 9 mai 2019 11:25-12:25****CHATTERJEE Kalyan**(Penn State) :__Repeated Coordination with Private Learning__

**Jeudi 18 avril 2019 12:30-13:30**- salle R2-01, campus Jourdan - 48 bd Jourdan, 75014 Paris
**GALICHON Alfred**(New York University) :__The equilibrium flow problem__**Larry Samuelson (Yale) and Lucas Vernet (Sciences Po)**- RésuméWe introduce the 'equilibrium flow problem' as an extension of the classical min-cost flow problem. Applications to trade on networks, demand theory, two-sided matching with and without transfers, and dynamic programming are provided. Existence and structural properties are discussed, as well as a regularized version of the problem.

**Jeudi 11 avril 2019 12:30-13:30**- salle R2-01 Campus Jourdan - 48 bd Jourdan, 75014 Paris
**LEVIN Dan**(Ohio State University) :__Misbehavior in Common Value Auction__**Jim Peck (Ohio State University)**- RésuméWe study optimal misbehavior by an auctioneer (shill bidding) or by (rings of) bidders in two auction formats in a pure CV environment. Specifically, we consider the dynamic English (EA), and the static Sophi auction, that are outcome equivalent without misbehavior. We first characterize the optimal misbehavior strategy of the ring or the auctioneer, and then compare/rank those formats by their immunity to such misbehavior. The recent theoretical and experimental literature documented and explained the observation that often dynamic auctions outperform their “equivalent” static (one shot) strategic implementation. A main result is that the static version is more immune to several forms of misbehavior, which might be one reason why it still flourishes.

**Mardi 9 avril 2019 14:00-17:00**- salle R1-14, campus Jourdan - 48 bd Jourdan, 75014 Paris
**LEVIN Dan**(Ohio State University) :__Séminaire Exceptionnel avec Dan Levin__- RésuméOVERVIEW of TALK: Dan Levin will review a large share of his experimental (and theory) work on common-value (CV) auctions, also known as “mineral-right auctions.” He will provide a wide background to non-experts to motivate the interest in CV auctions. This includes: adverse selection and the difficulty of (at least student) bidders to cope with it. What is meant by the Winner’s Curse (WC), by theorists and by experimentalists? Why do we care about bidders’ overbidding and losses? Are there ways to avoid the WC? Is the adverse selection the main cause of the WC? Do subjects learn with experience and what do they learn? What is the role information complexity in the auction environment and subjects’ heterogeneity in creating the WC? He will also discuss two recent papers on the topic (bolded below). References and useful links. http://www.econ.ohio-state.edu/levin/ Common Value Auctions and the Winner’s Curse, 2002. Princeton University Press, (with J. H. Kagel) Could we overcome the Winner’s Curse by (Behavioral) Auction Design? 2019 WP (With Philipp Reiss). An Experimental Study of Estimation and Bidding in Common-Value Auctions with Public Information, 2019. Journal of Economic Theory, 179(1) pp. 73-98. (With Gary Charness and David Schmeidler).

**Jeudi 4 avril 2019 12:30-13:30**- salle R2-01, campus Jourdan - 48 bd Jourdan 75014 Paris
**XING Zhaojun**(Bielefeld/Paris 1/PSE) :__Voronoi Language Games with Knightian Uncertainty__- Texte intégral [pdf]

**Jeudi 21 mars 2019 12:30-13:30**- salle R2-01 capus Jourdan - 48 bd Jourdan, 75014 Paris
**GLEYZE Simon**(PSE) :__Informationally Simple Implementation__**Agathe Pernoud (Stanford University)**

**Jeudi 14 mars 2019 12:30-13:30**- salle R2-01 campus Jourdan - 48 bd Jourdan, 75014 Paris
**LEDUC Mathieu**(PSE) :__Endogenous fragility in complex economic systems__**Matt Elliott (Cambridge University) and Ben Golub (Harvard University)**

**Jeudi 21 février 2019 12:30-13:30**- salle R1-09, campus Jourdan - 48 bd Jourdan, 75014 Paris
**MOMOT Ruslan**(HEC) :__User Privacy in Platforms__**Itay P. Fainmesser (Johns Hopkins Carey Business School), Andrea Galeotti (London Business School)**- Texte intégral [pdf]

**Jeudi 20 décembre 2018 12:30-14:00**- salle R2-01, campus Jourdan - 48 bd Jourdan 75014 Paris
**COMBE Julien**(UCL) :__The Design of Teacher Assignment: Theory and Evidence__**Olivier Tercieux (CNRS & PSE) and Camille Terrier (HEC Lausanne)**- RésuméThe reassignment of teachers to schools is a central issue in education policies. In several countries, this assignment is managed by a central administration that faces a key constraint: ensuring that teachers obtain an assignment that they weakly prefer to their current position. To satisfy this constraint a variation on the Deferred Acceptance (DA) mechanism of Gale and Shapley (1962) has been proposed in the literature and used in practice---for example, in the assignment of French teachers to schools. We show that this mechanism fails to be efficient in a strong sense: we can reassign teachers in a way that i) makes them better-off and ii) better fulfills the administration's objectives represented by the priority rankings of the schools. To address this weakness, we characterize the class of mechanisms that do not suffer from this efficiency loss and elicit a set of strategy-proof mechanisms within this class. To empirically assess the extent of potential gains associated with the adoption of our mechanisms, we use a rich dataset on teachers' applications for transfers in France. These empirical results confirm both the poor performance of the modified DA mechanism and the significant improvements that our alternative mechanisms deliver in terms of teachers' mobility and administration's objectives.
- Texte intégral [pdf]

**Jeudi 6 décembre 2018 12:30-13:30**- salle R2-01, campus Jourdan - 48 bd Jourdan, 75014 Paris
**LAMBERT-MOGILIANSKY Ariane**(PSE) :__Phishing for (quantum-like) Phools: Theory and evidence.__**V. I. Danilov (Russian Academy of Sciences), A. Calmettes (Scuola Superiore Sant'Anna) and H. Gonay (GetQuanty)**- RésuméThe presentation will be based on two articles: 1. Targeting in quantum persuasion problems JME (2018) joint with V.I Danilov Abstract: In this paper we investigate the potential for persuasion arising from the quantum indeterminacy of a decision-maker's beliefs, a feature that has been proposed as a formal expression of well-known cognitive limitations. We focus on a situation where an agent called Sender only has few opportunities to influence the decision-maker called Receiver. We do not address the full persuasion problem but restrict attention to a simpler one that we call targeting, i.e. inducing a specific belief state. The analysis is developed within the frame of a n-dimensional Hilbert space model. We find that when the prior is known, Sender can induce a targeted belief with a probability of at least 1/n when using two sequential measurements. This figure climbs to 1/2 when both the target and the belief are known pure states. A main insight from the analysis is that a well-designed strategy of distraction can be used as a first step to confuse Receiver. We thus find that distraction rather than the provision of relevant arguments is an effective means to achieve persuasion. We provide an example from political decision-making. 2. The power of distraction: An experimental test of quantum persuasion (2018) joint with A. Calmettes, H. Gonay forthcoming in LNCS Springer. Abstract: Quantum-like decision theory is by now a well-developed field. We here test the predictions of an application of this approach to persuasion as developed by Danilov and Lambert-Mogiliansky in (Danilov and Lambert-Mogiliansky 2018). One remarkable result entails that in contrast to Bayesian predictions, distraction rather than relevant information has a powerful potential to influence decision-making. We conducted an experiment in the context of donations to NGOs active in the protection of endangered species.We first tested the quantum incompatibility of two perspectives 'trust' and 'urgency' in a separate experiment. We next recruited 1371 respondents and divided them into three groups: a control group, a first treatment group and the main treatment group. Our main result is that 'distracting' information significantly affected decision-making: it induced a switch in respondents' choice as to which project to support compared with the control group. The first treatment group which was provided with compatible information exhibited no difference compared with the control group. Population variables play no role suggesting that quantum-like indeterminacy may indeed be a basic regularity of the mind. We thus find support for the thesis that the manipulability of people's decision-making is linked to the quantum indeterminacy of their subjective representations (mental pictures) of the choice alternatives.

**Jeudi 22 novembre 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan, 75014 Paris
**TALLON Jean-Marc**(PSE) :__Tailored recommendations__

**Jeudi 8 novembre 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan, 75014 Paris
**LEVY John**:__Concerned with careers? Career development, task assignment, and labour market competition__**Heski Bar Isaac (Rotman School of Management)**- RésuméFirms have discretion over the kinds of task to which they assign workers. This affects workers' opportunities to advance their careers and, consequently, their efforts. Both firms' task allocation and workers' effort decisions interact through the extent of labour market competition: more competition increases the return to effort, but decreases firms' incentives to assign workers to informative tasks. Competition can therefore increase or decrease firm profits, worker payoffs, and welfare. One consequence is that firms may choose strategies that lead to increased competition. When the employee pool is heterogeneous, firms might choose different strategies that attract different kinds of workers and differentiate themselves through the career opportunities that they offer.

**Jeudi 11 octobre 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**PEREZ Lucas**(PSE) :__Information Design with Agency__**Jacopo Bizzotto (University of Oslo) and Adrien Vigier (Norwegian Business School)**- RésuméWe study the problem of a principal who relies on an agent for the production of public information destined to the players of a continuation game. The agent has access to a status quo procedure to generate information. The principal is given the opportunity to design a better procedure, but her design is constrained by the status quo in two ways: First, the new procedure must use the same set of signals as the status quo procedure; Second, the agent incurs a cost of switching to the new procedure. The principal can reward the agent with monetary transfers up to a limited liability constraint, but we assume that procedures are not contractible, only signals are. The principal therefore faces a problem of information design with agency in which she must trade off informativeness about the states of the world to influence the continuation game with informativeness about the choice of the agent to reduce the cost of agency. We provide a general methodology for solving these problems and characterize optimal implementable procedures. We examine comparative statics with respect to switching cost. Finally, we apply our results to information acquisition, leak prevention and persuasion examples.

**Jeudi 4 octobre 2018 12:30-12:30**- salle R2-20, campus Jourdn - 48 bd Jourdan, 75014 Paris
**DWORCZAK Piotr**(Northwestern University) :__The Simple Economics of Optimal Persuasion__**Giorgio Martini (Stanford GSB)**- RésuméConsider a Bayesian persuasion problem in which the Sender’s preferences depend only on the mean of posterior beliefs. We show that there exists a price schedule for posterior means such that the Sender’s problem becomes a consumer-like choice problem: The Sender purchases posterior means using the prior distribution as her endowment. Prices are determined in equilibrium of a Walrasian economy with the Sender as the only consumer and a single firm that has the technology to garble the state. Welfare theorems provide a verification tool for optimality of a persuasion scheme, and characterize the structure of prices that support the optimal solution. This price-theoretic approach yields a tractable solution method for persuasion problems with infinite state spaces. As an application, we provide a necessary and sufficient condition for optimality of a monotone partitional signal. We show that the approach extends to competition in persuasion and persuasion problems with no restrictions on Sender’s utility.
- Texte intégral [pdf]

**Jeudi 20 septembre 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**KOESSLER Frédéric**(PSE) :__Cheap talk with partial language competence__**Jeanne Hagenbach (Sciences Po)**

**Jeudi 6 septembre 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**MUIR Ellen**(Standford University ) :__Equilibrium Adjustment Dynamics__**Simon Loertscher (University of Melbourne)**- RésuméAbstract: We study asset markets in which agents with private information about their values arrive over time. Each agent is endowed with some units of the asset and has a commonly known maximum demand. As a function of her value and the values of all others, each agent optimally either buys, sells, or does not trade. We show that the equilibrium under the socially optimal mechanism involves phases of market imbalances during which sellers, or less frequently buyers, are on the long side. Calling a market a buyers' (sellers') market when the terms of trade favor buyers (sellers), we show that buyers (sellers) being on the short side is neither necessary nor sufficient for a buyers' (sellers') market. The optimal mechanism involves (partial) price rigidity insofar as some traders' prices do not change even as other traders' allocations do. The optimal adjustment dynamics to out-of-equilibrium states involve fire trades -- the immediate liquidation of trades whose level exceeds the socially optimal storage threshold -- followed by slow and incremental adjustments to a typical state.

**Jeudi 14 juin 2018 12:30-13:30****La séance a été annulée.**- salle R2-20, campus Jourdan - 75014 Paris
**COUANAU Quentin**(PSE) :__TBA__

**Jeudi 7 juin 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan, 75014 Paris
**COLO Philippe**(PSE) :__Cheap Talk under Ambiguity: The IPCC in Climate-Change Agreements.__- RésuméAbstract : The Intergovernmental Panel on Climate Change (IPCC) has gained a central position on the report of scientific knowledge on climate change. This paper shows that the IPCC can take advantage of it to mitigate the inefficiency of the level of greenhouse gas emissions in international environmental agreements (IEA). I model an IEA as a game of contribution to a public bad. In making confidence statements over global warming predictions, the IPCC plays a cheap talk game with IEA participants. Under the assumption that players are maxmin expected utility maximisers I show that there is a unique sequential equilibrium in this game. This result backs up the idea that the IPCC could play a regulatory part in the management of climate change. In addition, it sheds new light on the role of strategic communication under ambiguity, supporting the view that it can contribute to restoring economic efficiency.

**Jeudi 31 mai 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014, Paris
**ERLANSON Albin**(Stockholm School of Economics) :__Costly Verification in Collective Decisions.__**Andreas Kleiner**- RésuméWe study how a principal should optimally choose between implementing a new policy and maintaining the status quo when the information relevant for the decision is privately held by agents. Agents are strategic in revealing their information, but the principal can verify an agent’s information at a given cost. We exclude monetary transfers. When is it worthwhile for the principal to incur the cost and learn an agent’s information? We characterize the mechanism that maximizes the expected utility of the principal. The evidence is verified whenever it is decisive for the principal’s decision. This mechanism can be implemented as a weighted majority voting rule, where agents are given additional weight if they provide evidence for their information. The evidence is verified whenever it is decisive for the principal’s decision. Additionally, we find a general equivalence between Bayesian and ex-post incentive compatible mechanisms in this setting. Finally, we extend our analysis of optimal mechanisms to imperfect verification.

**Jeudi 24 mai 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**GORELKINA Olga**(University of Liverpool) :__Collusion via Information Sharing and Optimal Auctions__

**Jeudi 17 mai 2018 12:30-13:30**- salle R1-15, campus Jourdan - 48 bd Jourdan 75014 Paris
**BOUACIDA Elias**(PSE) :__Consistency of Choice Correspondences: a Theoretical and an Empirical Study.__- RésuméExperiments in economics generally elicit a choice function: for a given choice set, a decision maker chooses exactly one alternative. It is a restriction on most models of revealed preferences, that start with a choice correspondence: for a given choice set, a decision maker chooses a subset of alternatives. This paper explores the impact of this restriction on the formulation of revealed preferences axioms and on the consistency of observed behaviors. In order to do so, a new methodology is built to elicit a choice correspondence in an incentive compatible way, called pay-for-certainty. We introduce a new theoretical requirements, increasing chosen set that guarantees the elicitation of a proper choice correspondence with the pay-for-certainty methodology. We implement the methodology in the laboratory and compare it to choice functions. 40.13% of observed choices are singletons and only 2.33% of subjects have single-peaked preferences. Increasing chosen set is verified by 60.47% of the subjects in the relevant case. 46.51% of all choices correspondences verify WARP whereas 58.82% of all choice functions do. In a second step, we use the pay-for-certainty methodology to explore the completeness of preferences assumptions, a task impossible with choice functions. We assess degrees of incomplete preferences. A lower bound introduced by Eliaz and Ok (2006) explains only 2.68% of observed behaviors. We characterized and upper bound. It is equivalent to the decision maker maximizing a strict partial preference. It explains 14.77% of observed behaviors. These results points to the relevance of observing choice correspondences instead of choice functions when a choice correspondence is needed to test a theory. .

**Jeudi 3 mai 2018 12:30-13:30**- salle R1-13, campus Jourdan - 48 bd Jourdan 75014 Paris
**GOSSNER Olivier**(CNRS and X and LSE) :__Dynamic Bank Runs.__- RésuméWe study a model of dynamic bank runs under private information. We characterize a unique equilibrium in threshold strategies, discuss the informational and fundamental determinants of bank stability and compare outcomes in the dynamic model to its static counterpart.

**Jeudi 12 avril 2018 12:30-13:30**- salle R1-13, campus Jourdan - 48 bd Jourdan 75014 Paris
**RINCON ZAPATERO Juan Pablo**(Université Carlos III Madrid) :__Recursive Utility and Thompson Aggregators (joint avec R. Becker)__**R. Becker**- RésuméWe reconsider the theory of Thompson aggregators proposed by Mari- nacci and Montruccio. First, we prove a variant of their Recovery Theorem estabilishing the existence of extremal solutions to the Koopmans equa- tion. Then, under more restrictive conditions, we demonstrate there is a unique solution to the Koopmans equation. Our proof is based on concave operator techniques as first developed by Kransosels’kii. This differs from Marinacci and Montruccio’s proof as well as proofs given by Martins de la Rocha and Vailakis.

**Jeudi 5 avril 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**HA-HUY Thai**(Université de Cergy ) :__A Not so Myopic Axiomatization of Discounting__**Jean-Pierre Drugeon**- Résumé«This article builds an axiomatization of inter-temporal trade-offs that makes an explicit account of the distant future and thus encompasses motives related to sustainability, transmission to offsprings and altruism. The focus is on separable representations and the approach is completed following a decision-theory index based approach that is applied to infinite dimension streams. This enlightens the limits of the commonly used flat tail intensity requesites for the evaluation of utility sequences: in this article, these are supersed and replaced by an axiomatic approach to optimal myopy degrees that in its turn precedes the determination of optimal discount. The overall approach is anchored on the new and explicit proof of a temporal decomposition of the preference orders between the distant future and the close future itself directly anchored on the determination of the optimal myopia degrees. The argument is shown to provide a novel understanding of temporal biases with the scope for a distant future bias when the finite dimensional gets influenced by the infinite dimensional. The reference to robust orders and pessimism-like axioms finally allows for determining tractable representations for the indexes.»

**Jeudi 29 mars 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**SANKTJOHANSER Anna**(TSE) :__Optimally Stubborn__- RésuméI consider a bargaining game with two types of players - rational and stubborn. Rational players choose demands at each point in time. Stubborn players are restricted to choose from the set of ``insistent'' strategies that always make the same demand and never accept anything less. However, their initial choice of demand is unrestricted. I characterize the equilibria in this game. Relative to the case with exogenous behavioral types, strong behavioral predictions emerge: in the limit, players randomize over at most two demands. However, unlike in a world with exogenous types, there is Folk theorem like payoff multiplicity.

**Jeudi 22 mars 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**BAUMANN Léonie**(University of Cambridge ) :__Identifying the best agent in a network.__- RésuméThis paper develops a mechanism for a principal to allocate a prize to the most valued agent when agents have a knowledge network. The principal does not know any agent's value but any two linked agents know each other's values. Agents compete for the prize and send costless private messages about their own value and the values of others they know to the principal. Agents can lie only to a certain extent and only lie if it increases their chances of winning the prize. A mechanism that determines each agent's chances of winning for any possible message profile is proposed. We show that with this mechanism, there exists an equilibrium such that the most valued agent wins with certainty if every agent has at least one link; if the network is a star or complete, then the most valued agent wins with certainty in every equilibrium.

**Jeudi 15 mars 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 PARIS
**KOPYLOV Igor**(University of California, Irvine) :__Combinatorial Subjective State Spaces__- RésuméI construct subjective state spaces S for preferences over finite menus. The additive representation of Kreps (1979) is relaxed to a weaker model called coherent aggregation. This model improves the identification of subjective states in several ways. First, the minimal size of S can be specified explicitly in terms of preferences. It allows combinatorial applications: starting from monotonic preferences over menus that have at most k elements, one can identify subjective state spaces that have up to k states. Second, coherent aggregation can be non-monotonic and hence, accommodate preferences for commitment. The case of singleton S delivers Gul and Pesendorfer's (2005) model of changing tastes. Third, the coherent aggregation model has equivalent interpretations in terms of incomplete dominance relations and choice functions that are both induced}endogenously by preferences. The induced dominance has a Pareto representation with subjective states. The induced choice function C is rationalized via strict maximization of subjective states. The path-independence of C characterizes the case where all subjective states are linear orders, as in Aizerman and Malishevski (1981).

**Jeudi 8 mars 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**ZHAO Wei**(HEC Paris) :__Optimal bridge players among separated networks.__- RésuméAbstract: This paper studies the optimal bridge problem among multiple separated separated networks and examines optimal targeting of players, groups and central planner. We find that the optimal choice (new links connect with other groups) of the group coincides with the individual’s rational choice. But, the best choice for the central planner may be different from the individual’s rational choice. We have shown how player’s centrality is going to be affected due to combination of separated networks, and how this is related to the self-loops and Bonacich centrality of these separated networks. We construct an exact index to identify the key bridge players linking up whom will mostly increase aggregate equilibrium effort and aggregate equilibrium welfare, and we find that the key bridge players may consist of neither each group’s key players nor central players. We provides its implications on company merge, optimal targeting, community merge and network design.

**Jeudi 1er mars 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**HA-HUY Thai**(Université de Cergy ) :__A Not so Myopic Axiomatization of Discounting__**Jean-Pierre Drugeon**- RésuméThis article builds an axiomatization of inter-temporal trade-offs that makes an explicit account of the distant future and thus encompasses motives related to sustainability, transmission to offsprings and altruism. The focus is on separable representations and the approach is completed following a decision-theory index based approach that is applied to infinite dimension streams. This enlightens the limits of the commonly used flat tail intensity requesites for the evaluation of utility sequences: in this article, these are supersed and replaced by an axiomatic approach to optimal myopy degrees that in its turn precedes the determination of optimal discount. The overall approach is anchored on the new and explicit proof of a temporal decomposition of the preference orders between the distant future and the close future itself directly anchored on the determination of the optimal myopia degrees. The argument is shown to provide a novel understanding of temporal biases with the scope for a distant future bias when the finite dimensional gets influenced by the infinite dimensional. The reference to robust orders and pessimism-like axioms finally allows for determining tractable representations for the indexes.

**Jeudi 15 février 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**CHONÉ Philippe**(ENSAE) :__Partial exclusivity__**joint with Laurent Linnemer and Thibaud Vergé**- RésuméThis papers offers a new rationale for exclusive agreements. Long-term business partners have a common interest to agree on an option to deal on an exclusive basis, while keeping the possibility to revert to a competitive process (auction) should the surplus of an internal deal turn out to be low. Partial exclusivity occurs in equilibrium in the sense that competition takes place with positive probability. Although there is no rent extraction à la Aghion and Bolton (1987), such lock-up agreements harm competitors (and total welfare) by depriving them of business opportunities. Compared to Bulow and Klemperer (1996), the possibility of long-term contracting calls into question the efficacy of auctions relative to negotiations

**Jeudi 8 février 2018 12:30-13:30**- salle R1-15, campus Jourdan - 48 bd Jourdan 75014 Paris
**MACÉ Antonin**(PSE) :__On the Weights of Sovereign Nations__**Rafael Treibich**- RésuméWe study the design of voting rules for committees representing heterogeneous groups (countries, states, districts) when participation in the committee is voluntary. While efficiency recommends weighting groups proportionally to their stakes, we show that accounting for participation constraints entails overweighting some groups, those for which the incentive to participate is the lowest. When collective decisions are not enforceable, cooperation is enabled by the satisfaction of more stringent constraints, that may require granting a veto power to certain countries. The model has important implications for the problem of apportionment, the allocation of voting weights to countries of varying populations, where it provides a rationale for setting a minimum representation for smaller countries.

**Jeudi 1er février 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**LARAKI Rida**(CNRS et Lamsade) :__Acyclic Gambling Games__**J. Renault**- RésuméWe consider two-player zero-sum stochastic games where each player controls his own state variable living in a compact metric space. The terminology of the title comes from gambling problems where the state of a player represents its wealth in a casino. Under natural assumptions (such as continuous running payoff and non expansive transitions), we consider for each discount factor the value $v_\lambda$ of the $\lambda$-discounted stochastic game and investigate its limit when $\lambda$ goes to 0 (players are more and more patient). We show that under a strong acyclicity condition, the limit exists and is characterized as the unique solution of a system of functional equations: the limit is the unique continuous excessive and depressive function such that each player, if his opponent does not move, can reach the zone when the current payoff is at least as good than the limit value, without degrading the limit value. The approach generalizes and provides a new viewpoint on the Mertens-Zamir system coming from the study of zero-sum repeated games with lack of information on both sides. A counterexample shows that under a slightly weaker notion of acyclicity, convergence of $(v_\lambda)$ may fail.

**Jeudi 25 janvier 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**BANERJEE Abhijit**(MIT) :__The Efficient Deployment of Police Resources: Theory and New Evidence from a Randomized Drunk Driving Crackdown in India__**Esther Duflo and Dan Keniston**

**Jeudi 18 janvier 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan, 75014 Paris
**PRADELSKI Bary**(ETH Zurich) :__Title: Micro influence and macro dynamics of opinion formation__**Bernhard Clemm von Hohenberg and Michael Maes**- RésuméSession organisée par Francis Bloch (Paris 1/PSE) Abstract: Social media platforms, comment boards, and online market places have created unprecedented potential for social influence and resulting opinion dynamics such as polarization. We propose an encompassing model that captures competing micro-level theories of social influence. Conducting an online lab-in-the-field experiment, we observe that individual opinions shift linearly towards the mean of others' opinions. From this finding, we predict the macro-level opinion dynamics resulting from social influence. We test our predictions using data from another lab-in-the-field experiment and find that opinion polarization decreases in the presence of social influence. We corroborate these findings with large-scale field data.

**Jeudi 11 janvier 2018 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**VEIEL Rafael**:__Strategic Type Spaces__- RésuméUniversal type spaces provide a representation of players’ information that allows to derive the set of interim correlated rationalizable (ICR) actions in any game. They also provide the appropriate structure for the study of ICR sensitivity to information. However, the resulting procedure to compute the set of ICR actions in a given game is often unnecessarily complex, as the universal type space contains types that are redundant for this game. Furthermore, no representation equivalent to the universal type space is known for alternative solution concepts such as interim independent rationalizability (IIR). We offer a way to study the set of rationalizable (in the sense of ICR or IIR, or other) actions in any given game. We introduce strategic type spaces as minimal representations of all strategically relevant information that allow to derive the set of rationalizable actions in this game. These strategic type spaces are characterized by a network structure, and provide an adequate framework for studying sensitivity of strategic behavior to information.

**Jeudi 14 décembre 2017 12:30-13:30**- salle R2-20, campus Jourdan - 48 bd Jourdan 75014 Paris
**ELLISON Sara**(MIT) :__Match Quality, Search, and the Internet Market for Used Books__**joint with S. Ellison**- RésuméAbstract : This paper examines the effect of the Internet on markets in which match-quality is important, including an analysis of the market for used books. A model in which sellers of unusual objects wait for high-value buyers to arrive brings out match quality and competition eects through which improved search technologies may increase both price dispersion and social welfare. A reduced-form empirical analysis nds support for a number of more nuanced predictions of the model in the context of the used book market, exploiting both cross-sectional dierences across books and time-series dierences in the wake of Amazon's acquisition and incorporation of a large used book marketplace. The paper develops a framework for structural estimation of a model based on the theory. The estimates suggest that the shift to Internet sales substantially increased both seller prots and consumer Surplus.

**Jeudi 7 décembre 2017 12:30-13:30**- salle R2-20, campus Jourdan, 48 bd Jourdan - 75014 Paris
**CARAYOL Nicolas**(Université de Bordeaux ) :__Evaluating the underlying qualities of items and raters from a series of reviews__**joint with M. Jackson**

**Jeudi 23 novembre 2017 12:30-13:30**- salle R1-13, campus Jourdan, 48 bd Jourdan - 75014 Paris
**PREVET Antoine**(PSE/Université Paris 1) :__Information Management by a Budget constrained Principal__- RésuméThis paper contributes to the debate on transparency in the public sector by considering one of its major features: a limited budget. In a principal-agent model with moral hazard and bounds on transfers, I study an information design problem where the principal can either choose to be transparent and fully reveal information before the agent takes action, or remain opaque. On the one hand, transparency makes it possible for the principal to engage her limited budget only when the expected gain is worth the implementation cost. On the other hand, opacity does not allow for tailor-made contracts but rather ensures optimal incentives. I show that transparency is more likely to be optimal for the principal when the task is less valuable and the budget is lower. Furthermore, the optimal information structure is derived, requiring that the agent be told only what action to undertake.

**Jeudi 16 novembre 2017 12:30-13:30**- salle R2-20, campus Jourdan, 48 bd Jourdan - 75014 Paris
**MOULIN Hervé**(University of Glasgow) :__Fair Mixture of Public Outcomes__**joint with H. Aziz and A. Bogomolnaia**- RésuméAbstract : We revisit probabilistic voting under dichotomous preferences where, as on Facebook, agents report that they like/dislike every outcome. We focus on two often conflicting normative concerns. Minorities should not be crushed by the opposing majority; and the size of the support for a given outcome should increase its weight: numbers matter. The Unanimous Fair Shares property adresses both concerns by giving to any group of like-minded agents an influence proportional to its size. Individual Fair Shares weakens UFS by only guaranteeing to each agent a 1/n-th influence on the outcome (where n is the total number of agents). Given that we cannot combine Efficiency, Incentive Compatibility (Strate- gyproofness) and Unanimous, or even Individual, Fair Shares, we propose second best mechanisms achieving two of these three design goals. The Conditional Utilitarian rule, a simple variant of the classic random dictator, is Strategyproof and guarantees Unanimous Fair Shares. It is much easier to compute and more efficient than the familiar Random Priority rule. In numerical simulations its inefficiency is consistently low. The efficient Egalitarian rule guarantees Individual Fair Shares and is Excludable Strategyproof: this weakens Strategyproofness by ensuring that an agent is excluded from consuming those public outcomes she reportedly dislikes. But numbers do not matter: the rule treats a unanimous group of agents exactly as if it contains a single agent. The efficient Nash rule (maximizing the product of utilities) offers much stronger welfare guarantees than both above, in particular it achieves core stability if any group of agents can enforce any outcome with a probability proportional to its size. But the rule fails even the excludable form of Strategyproofness. We also uncover several challenging open questions
- Texte intégral [pdf]

**Jeudi 2 novembre 2017 12:30-13:30**- salle R2-20, campus Jourdan, 48 bv Jourdan - 75014 Paris
**BERVOETS Sébastien**(CNRS/GREQAM ) :__Doping and competition uncertainty__

**Jeudi 19 octobre 2017 12:30-13:30**- salle R2-20, campus Jourdan, 48 bd Jourdan - 75014 Paris
**LOVO Stefano**(HEC) :__Herding and Crowdfunding (joint with Brett Green).__- RésuméAbstract: We explore a sequential fund-raising mechanism with common values and an all-or-nothing clause (i.e., a target level of funds must be raised in order for the project to be implemented). We show that in the maximum truth-telling equilibria the presence of the all-or nothing clause decreases the probability of abstention cascade but it increases the probability of pledging cascades. Compared to the socially optimal outcome these campaign lead to an excessive implementation of projects. Social optimum can be restored by allowing backers to opt in and out at the end of the campaign.

**Jeudi 5 octobre 2017 13:00-14:00**- salle R1-11, campus Jourdan - 48 boulevard Jourdan, 75014 Paris
**ELLISON Sara**(MIT) :__Costs of Managerial Attention and Activity as a Source of Sticky Prices: Structural Estimates from an Online Market__- RésuméAbstract: We study price dynamics for computer components sold on a price-comparison website. Our fine-grained data—a year of hourly price data for scores of rival retailers—allow us to estimate a dynamic model of competition, backing out structural estimates of managerial frictions. The estimated frictions are substantial, concentrated in the act of monitoring market conditions rather than entering a new price. We use our model to simulate the counterfactual gains from automated price setting and other managerial changes. Coupled with supporting reduced-form statistical evidence, our analysis provides a window into the process of managerial price setting and the microfoundation of pricing inertia, issues of growing interest in industrial organization and macroeconomics. JEL Codes: L11, C73, D21, L81
- Texte intégral [pdf]

**Jeudi 28 septembre 2017 13:00-14:00**- salle R1-11, campus Jourdan - 48 bd Jourdan, 75014 Paris
**RIVERA Thomas**(HEC) :__Robust Mechanisms for the Regulation of Bank Risk__- RésuméThis paper aims at designing mechanisms for the regulation of bank risk (probability of default) that can guarantee robustness to large misspecifications of the regulators information regarding the bank's assets. Assuming banks can (on average) discern the true level of risk of the other regulated banks, I construct a robust mechanism that allows the regulator to bound the worst case probability of bank failure by any arbitrary amount. In doing so, I show that any informationally robust mechanism must necessarily require banks to issue subordinated debt to other regulated banks and must provide a way to guarantee a minimal probability of joint bank failure between at least two of the banks. We show how the regulator can achieve the latter objective by providing a single bank with an explicit guarantee against losses in the event that another bank fails and how, when coupled with a minimum subordinated debt requirement and interest rate ceiling, such a guarantee can ensure an arbitrarily low probability of any regulated bank's failure. Finally, I discuss how the results are affected when banks have biased estimations of other banks' risks and the best bound on bank risk that the regulator can achieve given such biases.
- Texte intégral [pdf]

**Jeudi 21 septembre 2017 12:00-13:00**- salle R1-13, campus Jourdan, 48 boulevard Jourdan - 750145 Paris
**BOBTCHEFF Catherine**(PSE) :__Insurance Pools for Undiversifiable Risks__**David Alary (TSE) and Carole Haritchabalet (Université de Pau)**- RésuméAbstract : This paper discusses the decision of the European Commission not to renew the antitrust exemption for the setting up of syndicates in the insurance industry. Pools are constituted to provide insurance for undiversifiable and/or new risks for which insurers with private expertise are capacity constrained. Our objective is to study if such syndicates improve insurance supply. Organizing this supply amounts to sharing a common value divisible good between capacity constrained and privately informed agents with a reserve price. Pools turn out to operate as a uniform price auction with an “exit/re-entry” option that we compare to a discriminatory auction where no specific agreements are needed. Both auction formats lead to different coverage/premium tradeoffs. If at least one insurer provides an optimistic expertise, the pool offers both lower premiums and higher coverage. This result is reversed when all insurers are pessimistic about the risk. Static comparative results with respect to capacity constraints and reserve price are provided

**Jeudi 29 juin 2017 13:00-14:00**- salle R2-01, campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**BOUACIDA Elias**(PSE) :__Indifference or Indecision, an Experimental Investigation using Choice Correspondences__- RésuméAbstract. The completeness axiom is a staple in economics, despite its lack of both: 1. Normative appeal, as pointed by Aumann (1962); 2. Descriptive investigation, apart from two experiments using lotteries by Danan and Ziegelmeyer (2006) and Cettolin and Riedl (2016). In this paper, I am tackling the second question. I show how to elicit a choice correspondence -- in contrast to a choice function, as usually done in experiments -- in an incentive compatible way, while keeping the revealed preferences methodology. This allows the elicitation of -- some -- indifference and -- some -- indecision (aka incompleteness) in experiments using Eliaz and Ok (2006)'s characterization, which provides a lower bound on indecision. This procedure is applied to the choice between four incentivized tasks to be performed in the laboratory. I observe some indifference for a majority of subjects and some indecision for around a quarter of them.

**Jeudi 22 juin 2017 13:00-14:00**- salle R1-11, campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**HE Junnan**(Sciences Po) :__Strategic Mistakes: Implications for Market Design Research__**Yeon-Koo Che (Columbia) and Georgy Artemov (Melbourne**- RésuméAbstract. A field data from Australian college admissions shows that a non-negligible fraction of applicants choose strategies (or rank-ordered lists) that are unambiguously dominated, but that the majority of these mistakes are payoff irrelevant. In keeping with this result, we develop a theory suggesting that the presence of such mistakes jeopardizes the identication method based on truthful reporting hypothesis under a (seemingly) strategy-proof mechanism, but leaves the method based on weaker stability condition relatively unscathed. Monte Carlo simulation further confirms this point and quantifies the differences between these two methods in the structural estimation of preference parameters and in a hypothetical counterfactual analysis.

**Jeudi 18 mai 2017 13:00-14:00**- salle R1-11, campus Jourdan, 48 boulevard Jourdan, Paris 14e
**PEJSACHOWICZ Leonardo**(Paris 1) :__Breadth versus Depth.__- RésuméAbstract. We consider a fundamental trade-off in search: when choosing between multiple unknown alternatives, is it better to learn a little about all of them (breadth) or a lot about a single one (depth)? In choice settings where a distribution is exogenous, we find that breadth is optimal for ``small'' problems and that depth is optimal for ``large'' ones. But in IO settings, where firms endogenously choose distributions, we find breadth to be always optimal. Finally, we consider extensions to fat-tails and correlation, and find that in these extensions, breadth is superior.

**Jeudi 20 avril 2017 13:00-14:00**- campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**BASTIANELLO Lorenzo**(Paris 1) :__Target-based solutions for Nash bargaining__

**Jeudi 6 avril 2017 13:00-14:00**- salle F, bâtiment G, campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**BIRO Peter**( (Hungarian Academy of Science)) :__Stable Project Allocation under Distributional Constraints__**joint with Agoston and Szanto**- RésuméAbstract. In a two-sided matching market when agents on both sides have preferences the stability of the solution is typically the most important requirement. However, we may also face some distributional constraints with regard to the minimum number of assignees or the distribution of the assignees according to their types. These two kind of requirements can be challenging to reconcile in practice. Our research is motivated by two real applications, a project allocation problem and a workshop assignment problem, both involving some distributional constraints. We used integer programming techniques to nd reasonably good solutions with regard to the stability and the distributional constraints. Our approach can be useful in a variety of di erent applications, such as resident allocation with lower quotas, or controlled school choice.

**Jeudi 30 mars 2017 13:00-14:00**- Salle R1-11, Campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**STEPANOV Sergey**(Higher School of Economics ) :__Reputation and information aggregation.__

**Jeudi 23 mars 2017 13:00-14:00**- Salle R1-11, Campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**WOODWARD Kyle**(University of North Carolina ) :__Equilibrium and Approximation in Auction-Like Models__- RésuméI define auction-like models with continuum actions to satisfy a number of conditions which are met in common auction contexts, when action spaces are appropriately constrained. In these models, I prove that there exists a monotone pure-strategy Bayesian-Nash euqilibrium. The proof of equilibrium existence constructs equilibrium as a limit of equilibria of nearby discretized models, which implies that utility-relevant observable functions of actions are converging in probability to the continuum-action limit. The conditions defining auction- like models are intuitive: first, bidder utility cannot be adversely affected by small upward deviations; second, with small deviations bidders can guarantee themselves nearly the utility at a limit of strategies in the limit of the deviated strategies; third, if one bidder is discontinuously worse off in a limit of strategies, some other bidder is occasionally better off. Under mild additional conditions, the action space constraints are irrelevant, and the limiting strategies are an equilibrium in the unconstrained continuum-action model. I use these results to prove the existence of pure-strategy equilibria in divisible-good pay-as-bid auctions with private information, heterogeneous agents, and generic decreasing value functions. Equilibrium approximation implies that the distribution of observed allocations and seller revenue in discrete auctions may be close to the distribution of these outcomes predicted by the divisible-good model.

**Jeudi 16 mars 2017 13:00-14:00**- Salle R1-11, Campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**JEHIEL Philippe**(PSE, ENCP) :__Investigation with forgetful liars__- RésuméAbstract: I consider environments in which an individual not telling the truth does not remember which lie he made and I propose modelling how such an individual forms a belief about his past lie using the apparatus of the analogy-based expectation equilibrium so that after performing a lie the individual considers that he lied according to the aggregate distribution of lie performed in equilibrium across all possible observations that the individual may possible make. I study the extent to which how asking subjects several times about their information and committing to harsh punishments in case of contradiction allows to elicit information and implement the first-best ouctome. Various extensions allowing for a mix of rational and forgetful liars are considered.

**Jeudi 9 mars 2017 13:00-14:00**- salle R1-11, Campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**GOMES RENATO**(TSE) :__Drip Prices and Missed Sales__**joint with TIROLE Jean**- RésuméAbstract: Firms often sell a basic good as well as ancillary ones. Hold-up concerns have led to regulations on ancillary good pricing such as price transparency and caps on the add-on price. The hold-up narrative, however, is an inaccurate description of many retail settings in which add-ons are offered below cost (e.g. free shipping), and of the infrequent card surcharging in countries where the “no-surcharge rule” was lifted. We argue that the key to unifying these conflicting narratives is the seller’s concern about losing sales on the basic good. When the ancillary good is self-supplied (or equivalently supplied by a competitive upstream ancillary-good industry) and consumers are well-informed about prices (attentive or repeat customers), the firm passes through the cost of the ancillary good to the consumer if the (endogenous) markup on the basic good is small, but absorbs partly or fully any ancillary good cost increase if this markup is high, so as not to run the risk of losing sales. The ancillary good is always sold below cost. When the ancillary good is provided by a supplier with market power, the latter has an incentive to jack up the intermediate price so as to benefit from the seller’s cost absorption strategy. We show that the optimal regulation is a price floor on the ancillary good, equal to the intermediate price. We also show that absent regulation, there is an overprovision of the ancillary good when the ancillary good is supplied by a two-sided platform. When consumers are initially unaware of the ancillary good price, the price of the basic good acts as a signal of the ancillary good price: A high price for the basic good makes the firm more wary of missed sales and thus reassures the consumer about the ancillary good price. We study the optimal regulation

**Jeudi 2 mars 2017 13:00-14:00**- salle DSS, bâtiment B, Campus Jourdan, 75014 paris
**TOMALA Tristan**(HEC Paris) :__Competitive Information Design. Work in progress.__**joint with Frédéric Koessler (PSE) and Marie Laclau (PSE).**- RésuméAbstract. We study games between n information designers, each of whom can perform a statistical experiment about a piece of information, the pieces being independent. They aim at persuading a decision-maker to take their most favorable action. For such games with discontinuous payoffs, we show that there exists a (sub-game perfect) equilibrium with either an infinite number of messages or randomization over finite statistical experiments. We characterize the equilibrium distributions of actions for rectangular games in which the optimization problem of the decision-maker is separable across designers. Rectangular games have a (sub-game perfect) equilibrium in pure strategies with a finite number of messages.

**Jeudi 2 février 2017 13:00-14:00****La séance a été annulée.****BOUACIDA Elias**(PSE) :__*__

**Jeudi 19 janvier 2017 13:00-14:00****TALLON Jean-Marc**(PSE) :__Asset Allocation and Trade under Heterogeneous Ambiguity Aversion__**with S. Mukerji and H. Ozsoylev**- RésuméAbstract: We seek to explore several financial puzzles in a single model, that departs from basic portfolio choice models only in that it accomodates ambiguity and ambiguity sensitve preferences. The three anomalies are (1) the asset allocation puzzle wich is the observation that professionnals' recommandations for portfolio holdings do not accord well with the mutual fund theorem, (2) departures from CAPM pricing in a systematic way and (3) large volume of trade after public annoucement of earnings. We propose a model where investors have smooth ambiguity preferences and asset returns have ambiguous means. We show that investors optimally hold portfolio with a different mix of risky assets, due to their heterogeneous ambiguity aversion. This heterogeneity also leads, in an equilibrium model, to trade upon the arrival of public information (about the dividends).

**Jeudi 12 janvier 2017 13:00-14:00****FLECKINGER Pierre**(Ecole des Mines - PSE) :__Game of Frauds__**T. Lafay and C. Monnier**

**Jeudi 15 décembre 2016 13:00-14:00****COUANAU Quentin**(PSE) :__Cooperation versus competition under ambiguity aversion__- RésuméAbstract : We study a principal-agent model with two agents and two positively correlated tasks. We assume that agents face ambiguity (or knightian uncertainty) regarding the tasks and are averse to ambiguity. Our focus is on whether the presence of ambiguity aversion favors cooperation or competition between agents, as compared to risk aversion. We show that the effect of ambiguity is twofold. It lowers the power of incentives making agents' wages less sensible to performance, while it increases significantly the cost of implementing relative incentives. As a consequence, the principal does not necessarily choose relative incentives, even in situations where they would always be optimal under risk. Allowing for agents to cooperate, we show that ambiguity aversion favors cooperation between agents while risk aversion favors competition.

**Jeudi 8 décembre 2016 12:30-13:30**- Salle F, Bât G
**HAGENBACH Jeanne**(Sciences Po) :__Communication with Evidence in the Lab__**E. Perez-Richet**- RésuméWe study communication with evidence in the lab. Our experimental design involves a collection of sender-receiver games with various payoffs and permits partial disclosure. We use local and global properties of the sender's incentive graph to uncover behavioral regularities and explain performance across games. Sender types whose interests are aligned with those of the receiver fully disclose, while sender types whose interests are not aligned with those of the receiver remain silent or partially disclose. When partially disclosing senders mostly disclose favorable pieces of evidence and hide unfavorable ones. But the cognitive cost of partial disclosure, as measured by response times, is higher for both senders and receivers. Receivers take evidence into account and tend to be skeptical about vague messages in games whose graph is acyclic. They perform better in acyclic games, whereas senders perform better in cyclic games.

**Jeudi 1er décembre 2016 13:00-14:00**- Matching with Ownership
- RésuméAbstract : We consider a hybrid model at the intersection between the standard two-sided matching market as proposed by Gale and Shapley (1962) and a housing market as proposed by Shapley and Scarf (1974). Two types of agents have to be matched to a common side of objects. Agents of one type have preferences that depend on the object they are matched to but also which agent of the other type is matched to this object. To investigate the link with the two-sided matching literature, we define a natural concept of ownership structure that determines for each possible triplet of match, which agent owns the object. We then define a pairwise stability property of matchings with respect to an ownership structure i.e. where only an owner could ask another agent to join his current object. We first show that there are ownerships structures for which no such matching exists. However, we show that it exists for one-sided ownership structures i.e. when agents of one type are always the owners. In this framework, we show that contrary to a standard two-sided matching market, there may not be any core matching even if stable matchings exist. The existence is restored if we assume that agents of the side always owning the objects have lexicographic or separable preferences. To investigate the link with a housing market model, we define an initial allocation of the objects to one agent and define the standard notion of core in this framework. Even with one sided initial allocation, we show that the latter may also be empty. However, with such initial allocations, we show that there always exists a Pareto-efficient matching that cannot be blocked by coalitions of size two.

**Jeudi 17 novembre 2016 13:00-14:00****HERINGS Jean-Jacques**(Maastricht University) :__Equilibrium and Matching under Price Controls__- RésuméThe paper considers a one-to-one matching with contracts model in the presence of price controls. This set-up contains two important streams in the matching litera- ture, those with and those without monetary transfers, as special cases and allows for intermediate cases with some restrictions on the monetary transfers that are feasible. An adjustment process that ends with a stable outcome is presented, thereby prov- ing the existence of stable outcomes. The process contains the deferred acceptance algorithm of Gale and Shapley (1962) and the approximate auction mechanism of Demange, Gale, and Sotomayor (1986) as special cases. The paper presents a no- tion of competitive equilibrium, called Dreze equilibrium, for this class of models, an extension of the concept as developed by Dreze (1975) for economies with divisible commodities subject to price controls. It is shown that Dreze equilibrium allocations are equivalent to allocations induced by stable outcomes. One implication is the ex- istence of Dreze equilibria. Another implication is the equivalence of a competitive equilibrium concept and the concept of stable outcomes that is valid with and with- out monetary transfers as well as when monetary transfers are limited.
- Texte intégral [pdf]

**Jeudi 10 novembre 2016 13:00-14:00****SALAMANCA Andrés**(PSE) :__The value of mediated communication__- RésuméAbstract: Kamenica and Gentzkow (2011) consider a model in which a sender chooses a public communication device for signaling his information to an uninformed receiver, who then takes an action that aects the welfare of both individuals. In their model, the sender is fully committed to truthfully communicate the signal to the receiver, so that they abstract from incentive compatibility issues. By considering mediated communication, we provide an analytical framework overcoming this overly restrictive assumption. Specically, we are able to characterize incentive constraints by a set of linear inequalities, which allows to formulate the sender's problem as a linear programming problem. As a result, we can use an alternative geometric approach based on duality theory to transform the sender's problem into a simplied problem without incentive constraints that can be solved using concavication

**Jeudi 3 novembre 2016 13:00-14:00****ISPANO Alessandro**(CY Cergy) :__Competitive pricing and quality disclosure to cursed consumers__- RésuméAbstract: We study the disclosure decision and price-setting behavior of competing firms in the presence of cursed consumers, who fail to be sufficiently skeptical about a firm's quality upon observing non-disclosure of quality-relevant information. We show that neither competition nor the presence of sophisticated consumers necessarily offer protection to cursed consumers. Exploitation arises if markets are vertically differentiated, if there are many sophisticated consumers, and if it is more likely ex ante that product quality is high. Information campaigns that seek to educate consumers may encourage exploitation and decrease social welfare. Mandatory disclosure laws restore efficiency, but at the cost of redistributing rents from consumers to firms. Our simple model delivers a rich set of positive results, captures important markets, like those for food and consumer finance, and speaks to several recent policy initiatives aimed at consumer protection.

**Jeudi 13 octobre 2016 12:30-13:45****LAMBERT-MOGILIANSKY Ariane**(PSE) :__Phishing for phools - Dynamic consistency under non-classical uncertainty__**V.I. Danilov and V. Vergopoulos**- RésuméIn this paper we develop an expected utility theory in the context of non-classical(quantum) uncertainty. We replace the classical state space with a Hilbert space which allows introducing the concept of quantum lottery. Within that framework we formulate sufficient and necessary axioms in terms of choice behavior and establish a representation theorem. We show that demanding the consistency of choice behavior conditional on new information is equivalent with the von Neuman-Luder postulate applied to beliefs. As a consequent we find that rational choice behavior can be consistent with non-commutativity in updating. Finally, we discuss the value of our results for behavioral economics

**Jeudi 6 octobre 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**DOSIS Anastasios**:__*__

**Jeudi 29 septembre 2016 12:45-13:45****REDLICKI Bartosz**(U of Cambridge) :__Spreading Disinformation.__- RésuméIt is a common tactic, e.g., for political parties, to manipulate information without any official verification and then spread it by word of mouth. I develop a theoretical model which combines elements of Bayesian persuasion and cheap talk. An individual, called “manipulator”, manipulates information by designing an information policy, which is modelled by Bayesian persuasion. The information then diffuses via a chain of cheap talk communication in a population of agents with heterogeneous preferences. The manipulator’s objective is to spread disinformation so that as many agents as possible take high actions. He needs to take into account that his information policy can affect the agents’ own incentives to manipulate information during its diffusion. The manipulator’s optimal information policy has a high degree of manipulation if the majority of agents have a large upward bias. On the other hand, he may set a lower degree of manipulation in order to make the information more credible to agents with no upward bias or to improve information diffusion between agents with different preferences.

**Jeudi 23 juin 2016 12:45-13:45****GABAIX Xavier**(NYU Stern) :__tba__

**Jeudi 16 juin 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**SEROR Marlon**(University of Bristol) :__A Theory on the Evolution of Civic Culture and Political Cycles__- RésuméWe argue in this paper that political cycles in democracies impede the diffusion of civic values, thereby contributing to a persistent political instability. Indeed, we build a theory on the evolution of political cycles under repeated majoritarian elections whenever civic values are transmitted inter-generationally. We first establish that cycles in public expenditures result from asymmetric information on incumbents' civic attitude. We then show that recurrent suboptimal increases in the public provision weight on average less on the cultural transmission strategies of the agents with low civic values. This is because the latter have on average a lower propensity to contribute to the public provision, and consequently suffer less from the higher taxes induced by the cycles. Thus, cycles of a sufficiently high magnitude persist because they create weak civic majorities in which informed voters are not majoritarian. At the opposite, low magnitude cycles may not impede the formation of a majority of informed civic voters. The preceding cycles are then transitory and civic norms consolidate once they disappear. Finally, we suggest that the composition of the cycling expenditures matters as well in explaining the diffusion of civic values.

**Jeudi 9 juin 2016**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**ZACCOUR Georges**(HEC Montréal) :__TBA__

**Jeudi 2 juin 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**BRETON Michèle**(HEC Montréal) :__Mergers in Nonrenewable Resource Oligopolies and Environmental Policies: A green paradox__- RésuméWe examine the profitability of horizontal mergers within nonrenewable resource industries, which account for a large proportion of merger activites worldwide. We focus on the case of a tripoly where each firm owns a private resource stock of the resource, and firms compete in quantities. We show that a merger is profitable when the resource stock owned by each firm is small enough. We also consider an endogenous merger process and show that monopolization is profitable when the resource stock of each firm is small enough. In the case where pollution is generated by the industry's activity, we show that an environmental policy that increases firms' production cost or reduces the price received by firms can deter a merger and therefore result in larger industry emissions than under a laisser-faire scenario. The impact of such an environmental policy on merger profitability generally depends on the level of the resource stocks. When the stocks are large enough, a tax on extraction reduces the losses from a horizontal merger.

**Jeudi 26 mai 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**KOESSLER Frédéric**(PSE) :__tba__

**Jeudi 19 mai 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**GALASSO Emanuela**(World Bank ) :__Bundled Discounts and Foreclosure in Wholesale Markets__**Enrique Ide**- RésuméCan a multi-product firm offer bundled-discounts to foreclose a more efficient single-product supplier? We show that the degree of downstream competition is key. Bundled-discounts are successful in depriving single-product rivals of scale economies only to the extent that buyers are disorganized and their valuations heterogeneous. This holds when suppliers deal directly with final consumers, or when they sell to retailers that compete intensely in downstream markets, but it is lost in the case of retail monopolies. These results are robust to alternative timing assumptions and different ways to model scale economies.

**Jeudi 12 mai 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**RITZBERGER Klaus**(IAS Vienna) :__The Curse of Poverty and the Blessings of Wealth__- RésuméIf productivity in a society is low, then the only equilibrium involves inefficiently selfish (autarkic) behavior. If productivity is sufficiently high, then there are several equilibria that realize substantial welfare gains through reciprocal behavior. Equilibria for productive societies are distinguished by how they treat the wealth distribution. While conservative societies preserve the distribution, egalitarian behavior further improves welfare by transfers. Yet, both conservative and egalitarian (productive) societies exclude the poorest, even though that involves a welfare loss. So, there are a social and an individual poverty trap. The first is technological, the second behavioral.

**Jeudi 14 avril 2016 12:45-13:45****ENACHE Andreea**(Stockholm School of Economics) :__TBA__

**Jeudi 7 avril 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**VARDAR Boris**:__TBA__

**Jeudi 31 mars 2016 12:45-13:45****La séance a été annulée.****LAMBERT-MOGILIANSKY Ariane**(PSE) :__Dynamic consistency of expected utility under non-classical(quantum) uncertainty__**Danilov V.I. and V. Vergopoulos**- RésuméIn this paper we develop an expected utility theory in the context of non-classical(quantum) uncertainty. We replace the classical state space with a Hilbert space which allows introducing the concept of quantum lottery. Within that framework we formulate sufficient and necessary axioms in terms of choice behavior and establish a representation theorem. We show that demanding the consistency of choice behavior conditional on new information is equivalent with the von Neuman-Luder postulate applied to beliefs. As a consequent we find that rational choice behavior can be consistent with non-commutativity in updating. Finally, we discuss the value of our results for behavioral economics.

**Jeudi 24 mars 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**BLOCH Francis**(PSE) :__The formation of partnerships in social networks__**Bhaskar Dutta, Stephane Robin et Min Zhu**- RésuméThis paper analyzes the formation of partnerships in social networks. Agents randomly request favors and turn to their neighbors to form a partnership where they commit to provide the favor when requested. If favors are costly, agents have an incentive to delay the formation of the partnership. In that case, we show that for any initial social network, the unique Markov Perfect equilibrium results in the formation of the maximum number of partnerships when players become infinitely patient. If favors provide benefits, agents rush to form partnerships at the cost of disconnecting other agents and the only perfect initial networks for which the maximum number of partnerships are formed are the complete and complete bipartite networks. The theoretical model is tested in the lab. Experimental results show that a large fraction of the subjects (75\%) play according to their subgame perfect equilibrium strategy and reveals that the efficient maximum matching is formed over 78\% of the times. When subjects deviate from their best responses, they accept to form partnerships too early. The incentive to accept when it is optimal to reject is positively correlated with subjects' risk aversion, and players employ simple heuristics -- like the presence of a captive partner -- to decide whether they should accept or reject the formation of a partnership.

**Jeudi 17 mars 2016 12:30-14:30**- Campus Jourdan, bâtiment E, rez-de-chaussée, salle 101
**BONNET Celine**(Toulouse School of Economics, INRAE) :__Information Design: The Random Posterior Approach (1)__**GEEROLF François**(UCLA) :__Paretos from Production Functions (2)__- Résumé(1)Information affects behavior by affecting beliefs. Information design studies how to disclose information to a group of players to incentivize them to behave in a desired way. This paper is a theoretical investigation of information design, culminating with a representation theorem and a fundamental application of it. We adopt a random posterior perspective, viewing information design as belief manipulation rather than information disclosure. The representation theorem shows that it is as if the designer manipulated beliefs in a specific way, shaping the approach in games, as did Kamenica and Gentzkow (2011) in one-agent problems. The representation theorem can also be implemented in specific problems, for example in the beauty contest and multiple-agent problems. We focus on an application that we dub the Mother's Problem.

(joint with J. Perego and I. Taneva)

(2)This paper shows that Pareto distributions can arise from production functions rather than from the distribution of primitives. A version of Garicano (2000)’s knowledge-based production hierarchies microfounds such a production function. It generates under very limited assumptions on the distribution of primitives (here, agents’ skills), Pareto distributions for span of control of CEOs as well as intermediary managers, and in particular Zipf’s law for firm sizes when the number of layers of management becomes large. This breakdown of the aggregate firm size distribution receives important empirical support in the French matched employer-employee administrative data. This novel justification of Pareto distributions can shed a new light on why outcomes are often so out of proportion compared to underlying primitives. For example, it provides a framework to study the recent rise in top income inequality. The analysis also has striking implications on the literature on firm heterogeneity: Pareto distributions are in fact the benchmark distributions arising in the case of perfect homogeneity, while heterogeneity in primitives should instead be backed out in the deviations from Pareto distributions.

**Jeudi 10 mars 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**VENEL Xavier**(Paris 1 et PSE) :__Dynamical strategic interaction in social networks__- RésuméWe consider a dynamical model of influence with a set of non-strategic agents and two strategic agents. The non-strategic agents are organized in a fixed network describing how they influence each other. The strategic agents have opposite opinion and try to drive the opinion of the society by targeting key-players in the network. We formulate this problem as a two-player zero-sum stochastic games allowing the strategic agents to change their target at every stage of the game. We prove the existence of the uniform value: if the player are sufficiently patient, both players can guarantee the same mean-average opinion without knowing the exact discount factor.

**Jeudi 18 février 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**FAVART Thomas**(PSE) :__Collusion in Capacity Under Irreversible Investment__- RésuméThis work studies the possibility of collusion under irreversible investment in production capacity. The irreversibility of investment has two effects. It reduces the profitability of short run deviation, as the deviating firm has to invest in order to increase its capacity. It also creates a long run effect. Once the deviating firm has invested, it is committed to its new capacity. The deviation may thus lead to a preemption of the punishing firm. When the firms become more patient, the short run profitability of a deviation increases whereas the long run profitability is reduced. When the second effect dominates, collusion is harder to sustain for more patient firm. This contrasts with the literature on collusion. Finally, this work shows that the smallest firm has the most incentives to deviate.

**Jeudi 11 février 2016 12:45-13:45****HERNANDEZ Penelope**(Valencia) :__Conflict and segregation in networks: individual preferences and social influence__- RésuméWe examine the interplay between a person's individual preference and the social influence others exert. We provide a model of network relationships with conflicting preferences, where individuals are better off coordinating with those around them, but not all prefer the same action. We test our model in an experiment, varying the level of conflicting preferences between individuals. Our findings suggest that preferences are more salient than social influence, under conflicting preferences: subjects relate mainly with others who prefer the same. This leads to two undesirable outcomes: network segregation and social inefficiency. The same force that helps people individually hurts society.

**Jeudi 4 février 2016 12:45-13:45****LAMY Laurent**:__On the benefits of set-asides__**Philippe Jehiel**

**Jeudi 28 janvier 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**NAX Heinrich**(ETH Zurich) :__A behavioral study of « noise » in coordination games__**Michael Mäs**- Résumé‘Noise’ in this study, in the sense of evolutionary game theory, refers to deviations from prevailing behavioral rules. Analyzing data from a laboratory experiment on coordination in networks, we tested ‘what kind of noise’ is supported by behavioral evidence. This empirical analysis complements a growing theoretical literature on ‘how noise matters’ for equilibrium selection. We find that the vast majority of decisions (96% ) constitute myopic best responses, but deviations continue to occur with probabilities that are sensitive to their costs, that is, less frequent when implying larger payoff losses relative to the myopic best response. In addition, deviation rates vary with patterns of realized payoffs that are related to trial-and-error behavior. While there is little evidence that deviations are clustered in time or space, there is evidence of individual heterogeneity.

**Jeudi 21 janvier 2016 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**RIVERA Thomas**(HEC) :__Incentives and the Structure of Communication__- RésuméThis paper analyzes the incentives that arise within an organization when communication is restricted to a particular network structure (e.g., a hierarchy). We show that restricting communication between the principal and agents may create incentives for the agents to misbehave when transmitting information and tasks throughout the organization. Such incentives can render the principal's most preferred outcome infeasible and therefore introduces a trade off between the cost of communication borne by the principal and the benefit of curbing incentives to deviate induced by the communication structure. To remedy this issue, we provide necessary and sufficient conditions on the topology of the network of communication such that restricting communication to a particular network does not restrict the set of outcomes that the principal could otherwise achieve. In this sense, we show that for any underlying incentives and any outcome available when communication is unrestricted, there exists a (finite) communication scheme restricted to a particular network that implements this outcome (i.e., does not induce agents to misbehave in the communication phase) if and only if that network satisfies our conditions

**Jeudi 14 janvier 2016 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 10
**BEHRINGER Stefan**:__Public Good Provision with many Agents: The k-success technology__**Yukio Koriyama, Ecole Polytechnique**- RésuméIn this paper, we consider a class of public good provision problems in which the production function takes the form of k-success technology, an extension of the direct provision technology considered in Behringer (2013). These models are suitable to describe the free-rider problems in which there are a large number of agents who are both users and beneficiaries of a public good at the same time, e.g. open-source software or social networks. We provide results on asymptotic efficiency which connect a negative result of Mailath and Postlewaite (1990) and a positive result of Hellwig (2003), as well as a set of simple examples which allow us welfare comparison with the standard technologies.
- Texte intégral [pdf]

**Jeudi 17 décembre 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**LE CHAPELAIN Alexis**(Science-po) :__*__

**Jeudi 10 décembre 2015 12:45-13:45****RAI Birendra**(University of Mannheim) :__Preference Domains and Bilateral Bargaining: An Experiment__- RésuméTheoretical results and experimental findings from competitive settings suggest that agents with other-regarding preferences (ORP) often behave as if they have self-regarding preferences. We say an institution (game form) is `robust' wrt ORP if the above holds. We experimentally explore whether and when such robustness arises in non-competitive environments using a series of ultimatum games where proposer's monetary payoff upon rejection varies. The data provides some insights into when and why robustness holds wrt which type of ORP.

**Jeudi 26 novembre 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 10
**IARIA Alessandro**(CREST) :__Estimating Demand Parameters with Choice Set Misspecification__- RésuméWe describe methods to estimate demand parameters in the presence of choice set misspecification due to unobserved individual choice sets. We show that a consumer’s probability of making a choice from her true choice set can be written as the probability she makes the same choice from a universal choice set, an assumption commonly made in the empirical literature, less a bias term capturing the probability that she would choose any of the elements in her true choice set. We show how to estimate models that avoid this bias under Logit demand using “sufficient sets,” instances where (observationally identical) individuals make at least two choices from the same unobserved choice set. Examples of sufficient sets include a variation of the Fixed Effect Logit approach due to Chamberlain (1980), using observed individual-specific purchase histories, and assuming common unobserved choice sets across individual purchases in a given retailer and time period. We illustrate these ideas using household-level scanner data exploiting variation in choice sets based on the introduction of a new product variety and heterogeneous product offerings across stores in the market for ketchup in the UK. Our results show that accounting for choice set misspecification is critical for accurately estimating own- and cross-price elasticities of demand.

**Jeudi 19 novembre 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**MARFAN Manuel**(PSE) :__Multidimensional Delegation__- RésuméWe study a principal agent model without transferable utility in which both the agent's private information and the decision to be made is multidimensional. We study properties of the optimal delegation set, that is, the set of actions that will be taken on equilibrium, and compare with the single dimensional case. Finally, we study the use of "constant caps" as seen in Frankel (2014), and obtain necessary and sufficient conditions for their optimality.

**Jeudi 12 novembre 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**DOSIS Anastasios**(Essec) :__Nash Equilibrium in? Competitive Markets with Adverse Selection__- RésuméWe generalise the results of Rothschild and Stiglitz (1976). We define the generalised Rothschild - Stiglitz allocation (RSA), and we show that, in every possible market, if the RSA is efficient, it is also a pure Nash equilibrium. In markets in which the RSA is not efficient, we specify a broader than Rothschild and Stiglitz (1976) class of environments in which no pure Nash equilibrium exists.

**Jeudi 5 novembre 2015 12:45-13:45**- Campus Jourdan, bâtiment B, 1er étage, salle CMH
**SZECH Nora**(Kalsruhe) :__*__

**Jeudi 29 octobre 2015 12:45-13:45**- Campus Jourdan, bâtiment B, 1er étage, salle CMH
**HARRINGTON JOE**(Warton Business School) :__A Theory of Collusion with Partial Mutual Understanding__- RésuméUnlawful collusion is when Örms have a mutual understanding to coordinate their behavior for the purpose of achieving a supracompetitive outcome. Given the legal focus on mutual beliefs, this paper initiates a research program to explore how much and what type of mutual beliefs among Örms allows them to e§ectively collude. Focusing on price leadership as the collusive mechanism, it is assumed that Örms commonly believe that price increases will be at least matched but lack any shared understanding about who will lead, when they will, and at what prices. Su¢ cient conditions are derived which ensure that supracompetitive prices emerge. However, price is bounded below the maximal equilibrium price.

**Jeudi 15 octobre 2015 12:45-13:45**- Campus Jourdan, bâtiment B, 1er étage, salle CMH
**TAKAHASHI Hidenori**(Mannheim) :__Strategic Design Under Uncertain Evaluations: Structural Analysis of Design-Build Auctions__- RésuméI investigate firms’ competition over price and product design under uncertain design evaluations in the context of Design-Build (DB) auctions. Reviewers’ design evaluations contain uncertainty from a bidder’s perspective, leading luck to dampen differences in firms’ chances of winning. I model bidders’ behavior and show semiparametric identification of the model primitives. Uncertain evaluations reduce the expected price of design quality, and exacerbate an auctioneer’s uncertainty in auction outcomes. A simple adjustment in the auction mechanism may completely shut down the impact of uncertain evaluations on bidding incentives, restoring efficient allocations of projects.

**Jeudi 8 octobre 2015 12:45-13:45**- Campus Jourdan, bâtiment B, 1er étage, salle CMH
**BOUACIDA Elias**(PSE) :__An application of behavioral welfare economics to consumption data__- Texte intégral [pdf]

**Jeudi 18 juin 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**WREN-LEWIS Liam**(PSE) :__*__

**Jeudi 11 juin 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**COMBE Julien**(PSE) :__The Design of Teacher Assignment : Theory and Evidence__**Co-author(s) : Olivier Tercieux et Camille Terrier**

**Jeudi 4 juin 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**DARGNIES Marie Pierre**(Université Paris-Dauphine) :__Gender Di fferences in Reactions to Feedback and Willingness to Compete__**Co-author(s) : Noémi Berlin**- RésuméIn Western societies, it is generally known men have a greater taste for competition than women. However, the determinants of the decision to enter competitions are still not fully understood. Providing feedback on relative performance reduces the gender gap in competitive entry. The aim of this paper is twofold. We rst evaluate how participants update their beliefs after receiving feedback informing them whether their performance is below or above the median performance. Second, we are interested in how men and women react to this information in terms of competitive entry. Our rst result is that participants, and women in particular, react more strongly to the feed- back they receive than would a Bayesian agent. As far as entry into the competition is concerned, below-median participants adjust their entry decision according to the competition they expect to face, while above-median participants do not. However, the behavior behind these results is quite di erent for men and women: women mainly re- act to information on their own performance, while men seem to respond more to their beliefs over the competition they will face. Moreover, most of the e ect of feedback and the information regarding the level of the competition on the decision to compete seems to operate via beliefs.

**Jeudi 28 mai 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**SEVERINOV Sergei**(University of British Columbia) :__Optimal Mechanisms for Budget Constraint Buyers__**Co-author(s) : A. Boulatov**- RésuméThe paper deals with the mechanism design for selling to buyers who have known budget constraints. With asymmetric budgets, our problem is that of asymmetric optimal mechanism design. We derive and characterize the optimal mechanism. It belongs to one of two classes. When the budget differences are small, the optimal mechanism discriminates only between high-valuation buyers for whom the budget constraint is binding. All low valuations buyers are treated symmetrically despite budget differences. In case of large budget differences, the optimal mechanism discriminates in favor of buyers with small budgets when the valuations are low, and it discriminates in favor of buyers with larger budgets when the valuations are high.

**Jeudi 7 mai 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**CAMARA Fanny**(Goethe University of Frankfurt) :__When Competition Hinders Transparency: Evidence from the Pharmaceutical Industry__

**Jeudi 9 avril 2015 12:45-13:45****RENOU Ludovic**(University of Essex) :__Revealed preferences over risk and uncertainty__**Co-authors : Matthew Polisson & John K.-H. Quah)**- RésuméConsider a finite data set where each observation consists of a bundle ofcontingent consumption chosen by an agent from a constraint set of such bundles. We develop a general procedure for testing the consistency of this data set with abroad class of models of choice under risk and under uncertainty. Unlike previous work, we do not require that the agent has a convex preference, so we allow for risk loving and elation seeking behavior. Our procedure can also be extended to calculate the magnitude of violations from a particular model of choice, using an index firstsuggested by Afriat (1972, 1973). We then apply this index to evaluate different models (including expected utility and disappointment aversion) in the data collected by Choi et al. (2007). We show that more than half of all subjects exhibiting choice behavior consistent with utility maximization are also consistent with models of expected utility and disappointment aversion.

**Jeudi 26 mars 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**BONTEMPS Christian**(Toulouse) :__*__

**Jeudi 19 mars 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**MARLATS Chantal MARLATS**(Paris 2-Assas) :__Observation delays in teams__**Co-authors : Sid Gordon & Lucie Ménager**- RésuméWe study a dynamic team problem in which two players contribute to a common project that leads to a success with some probability increasing with effort. Players are informed of the success or the failure of their partner with a time lag. We interpret this delay as a technological constraint on information transmission. We find that the equilibrium payoff follows a regular cyclical pattern and the strategies are bang bang. This contrasts with the case without delay in which payoffs and efforts are constant over time. Also, there exists an asymmetric equilibrium if and only if the time lag is sufficiently small.

**Jeudi 5 mars 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**BERGES Alexis**(PSE) :__Investigating the impact of uncertainty on firms with dynamic costs on the electricity market__- RésuméWe provide an empirical analysis of the French electricity market. Specifically, we look at the impact of uncertainty on supplier strategies and take this as evidence that dynamic costs matter. Our approach to separate out the uncertainty from market demand expectations and predictability of renewables generation is novel. Both proxies for uncertainty used are new, uncertainty from market demand is inferred from the prediction errors that firms make in a demand estimation and uncertainty from renewable production is computed in a bottom-up approach from local weather forecasts. Instead of opting for a time series regression, we understand all hourly auctions as a cross-sectional dataset and control for the time of the day by using continuous transition variables for daytime periods. Similarly, we control for seasonality using continuous variables rather than dummies. Thereby, we are able to leverage our dataset and increase the sample size for each of our regression and improve the precision of our estimates.

**Jeudi 12 février 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**JEHIEL Philippe**(PSE) :__Investment strategy: Sampling, categorizing, and selection bias__- RésuméIn most cases, it is hard to have a good understanding of all possible (probablistic) consequences of investment decisions. I propose an approach in which decision makers first categorize projects based on similarity as perceived from small samples, then get access to aggregate statistics about the effect of the investment in the various categories of projects, and finally make investment decisions in randomly drawn projects based on these aggregate statistics. I compare the resulting outcome to that arising in a benchmark Bayesian model and highlight the mistakes of the selection bias type that such heuristics induce. I also apply the approach to multi-player contexts so as to shed light on the so called no trade theorem.

**Jeudi 5 février 2015 12:45-13:45****GORELKINA Olga**(Max Planck Institute, Bonn) :__Selling Money on Ebay__**Co-author(s) : Alia Gizatulina**- RésuméTo determine whether subjects display social preferences in the field, we conduct a trading experiment with a pool of German Ebay users. Acting as a seller, we offer Amazon gift cards with values from 5 to 500 euros. Randomly arriving buyers, the subjects of our experiment, make take-it-or-leave-it price offers. From each offer we infer the fraction of the gains from trade that a buyer proposes to leave to the seller. We find that a large numbers of offers correspond to approximately equal sharing between the buyer and the seller. The amount of money at stake has no significant effect on the relative prices offered. More experienced buyers are more likely to display behavior consistent with the theoretical model.

**Jeudi 15 janvier 2015 12:45-13:45****GORDON Sidartha**(Sciences Po) :__Information Choice and Diversity: The Role of Strategic Complementarities__- RésuméWe study a class of games where players face restrictions on how much information they can obtain on a common payoff-relevant state, but have some leeway in covertly choosing the dependence (or similarity) between their signals, before simultaneously choosing actions. Using a new dependence stochastic ordering between signals, we show that each player prefers and chooses information that is more dependent on the information of other players whose actions are either isotonic and complements with her actions or antitonic and substitutes with her actions. Similarly, each player prefers and chooses information that is less dependent on the information of other players whose actions are antitonic and complements with her actions or isotonic and substitutes with her actions. We then provide sufficient conditions for certain information structures to arise in equilibrium, in particular for public information and for private information. We discuss applications to collective action, oligopolies, supply chains, financial markets, and spec- ulative currency attacks. Equilibrium information structures may be inefficient. Making which signals were chosen (but not their realizations) publicly observable may restore efficiency.

**Jeudi 8 janvier 2015 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**GAGNEPAIN Philippe**(PSE - Paris 1) :__Knowledge Spillovers in Cost-Reduction Incentives__**Co-author(s) : Luis Aguiar**- RésuméThe goal of this paper is to identify and measure the relevance of knowledge spillovers in the French urban transport industry, where most regulated transportation networks are operated by firms that belong to the same company. We build and estimate a structural cost regulation model under incomplete information where the service is regulated by an authority and is provided by a single operator that may be owned by a larger company. We identify the knowledge spillovers which arise for some operators being linked to a same group, and see how they influence the firms’ decisions of exerting effort in order to reduce their operating costs. Our model provides us with estimates of the operators’ inefficiencies, the effort of the managers and the knowledge spillovers. Our results show that knowledge spillovers are indeed relevant for the existing industrial groups present in the French urban transport industry. Simulation exercises provide evidence of significant reductions in total operating cost following the enlargement of industrial groups and/or mergers between existing groups.

**Jeudi 27 novembre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**BLOCH Francis**(PSE -Paris 1) :__The formation of partnerships in social networks__**Co-autheur : Bhaskar Dutta**

**Jeudi 20 novembre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**HOUNGBONON Georges Vivien**(PSE) :__The impact of entry and merger on the pricing of multiproduct firm__- RésuméIn this paper, we design a hedonic pricing model to estimate the dynamic impact of entry and merger on access and usage prices in a context of mixed bundling between two independent services with respectively high and low valuations. Using telecommunications tariffs data from OECD operators and a double difference matching estimation strategy, we find that entry does not raises the access price of the services, irrespective of their valuation and whether in standalone or in bundle. However, merger does increase the bundle access price for the low valuation service while decreasing the standalone access price for the high valuation service. Although both the standalone and bundle usage prices of the low valuation service are also not affected by entry, they fall following a merger. For the high valuation service, entry decreases its standalone usage price while raising its bundle usage price. The inverse holds for merger.

**Jeudi 13 novembre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**SCHOLZ Eva-Maria**(Université Catholique de Louvain) :__How to license a downstream technology when upstream firms are capacity constrained?__- RésuméIn this paper we study the relationship between capacity constraints and licensing strategies. To do so, we focus on the licensing strategy of an outside patentee who licenses a process innovation to the downstream sector of a vertical Cournot oligopoly. Downstream firms source their input requirements from an upstream sector in which a subset of firms is capacity constrained. It is shown that (1) the patentee licenses large innovations via per-unit royalty contracts and small innovations via fixed-fee contracts; (2) an increase in the number of capacity constrained upstream firms makes it more likely that the patentee opts for a per-unit royalty contract; (3) depending on the number of capacity constrained upstream firms, privately optimal per-unit royalty contracts may maximise economic welfare.

**Jeudi 6 novembre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**HAGENBACH Jeanne**(Polytechnique) :__COMMUNICATION WITH EVIDENCE IN THE LAB__**Co-author : E. Perez-Richet**- RésuméWe examine strategic communication in sender-receiver games where senders can only transmit information using verifiable statements. Applying the method of Hagenbach, Koessler, Perez-Richet [Ecta14], we represent players’ incentives by a graph mapping which type of a sender would like to be believed as which other type, the masquerading graph. In theory, when this graph is acyclic, the receiver can deter the use of any vague message by attributing it to a type that no potential sender of this message would like to be. If receivers make such skeptical inferences, full revelation of information obtains in equilibrium. On the contrary, when the masquerading graph is cyclic, the use of vague messages is harder to discourage and full disclosure more difficult to sustain. This paper investigates the relevance of this graphical distinction in the lab. Our experimental design involves sender-receiver games with various payoffs and allows examining how subjects use and interpret evidences for the different resulting graphs. We show that information is more often revealed in games whose graph is acyclic than cyclic, and that receivers interpret vague evidences skeptically in most cases. While senders understand well when their type should be identified, they however fail using vague messages to fool receivers as finely as they could.

**Jeudi 16 octobre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**MARFAN Manuel**(PSE) :__*__

**Jeudi 9 octobre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**ALAIN Marie-Laure**(Polytechnique) :__The Impact of Retail Mergers on Food Prices: Evidence from France__**Co-author(s) : Claire Chambolle, Stephane Turolla and Sofia B. Villas-Boas**- RésuméThis paper analyzes the impact of a merger in the French retail sector on food prices, using a consumer panel data. We perform a difference-in-differences analysis by comparing price changes in stores for which the local market structure is affected by the merger to unaffected stores. In addition, we empirically investigate economic forces behind the observed price changes. On average, we find that the merger significantly raised competitors’ prices contemporaneously with merging firms’ price increases. Further, we show that competitor prices increase more in local markets that experience larger structural changes in concentration and chain differentiation.
- Texte intégral [pdf]

**Jeudi 2 octobre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**STEPANOV Sergey**(Higher School of Economics, Moscow) :__Biased Performance Evaluation in A Model of Career Concerns: Incentives versus Ex-Post Efficiency__

**Jeudi 25 septembre 2014 12:45-13:45**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**GOMES Renato**(Toulouse School of Economics) :__Competitive Screening under Heterogeneous Information__**Co-author(s) : Daniel Garrett and Lucas Maestri**- RésuméWe study competition in price-quality menus when consumers privately know their valuation for quality (type), and are heterogeneously informed about the offers available in the market. While firms are ex-ante identical, the menus offered in equilibrium are ordered so that more generous menus leave more surplus uniformly over types. More generous menus provide quality more efficiently, serve a larger range of consumers, and generate a greater fraction of profits from sales of low-quality goods. By varying the mass of competing firms, or the level of informational frictions, we span the entire spectrum of competitive intensity, from perfect competition to monopoly.

**Jeudi 3 juillet 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**MOULIN Hervé**(Rice University and Glasgow University) :__Segregation, Entropy, and Proportional Rationing__

**Jeudi 19 juin 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**ERDAMAR Bora**(PSE) :__On Manipulation from an Unacceptable Social Choice to an Acceptable One__**Co-author(s) : Remzi Sanver and Shin Sato**- RésuméNon-manipulability in collective decision making problems has been analyzed mainly through the axiom of strategy-proofness. In this paper, we propose a new concept of non-manipulability. We postulate that each agent misreports his preferences if and only if the misrepresentation leads to a change of the social outcome from an unacceptable one for this agent to an acceptable one. For the formulation of this idea the preference-approval framework is used. Possibility and impossibility results for the existence of a non-manipulable rule are provided.

**Jeudi 12 juin 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**LEVIN Dan**(Ohio State University) :__Separating Bayesian Updating from Non-Computational Reasoning: An Experimental Investigation__**Co-author(s) : James Peck et Asen Ivanov**- RésuméThrough a series of decision tasks involving colored cards, we provide separate measures of Bayesian updating and non-computational reasoning. We apply these measures to (and are the first to study) a common-value Dutch auction. This format is more salient than the strategically equivalent first-price auction and silent Dutch formats in hinting that one should condition one’s estimate of the value on having the highest bid. Both Bayesian updating skills and non-computational reasoning skills are shown to help subjects correct for the winners curse, as does the saliency of the active-clock Dutch format.

**Jeudi 5 juin 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**TREIBICH Rafael**(Polytechnique) :__Welfare Egalitarianism with Other-Regarding Preferences__- RésuméWe study the fair allocation of a one dimensional and perfectly divisible good when individuals have other-regarding preferences. A social ordering function associates any profile of ordinal preference relations with a complete ordering of all possible allocations. In both a model of average and positional externalities, we characterize a class of social ordering functions which satisfy appealing efficiency, fairness and consistency properties. These rankings require giving full priority to the worst off agent in the economy, where individual welfare is measured in a specific constructed way.

**Jeudi 22 mai 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**ISPANO Alessandro**(TSE & CREST) :__Competing over Gains, Cooperating over Losses__**Co-author(s): Peter Schwardmann (Toulouse School of Economics).**- RésuméWe let experimental subjects play a social dilemma game and vary whether payoffs are framed as losses or as gains. Subjects cooperate substantially more in the loss domain. We account for this finding with a theory that features noisy equilibrium behavior and reference-dependent risk preferences. We estimate a structural econometric model that incorporates risk preferences into an augmented cognitive hierarchy model. The estimated risk preference parameters are in line with evidence from individual decision-making tasks. Our model fits the experimental data well and is consistent with other puzzling results in the experimental literature on games played over losses. These findings have several important applications, including price competition between firms during recessions or concessions in climate change negotiations.

**Jeudi 15 mai 2014 12:45-15:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**LACLAU Marie**(PSE) :__Bayesian games with multiple priors__**Co-author : Ludovic Renou (Essex).**- RésuméWe study Sender-Receiver games where the Receiver has multiple priors over the states of the nature (the Sender may have a single or multiple priors). We assume full Bayesian updating and that the Receiver has max-min preferences. We establish necessary and sufficient conditions for a splitting lemma to hold. We derive a characterization of the equilibrium payoffs of the Sender.

**Jeudi 24 avril 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**REGUAND Mar**(Stanford GSB) :__Blowin' in the Wind: Sequential Markets, Market Power and Arbitrage__**Co-author(s): Koichiro Ito (Boston University)**- RésuméWe study strategic behavior of wind farms in electricity markets, and examine their dynamic adjustments through sequential markets. Wind production is uncertain and volatile, with the degree of uncertainty being reduced over the day. Therefore, sequential forward markets can improve market efficiency through information updating. However, pre-existing distortions such as market power and limited arbitrage may distort incentives to reveal accurate production forecasts. By using micro-level data in the Spanish electricity market, we show that wind farms exploit a forward market price premium, overstate their production over 20% at the day-ahead market, and only slowly adjust their commitments to expected production, increasing the dynamic inefficiencies from wind misplanning. Consistent with the premium being driven by market power, wind farms that have market power do not exploit the price premium, whereas competitive fringe farmers do the arbitrage. Our results show how pre-existing distortions can have unintended consequences in the market, making the integration of wind power even more challenging.

**Jeudi 10 avril 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**BOUSQUET Léa**(PSE) :__Time Inconsistency and Naivety within Hotelling Competition__- RésuméA monopoly offering a two-part tariff contract to time inconsistent consumers (both sophisticated and naive) do not exploit the time inconsistency bias but the naivety bias. Moreover, with perfect competition, firms do not succeed anymore to exploit the naivety bias even if an allocation inefficiency remains. So, we are interested to study the impact of Hotelling competition on the exploitation of these bias. We adapt DellaVigna and Malmendier (2004) framework to study the impact of imperfect competition and introduction of transport costs. These costs can be introduced at the contract period or at the consumption period. At the contract period, we observe that the competition even imperfect erases the exploitation of the naivety bias but the introduction of the transport costs conducts to a distortion between consistent and time inconsistent consumers. At the consumption period, transport costs generate an heterogeneity between consumers, firms are no longer able to provide a perfect commitment to sophisticated consumers. Consequently, allocation with both sophisticated and naive are not efficient.

**Jeudi 27 mars 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**WATSON Joel**(UC San Diego) :__Contractual Chains__- RésuméThis paper examines a model in which individuals contract with one another along the branches of a ?xed network and then select veri?able productive actions with externally enforced transfers. Special cases of the model include most settings of contracting with externalities that have been studied in the literature. The model allows for a type of externality that the previous literature has not explored fully—where a party is unable to contract directly with others whose actions affect his payoffs. The paper investigates the prospects for ef?cient outcomes under various contract-formation protocols (“contracting institutions”) and network structures. There are contracting institutions that always yield ef?cient equilibria for any connected network. A critical property is that the institutions allow for sequential contract formation or cancellation. The equilibrium construction features assurance contracts and cancellation penalties.

**Jeudi 20 mars 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**LEFEBVRE Perrin**(PSE) :__Lobbying, Informational Polarization and Political Contributions__**Co-author(s) : David Martimort**- RésuméCompeting interest groups can use two instruments to influence a decision maker: information supply, and monetary contributions. When engaging in informational lobbying, groups have to choose which elements of a political issue they want to focus on. Each element is ex ante more or less likely to serve each group's cause. We identify a tradeoff between the quality and the quantity of information the decision maker can get from group: the decision maker prefers ex ante neutral information, but it is more easy to induce groups to gather information about a priori favorable arguments. As a result of this tradeoff, the organization of lobbying will exhibit informational polarization, each group at equilibrium specializing in fields of expertise more likely to support their respective interest. Contributions do not qualitatively modify this result, but have an impact on groups' incentives to supply information. A group more efficient in the use of political contributions will tend to provide less information, while the effect of his competitors strength on a group's informational strategy is ambiguous. Depending on groups' efficiency in the use of contributions, information and contributions can appear as complement or substitute instruments.

**Jeudi 13 mars 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**DE BELSUNCE Henri**((International Max Planck Research School for Competition and Innovati) :__On Essential Patents and Vertical Structure in the Context of Cumulative Innovation__- RésuméThis paper focuses on the incentives to innovate in the context of cumulative innovation. It shows that the presence of essential patents exerts a detrimental effect on the innovation incentives of vertically integrated firms that hold a license reading on the essential technology and compares this effect under a property vs. liability regime. An application of the model shows that the effect also holds for licensees of patent pools hat feature a partial termination clause, e.g. MPEG-2. The partial termination clause extends the standard grantback requirement of patent pools to include related patents of pool licensees. The paper helps to explain the move of the DOJ in 2006 from a property rule to a liability rule for patent litigation cases involving essential patents, since the negative externality on cumulative innovation is exacerbated when courts employ a strict property rule for patent litigation cases. Furthermore, it suggests a possible explanation for the drop in innovation rates of pool outsiders who research in the technological proximity of the pool. Also, the paper gives a possible explanation for why vertically integrated firms are overrepresented (underrepresented) among the pool members (outsiders) as compared to pools that do not feature this clause. The model lends support to the finding that innovations made by outsiders are of higher quality then those of pool members. The pro-competitive effects, on which the Department of Justice based its positive ruling for the partial termination clause in the case of the MPEG-2 pool, are also a feature of the model.

**Jeudi 6 mars 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**SCHROEDER David**(Birkbeck College, University of London) :__A new measure of equity and cash flow duration: The duration-based explanation of the value premium revisited__- RésuméThis paper proposes a new method for estimating an expected equity and cash flow duration for individual shares. In analogy to bond duration, we derive duration from a share’s price sensitivity to changes in the discount rate. To ensure that the duration estimates are forward-looking, expectations of future cash flows are taken from equity analysts. In line with the duration-based explanation of the value premium, we find that value stocks have shorter durations than growth stocks. However, there is no significant relation between a firm’s cash flow duration and stock returns, which implies that cross-sectional differences in the firms’ cash flow timing cannot explain the value premium. Instead, we show that the firms’ equity duration can explain the cross-section of stock returns, and the value premium in particular.

**Jeudi 13 février 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**LEHLE Vincent**(Paris Dauphine) :__Afriat’s Theorem for Indivisible Goods__**Co-author : Françoise Forges**- RésuméWe identify a natural counterpart of the standard GARP for demand data in which goods are all indivisible. We show that the new axiom (DARP, for “discrete axiom of revealed preference”) is necessary and sufficient for the rationalization of the data by a well-behaved utility function. Our results complement the main finding of Polisson and Quah, Am. Econ. J.: Micro. 5(1) p.28-34 (2013), who rather minimally modify the original consumer problem with indivisible goods so that the standard GARP still applies.

**Jeudi 6 février 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**HOUNGBONON Georges Vivien**(PSE) :__Is there an optimal level of competition for maximal investment? Empirical evidence from the mobile telecommunications industry__**Co-author : François JEANJEAN (Orange, Regulatory Affairs)**- RésuméIn this paper, we use data from the mobile telecommunications industry to test whether there is an optimal level of competition which guarantees maximal investment in new technologies. Using the Lerner index of monopoly power as a measure of competition, we find an optimal level of competition at 62%. That is, investment rises from a low level of competition up to 62% and then falls above this level. The optimal level of competition is higher for firms which supply access to the fixed telecommunications network (65%), and lower for market leaders (56%). For laggards, the optimal level is much higher, at 80%. Our results suggest that antitrust agencies should consider the welfare implications of their competition policies when the Lerner index is close to 38%.

**Jeudi 23 janvier 2014 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**TSUR Matan**(University of Vienna) :__Financial Contracts and Bargaining__- RésuméI consider a firm with many projects; the output of each one will be sold to a prospective buyer. The terms of trade will be determined via some bargaining procedure. Projects require financing and the firm can borrow against the future revenue from the sales. I study how the firm’s financial obligations to an outside investor(s) influence the bargaining outcomes with prospective buyers. Does the firm prefer to bundle projects together and finance them jointly, e.g. taking out a single loan tied to the joint revenue, or finance them separately, e.g. with separate loans tied to different projects? The answer depends on the types of securities that are available. When constrained to standard debt, the firm extracts a larger share of the surplus by financing projects separately. But the advantage of separate financing is not preserved when the firm can design more complicated securities- convex combinations of debt and equity. The optimal securities are also characterized.

**Jeudi 12 décembre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**BOYER Pierre**(Mannheim) :__Efficiency, Welfare, and Political Competition__**Co-author: Felix Bierbrauer**- RésuméWe study political competition under the assumption that voters have private information on their valuation of publicly provided goods. Preferencesare quasilinear in private goods consumption. We show that equilibrium policies are surplus-maximizing. This is in contrast to the conventional wisdom that incentive compatibility constraints make it harder to obtain surplus-maximizing outcomes. Our result extends to a Mirrlees (1971)-model of income taxation. In this model, however, the surplus-maximizing outcome can be interpreted as a political failure to provide for welfare-enhancing redistribution among individuals who differ in their productive abilities.

**Jeudi 5 décembre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**TAKAHASHI Yuya**(Mannheim) :__Testing for Equilibrium Multiplicity in Dynamic Markov Games__**Co-authors : Taisuke Otsu and Martin Pesendorfer.**- RésuméThis paper proposes several statistical tests for finite state Markov games to examine the null hypothesis that the data are generated from a single equilibrium. We formulate tests of (i) the conditional choice probabilities, (ii) the steady-state distribution of states and (iii) the conditional distribution of states conditional on an initial state. In a Monte Carlo study we find that the chi-squared test of the steady-state distribution performs well and has high power even with a small number of markets and time periods. We apply the chi-squared test to the empirical application of Ryan (2012) that analyzes dynamics of the U.S. Portland Cement industry and test if his assumption of single equilibrium is supported by the data.

**Jeudi 28 novembre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**HOLLARD Guillaume**(Paris 1) :__Non-strategic players are the rule rather than the exception__- RésuméIndependent studies on experimental games provide converging, and puzzling evidence: a noticeable fraction of players behave in a non-strategic way. It is hard to believe that so many players act in way that is hard to reconcile with any theory of strategic behavior. This study is designed to understand why. We test commonly suggested explanations such as insucient stakes, lack of attention or misconceptions about the game. None appear to be entirely convincing. Using reaction-time data, we show that non-strategic subjects spend time thinking about the games and do pay attention to changes. Moreover, players classied as non-strategic in a first set of games continue to act non-strategically in subsequent games. Our results suggest that the existence of non-strategic players in one-shot games is a robust feature of human cognition. Bearing in mind that our subjects are international chess players, we wonder why their strategic chess-playing ability does not transfer to laboratory games. Theoretical and economic consequences are discussed.
- Texte intégral [pdf]

**Jeudi 21 novembre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**COLLIARD Jean-Edouard**(European Central Bank) :__Monitoring the supervisors: optimal regulatory architecture in a banking union__- RésuméI study the optimal architecture of bank supervision in a federal system. A central supervisor relies on off-site monitoring to learn about a bank's soundness, and then decides whether she should perform an on-site inspection or whether it is better left to local authorities. Local supervisors are more efficient at performing on-site examinations, but do not internalize cross-border externalities. The optimal degree of supervisory centralization depends on the severity of these externalities, the opacity of the supervised bank and the specificity of its assets. Cross-border externalities are endogenous as in equilibrium investors react to the supervisory architecture. Better supervision leads to more financial integration, which worsens the incentives of local supervisors. The economy can be trapped in an equilibrium with low supervision and low integration, while a forward-looking design of the supervisory architecture would coordinate economic agents on a superior equilibrium.
- Texte intégral [pdf]

**Jeudi 7 novembre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**TERCIEUX Olivier**(PSE) :__Efficiency and Stability in Large Markets__**Co-autheur : Y-K. Che**

**Jeudi 31 octobre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**EHLERS Lars**(University of Montréal) :__Matching Markets under (In)complete Information__**Co-autheur(s): Jordi Masso**- RésuméWe introduce incomplete information to centralized many-to-one matching markets. This is important because in real life markets (i) any agent is uncertain about the other agents' true preferences and (ii) most entry-level matching is many-to-one (and not one-to-one). We show that given a common prior, a strategy prole is an ordinal Bayesian Nash equilibrium under incomplete information in a stable mechanism if and only if, for any true prole in the support of the common prior, the submitted prole is a Nash equilibrium under complete information in the direct preference revelation game induced by the stable mechanism.
- Texte intégral [pdf]

**Jeudi 24 octobre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**TARNAUD Razvan**(Paris 1) :__Information Aggregation using Logarithmic Market Scoring Rule__- RésuméThe success of early prediction markets have urged for the developments of market structures more adapted to beliefs elicitation. Thus, market scoring rules are more and more used in the pursuit of this goal. Although, they are often implemented in terms of asset trading, they are equivalent to a sequential probability report process that is studied in this paper. For the logarithmic market scoring rule, it is proven that the sequence of probabilities reported by agents converges under some assumptions on the regularity of the utility functions and the small variations of the risk aversion levels. In a multiple period setting, if agents update their probability beliefs with their limit reports, the sequence of probability beliefs of those who have not quoted yet converge towards a unique probability forecast, provided that their risk aversion levels are high enough. Interpreting this process as a strategic game with incomplete information, it is demonstrated that it converges towards a local equilibrium at each period and a global one which aggregates the private information hold by agents.

**Jeudi 10 octobre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**MONTEIRO Juan Pablo**(Pontificia Universidad) :__When do non-linear contracts deter the expansion of (in)efficient rivals?__- RésuméWe study the anti-competitive effects of non-linear contracts (e.g., two-part tariffs, rebates, quantity discounts) between a buyer and a dominant firm in the presence of small rivals wishing to expand. While it is known that in a rent-shifting environment with imperfect information two-part tariffs are no different than the exclusionary contracts of Aghion and Bolton, we show that alternative non-linear schemes can be remarkably different. Quantity discounts that work entirely through the spot market (such as rebates) prevent the buyer and the incumbent to commit ex-ante to the transfer of rents. The effect is so strong that these contracts are rarely anticompetitive; more so the larger the dominant firm's bargaining power and/or outside option are. The reason we see these contracts is because they can be used to prevent the expansion of inefficient rivals, or more generally, to redistribute rents ex-ante between the buyer and the dominant firm.
- Texte intégral [pdf]

**Jeudi 3 octobre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**ROKETSKIY Nikita**(University College London) :__A Quantitative Analysis of the Used Car Market__- RésuméWe quantitatively investigate the allocative and welfare effects of secondary markets for cars. An important source of gains from trade in these markets is the heterogeneity in the willingness to pay for higher-quality (newer) goods, but transaction costs are an impediment to instantaneous trade. We explore how the income distribution affects this heterogeneity---income is an important determinant of willingness to pay for quality. Calibration of the model successfully matches several aggregate features of the U.S. and French used-car markets. Counterfactual analyses show that transaction costs have a large effect on volume of trade, allocations, and the primary market. Aggregate effects on consumer surplus and welfare are relatively small, but the effect on lower-income households can be large."

**Jeudi 26 septembre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**MARTIN Daniel**(PSE) :__A Testable Theory of Imperfect Perception__- RésuméWe provide a new characterization of Bayesian expected utility maximization. The signature of the resulting theory is the impossibility of raising utility by switching wholesale from one action to another. We provide applications to robustness, to the recovery of utility from choice data, and to model classification.
- Texte intégral [pdf]

**Jeudi 12 septembre 2013 12:45-13:45**- Campus Jourdan, bâtiment G, rez-de-chaussée, salle 8
**KIMBALL Miles**(University of Michigan) :__Beyond Happiness and Satisfaction: Toward Well-Being Indices Based on Stated Preference__- RésuméThis paper proposes foundations and a methodology for survey-based tracking of well-being. First, we develop a theory in which utility depends on “fundamental aspects” of well-being, measurable with surveys. Second, drawing from psychologists, philosophers, and economists, we compile a comprehensive list of such aspects. Third, we demonstrate our proposed method for estimating the aspects’ relative marginal utilities — a necessary input for constructing an individual-level well-being index — by asking ~4,600 U.S. survey respondents to state their preference between pairs of aspect bundles. We estimate high relative marginal utilities not only for happiness and life satisfaction, but also for aspects related to family, health, security, values, and freedoms.
- Texte intégral [pdf]

**Jeudi 18 juillet 2013 12:45-13:45****RICHTER Michael**(Yeshiva University) :__Mechanism Design With Budget Constraints__- RésuméThis paper finds welfare- and revenue-maximizing mechanisms for assigning a divisible good to a population of budget-constrained agents where agents' have independently distributed private valuations and budgets. Both of these optimal mechanisms feature a linear price for the good. The welfare-maximizing mechanism additionally has a uniform lump sum transfer to all agents and a higher linear price than the revenue-maximizing mechanism. This transfer increases welfare because it relaxes the key di culty in the aforementioned setting: agents with high valuations cannot purchase an effi cient amount of the good because of their budget constraints. The welfare-maximizing result can therefore be interpreted as a version of the second welfare theorem. I show that both optimal mechanisms can be implemented using dominant strategies. In addition, I consider extensions where I relax the independence condition and introduce linear production.
- Texte intégral [pdf]

**Jeudi 13 juin 2013 12:45-13:45****FUDENBERG Drew**(Harvard) :__Recursive Stochastic Choice__**Co-author(s): Tomasz Strzalecki**- RésuméThis paper provides axiomatic characterizations of two sorts of recursive stochastic choice rules, where the agent makes his current decisions using a forward-looking value function that takes into account his future randomizations. Both of the choice rules we examine generalize logistic choice and are equivalent to it in static problems. The rules differ in how the agent views future choice sets and how he views his future randomizations. One rule is equivalent to the discounted logit used in applied work, and exhibits a preference for exibility;" the other is error-averse" and penalizes the addition of undesirable choices to a menu.
- Texte intégral [pdf]

**Jeudi 6 juin 2013 12:45-13:45****BERGES Alexis**(PSE) :__Dynamics on the electricity market : supply function equilibria and__**Co-author(s): D. Martimort**- RésuméWe investigate the influence of technology on the solutions to supply function equilibria models for the electricity market. We first stress the importance of considering adjustment costs, i.e. costs that depend not only on the quantity produced but also on the speed at which production varies: C(q, dq/dt). For linear demand and stochastic demand shocks, we then solve for a monopoly and a symmetric oligopoly. Our solutions are unique and depend on the distribution of shocks, accounting for interday and intraday supply function variation. Our results provide strong predictions on the dynamic evolution of bids on the day-ahead electricity market. We introduce a novel selection rule to choose from the continuum of equilibria typically obtained in usual supply function equilibria models, namely a solution’s stability to adjustment costs.

**Jeudi 23 mai 2013 12:45-13:45****POSTLEWAITE Andrew**(U. Penn) :__Optimism and Pessimism with Expected Utility__**Co-author(s): David Dillenberger et Kareen Rozen**- RésuméSavage (1954) provides axioms on preferences over acts that are equivalent to the existence of a subjective expected utility representation. We show that there is a continuum of other expected utility" representations in which for any act, the probability distribution over states depends on the corresponding outcomes and is

**Jeudi 16 mai 2013 12:45-13:45****GORDON Sidartha**(Science Po) :__Figures of Speech and Informative Strategic Communication__- Résumé> We provide a class of tractable communication games where each sender > type chooses a possibly truth-distorting figure of speech, which the > receiver interprets before choosing an action. Because language is > inherently vague, a figure of speech mapping determines > informativeness of communication: Exaggeration results in better > information transmission than understatement. A sender's choice of a > figure of speech trades off the action he wants to induce and his > lying aversion. There can be between one and five equilibria, all of > them fully separating, yet only partially revealing, which can be > ranked by level of informativeness. At most two of these equilibria, > the less informative ones, are ironic. The other (at most three) > equilibria are straight-talking, either exaggerating or understated. > We find that a receiver may prefer a sender who is more dissimilar to > him, and also a sender who is less honest. He may prefer a language > that has more vagueness, and a sender who is less competent. Finally, > we study the limit of the equilibria as either (i) the language > vagueness vanishes or (ii) the lying aversion vanishes.

**Jeudi 25 avril 2013 12:45-13:45****DE CORNIERE Alexandre**(Oxford) :__Integration and Search Engine Bias__**Co-author(s): Greg Taylor (Oxford)**- RésuméCompetition authorities all over the world worry that integration between search engines (mainly Google) and publishers could lead to abuses of dominant position. In particular, one concern is that of own-content bias, meaning that Google would bias its rankings in favor of the publishers it owns or has an interest in, to the detriment of competitors and users. In order to investigate this issue, we develop a theoretical framework in which the search engine (i) allocates users across publishers, and (ii) competes with publishers to attract advertisers. We show that the search engine is biased against publishers that display many ads---even without integration. Although integration may lead to own-content bias, it can also reduce bias by increasing the value of a marginal consumer to the search engine. Integration also has a positive effect on users by reducing the nuisance costs due to excessive advertising. Its net effect is therefore ambiguous in general, and we provide sufficient conditions for it to be desirable or not.
- Texte intégral [pdf]

**Jeudi 18 avril 2013 12:45-13:45****LEFEBVRE Perrin**(PSE) :__Horizontal differentiation in lobbying__- RésuméThis paper considers strategic information gathering by competing interest groups that want to influence an uninformed decision-maker. Policy decision depends on different parameters, and groups can choose which parameters to specialize into when searching for information. Common wisdom wants groups to differentiate by specializing in different dimensions in order to soften competition. I show that differentiation is indeed the only equilibrium when dimensions are equally relevant; but far from softening competition, it increases the groups search cost without guaranteeing better policies. Compared to a situation where groups concentrate on the same information, differentiation favors the decision-maker and hurts interest groups. I also characterize the different types of equilibria when the decision-maker attributes different weights to different dimensions.

**Jeudi 11 avril 2013 12:45-13:45****LISKI Matti**(Aalto University School of Economics, Helsinki) :__Resource relationships__**Co-author(s): Reyer Gerlagh**- RésuméWe consider a model of resource dependence where only the seller knows the resource reserve. The model captures phenomena such as trust in the relationship and "bribing" for continuation through generous supplies. It also explains supply shocks in equilibrium: privately informed sellers have incentives to reveal their types too late through a supply disruption after which their exploit the consumers inability to immediately adjust demand. Two puzzles that a standard exhaustible-resource theory cannot explain are resolved: sellers have an incentive to overstate their resources rather than emphasize scarcity, and buyer's can switch to alternatives before exhausting the resource thereby leaving socially valuable resource in the ground.

**Jeudi 4 avril 2013 12:45-13:45****GENDRON-SAULNIER Catherine**(Université de Montreal) :__Information Sourcing in a Strategic Environment__- RésuméMany models in economics assume that individuals are endowed with exogenous information. Yet, individuals are making choices about what to learn and from which sources of information. In this paper, we study the sourcing decisions of individuals interacting in a game where the payoff function is quadratic and the information structure is gaussian. Our game is as follows. There are two players that have to make two sequential decisions. First, each player chooses his sources of information independently of, and simultaneously with, the other player. Then each player chooses an action after observing the realization of the signals send by the sources he chose in the first step, but not the signals of the other player (unless they have chosen the same information sources). As a motivating example we consider the situation of two oligopolists competing in price on a market for a differentiated good that acquire information on their demand function. We answer the question of whether the players make the same sourcing decisions and how similar the information they acquire is. We also perform some static comparative exercises and provide some welfare analysis.

**Jeudi 21 mars 2013 12:45-13:45****DE NIJS Romain**(Hass School of Business) :__An empirical assessment of the effects of the French Hadopi law__**Co-author(s) : C. Bellégo**- RésuméThis article uses the French anti-piracy law known as Hadopi, to study the effects of online piracy on movie box office performances. Results from difference-in-differences estimations indicate that the law increased admissions in movie theaters by 13 percent in France relative to control countries. Results also show that the Hadopi effects tend to decrease over time and are particularly strong for American movies that are the most heavily pirated movies. Last, the Hadopi effects are very unlikely to be explained by supply side reactions of the industry nor by a concomitant but unrelated to Hadopi positive shock in the French market for movies in theaters.
- Texte intégral [pdf]

**Jeudi 7 mars 2013 12:45-13:45****ZHOU Junjie**(PSE) :__Economics of Leadership and Hierarchy__- RésuméThis paper explores leadership within hierarchical organizations. For each hierarchy, I consider a dynamic signaling game in which each player observes only the actions of his direct superiors before choosing his action. At the top of the hierarchy are the leaders, who learn the state from nature. The hierarchy controls the flow of information and the timing of the game, and determines the equilibrium output and welfare. I show that the welfare-optimal hierarchy is the chain, because it maximizes the incentive of players to ``lead by example'' for their subordinates. The chain remains optimal even in the presence of verifiable or unverifiable costly information acquisition by the leaders. Lastly, I characterize optimal hierarchies when the number of layers or the number of leaders is limited. Applications to fund-raising are also discussed.
- Texte intégral [pdf]

**Jeudi 28 février 2013 12:45-13:45****LECOUTEUX Guilhem**(Ecole Polytechnique) :__The formation of preferences in game theory__- RésuméFollowing Hausman's analysis of the concept of preferences (Preference, Value, Choice and Welfare, Cambridge University Press, 2012), we suggest modelling the formation of preferences in game theory, by introducing a distinction between utility --- defined as individual welfare --- and preferences --- defined as a subjective total comparative evaluation. We provide then an analytical framework where players can choose their own preferences before playing a game, and introduce the $\Sigma$ -equilibrium as a new notion of equilibrium in non cooperative game theory in order to model such a decision process. This notion of equilibrium seems to present more robust methodological foundations and also to have a better predictive power than Nash equilibrium in some standard games, and that is why we suggest it as a substitute to Nash equilibrium. We show that in a very large class of games, if we assumed that players choose their preferences in order to maximize in fine their utility, they will generally decide to maximize a different function than their own utility function: this means that for a very large class of games, maximizing one's own utility is self-defeating. We show that players choose their preferences in order to get a strategic advantage on the other players, such that the strategy that satisfies their preferences is in fact the strategy they would play if they were a Stackelberg leader. A direct corollary of this result in welfare economics is that --- if we assume that players are rational --- the preferences of the individuals are in general distinct from their utility: therefore preference satisfaction is not welfare, and we cannot easily infer individual welfare from choices and preferences.

**Jeudi 21 février 2013 12:45-13:45****SCHUTZ Nicolas**(University of Mannheim) :__Cross-Border Price Effects of Mergers and Acquisitions - A Quantitative Framework for Competition Policy__**Co-auteurs : Holger Breinlich and Volker Nocke**- RésuméDecisions of national competition authorities have important effects on other jurisdictions. We provide a framework to quantify the domestic and cross-border effects of mergers, and to draw conclusions for the coordination of national merger policies. We develop a two-country model with many sectors. In each sector, producers vary in terms of their marginal costs, and are engaged in Cournot competition. We allow for pro?table mergers to take place subject to the non-violation of a given national competition policy. Because of trade costs and perceived differences in qualities between domestic and foreign products, mergers may have different consumer surplus effects in the home and the foreign country. We calibrate the model using data for the year 2002 for 167 manufacturing sectors in the U.S. and Canada, two well-integrated markets where cross-border effects of M&As are likely to be important. We choose parameters to match relevant moments in the data, including industry sales, concentration ratios and trade ?ows. We ?nd that merger decisions taken in isolation by national competition authorities lead to ine cient outcomes compared to a scenario where competition policy is coordinated, and that these ine ciencies are quantitatively important. While there is considerable variation across industries, we ?nd that in the majority of industries a merger approval policy based on domestic consumer surplus is too restrictive from the viewpoint of the neighboring country.
- Texte intégral [pdf]

**Jeudi 14 février 2013 12:45-13:45****LASLIER Jean-François**(Ecole Polytechnique) :__Vote au pluriel: How people vote under different voting rules__- RésuméThe seminar presents a survey of recent research on voter behavior. Using traditional laboratory experiments, innovative "in situ" experiments, and on-line surveys, we observe how people vote under different voting rules. The goal of the investigation is two-fold. In an Institutional Design perspective, we want to understand the consequences of possible political institutions (who is elected ?). In a Political Psychology perspective, we want to understand the individuals' conception of the act of voting. The set of candidates being given, we describe what kind of candidates are favored by what kind of rules. As to voters, we find that they use heuristics and behave partly strategically and partly sincerely, the mixture depending on the detailled circumstances. We note that voters have both self-serving and ideological preferences over voting rules.

**Jeudi 7 février 2013 12:45-13:45****GARRETT Daniel**(TSE) :__Dynamic Mechanism Design with Dynamic Arrivals and Changing Values__- RésuméI will discuss dynamic mechanism design settings where agents preferences evolve stocastically and where agents arrive over time. Agents need not participate at the arrival date and can be strategic with respect to the first date of participation in the mechanism. Profit-maximizing allocations, which maximize a “dynamic virtual surplus” are derived. Applications to the durable goods problem (see `Incoming Demand with Private Uncertainty’) and overselling of capacity in the airline industry (see `Overbooking’ with Jeff Ely and Toomas Hinnosaar) will be discussed.

**Jeudi 24 janvier 2013 12:45-13:45****DAVE Chetan**(NYU Abu Dhabi) :__Confirmation Bias with Motivated Preferences__**co-authored with Gary Charness**- RésuméModels of confirmation bias as an individual decision making error predict systematic departures from Bayesian updating and can imply that no learning may ever take place. While some experimental evidence for a confirmation bias has been found, an implicit assumption in many stochastic economic models is that strategic (or market) pressures cause individuals to not exhibit systematic deviations from Bayesian updating. We investigate in the laboratory the effect of subtle background strategic considerations on signal extraction, finding that people are affected by the presence of motivated preferences (here a preference for one state having manifested). Our results suggest that players with motivated preferences over probabilistic end states deviate less from Bayesian updating. However, such players still exhibit a confirmation bias when the directions of updates are analyzed.

**Jeudi 10 janvier 2013 12:45-13:45****JEONG HEO Eun**(University of Rochester) :__Probabilistic Assignment Problem with Multi-unit Demands: A Generalization of the Serial Rule and a Characterization__- RésuméAbstract: We study a probabilistic assignment problem when agents have multi-unit demands for objects. We first introduce two fairness requirements to accommodate di erent demands across agents and show that each of these requirements is incompatible with stochastic dominance efficiency (henceforth, we use the prefix sd" for stochastic dominance) and weak sd-strategyproofness, unless all agents have unitary demands. We next introduce a new incentive requirement which we call limited invariance. We explore implications of these requirements, together with consistency and converse consistency. Our main result is that the generalized serial rule, which we propose as an adaptation of the serial rule to our setting, is the only rule satisfying sd-efficiency, the sd proportional-division lower-bound, limited invariance, and consistency. Uniqueness persists if we replace the sd proportional-division lower-bound by sd normalized-no-envy, or consistency by converse consistency, or both. The serial rule in Bogomolnaia and Moulin (2001) is characterized as a special case of our generalized serial rule.

**Jeudi 13 décembre 2012 12:45-13:45****WREN-LEWIS Liam**(PSE) :__Hold-up problems in international electricity trade__- RésuméThis paper investigates the underdevelopment of electricity trade between African coun- tries. I build a model where hold-up problems originating from incomplete contracts decrease countries' willingness to trade. In particular, an inability to commit to future trade reduces the specialisation of investment. The paper then uses the model to explore potential policy solutions including deregulation and the liberalisation of ownership.

**Jeudi 6 décembre 2012 12:45-13:45****CANTALLA David**(Centro de Estudios Economicos, El Colegio de Mexico) :__Housing Markets with Endogenous Budget Constraints__**Co-auteur : Juan Sebastian Pereyra**- RésuméThis paper addresses the existence of competitive equilibrium in the variant of the Assignment Game where budget-constrained agents are endowed with one good, that can be sold to buy another good. We de?ne the Money Scarcity (MS) condition and its Strong (SMS) and Weak (WMS) versions. The WMS condition ensures the existence of an equilibrium in cases where Quinzii’s indis- pensability of money assumption does not hold. We consider two mechanisms to ?nd a competitive equilibrium which are extended versions of Gale’s Top Trad- ing Cycles algorithm and of the Exact Auction Mechanism of Demange, Gale and Sotomayor. Under MS, the ?rst mechanism is strategy-proof; the second is not. We strengthen the MS condition (SMS) to guarantee the existence of Nash equilibria in the revelation game induced by the second mechanism, where the outcome coincides with the assignment found by the ?rst mechanism; moreover, this assignment is Pareto e?cient.

**Jeudi 29 novembre 2012 12:45-13:45****POZZI Andrea**(Einaudi Institute for Economics and Finance) :__The Effect of Internet Distribution on Brick-and-mortar Sales__- RésuméI study the introduction of an online shopping service for a large supermarket chain that also operates a wide network of brick-and-mortar stores. The establishment of the Internet channel led to a 13 percent increase in overall revenues, inducing only limited cannibalization on traditional sales. I provide insights on the mechanism behind this result. I show that selling on the web removes travel costs and helps the retailer to attract new business from customers living far away from its stores. I also document that revenues increase more in markets where the chain faces more competitors. This suggests that the new sales represent in part business diverted from rival supermarkets.

**Jeudi 15 novembre 2012 12:15-13:45****MARFAN Manuel**(PSE) :__Rules versus discretion under limited commitment__- RésuméThis paper studies decision making by an uninformed authority that will be better informed in the future, and whose expected actions will affect a third party’s decision making. The trade-off between rules and discretion is studied in this setting, when information is not verifiable by the third party, thus inducing the authority to misreport it. A necessary and a sufficient condition for continuity of the decision rule as a function of the information parameter are given, and under some structural conditions the optimal mechanism is characterized.

**Jeudi 25 octobre 2012 12:45-13:45****SPECTOR David**(PSE) :__"Non-verifiable self-reporting and collusion"__- RésuméMany cartels involve companies reporting their sales to their competitors in a non-verifiable way, long before individual sales information becomes public. This paper assesses how the possible collusive outcomes depend on the possibility of engaging in such cheap talk. It is shown that for some parameter values, monopoly prices can be sustained in equilibrium only if such cheap talk is possible. The reason is that if communication is precluded, the occurrence of wasteful price wars may be the only way to deter deviations, whereas the possibility of communicating frequently, even if only by issuing non-verifiable reports, allows companies to fine-tune more efficient disciplining devices such as voluntary, temporary self-exclusion.

**Jeudi 11 octobre 2012 12:45-13:45****HEINRICH Nax**(PSE) :__"The Evolution of Core Stability in Decentralized Matching Markets"__**Co-auteur : Peyton Young, Bary Pradelski**- RésuméDecentralized matching platforms on the internet allow large numbers of agents to interact anonymously at virtually no cost. Very little information is available to market participants and trade takes place at many different prices simultaneously. We propose a decentralized learning process in such environments that leads to stable and efficient outcomes. Agents on each side of the market make demands of potential partners and are matched if their demands are mutually compatible. Matched agents occasionally experiment with higher demands, while unmatched agents lower their demands in the hope of attracting partners. This learning process implements core allocations even though agents have no knowledge of other agents’ strategies, payoffs, or the structure of the game, and there is no central authority with such knowledge either.
- Texte intégral [pdf]

**Jeudi 27 septembre 2012 12:45-13:45****WILNER Lionel**(Crest -LEI) :__"Do consumers correctly expect price reductions? Testing dynamic behavior"__**Co-auteur : Philippe Février**- RésuméBoth theoretical and empirical dynamics literature require agents to solve complex dynamic programs given their information set, which assumes implicitly that agents are fully rational and have perfect expectations. We claim that this assumption is easily testable provided that market-level data on prices and purchases are available. Using data on albums exhibiting frequent episodes of sales, we perform the test and find that consumers hold simple expectations on the timing of sales: everything happens as if consumers were expecting a Markov price process. On the contrary, the firm is more sophisticated and does not follow such a policy. Our results are coherent with the idea that agents rationality is bounded because of limited memory or limited capacity. These results have important implications in terms of demand estimation, firms' optimal pricing strategy and welfare.

**Jeudi 20 septembre 2012 12:45-13:45****WEINSTEIN Jonathan**(Northwestern University) :__"A Bayesian Foundation for Classical Hypothesis Testing."__- RésuméA decision-maker can ensure dynamic consistency by following Bayes' rule, but he may wish to balance such consistency against other goals. That is, when the decision-maker is surprised by a pattern unaccounted for in his prior, he may wish to change his beliefs in a way which violates Bayes' rule, but he may also wish to limit his inconsistency. We show that if such non-Bayesian events, or \paradigm shifts," are rare, in the sense that they occur only with a small probability according to the decision-maker's initial belief, the decision-maker will be \approximately" dynamically con- sistent. Our notion of \approximate" dynamic consistency is that the possible arbitrage against the decision-maker is small compared to the size of his transactions. The quantity is equivalent to the level of a classical hypothesis test, so our results provide a decision-theoretic foundation for the classical criteria for rejecting a null hypothesis. Our ndings give the decision-maker some latitude to revise his model while bounding the pain of inconsistency, and unify the classical and Bayesian modes of inference.
- Texte intégral [pdf]

**Jeudi 5 juillet 2012 12:45-13:45****SALOMON Antoine**(Imagine, Ecole des Ponts ParisTech) :__Strategic learning and bandit games__

**Jeudi 14 juin 2012 12:45-13:45****TESCHL Miriam**(University of Vienna) :__Theory of choice under internal conflict__

**Jeudi 31 mai 2012 12:45-13:45****LEFEBVRE Perrin**(PSE) :__Downstream lobbying and informational frictions__

**Jeudi 24 mai 2012 12:45-13:45****MICHELUCCI Fabio**(CERGE-EI, Pragues) :__The English Auction, Rushes, and a Selaed Bid Efficient Auction__- Texte intégral [pdf]

**Jeudi 10 mai 2012 12:45-13:45****KRANTON Rachel**(Duke University and PSE) :__Striving for Social Status in Networks__

**Jeudi 5 avril 2012 12:45-13:45****KUSHNIR Alexey**(University of Zurich) :__A geometric approach to mechanism design__

**Jeudi 29 mars 2012 12:45-13:45****VANASCO Victoria**(UC Berkeley and PSE) :__Learning by Lending: The Informational Role of Financial Intermediaries__

**Jeudi 22 mars 2012 12:45-13:45****LACLAU Marie**(HEC Paris) :__Communication in repeated network games with private monitoring__

**Jeudi 15 mars 2012 12:45-13:45****KANDORI Michihiro**(University of Tokyo) :__Asynchronous Revision Games__- Texte intégral [pdf]

**Jeudi 8 mars 2012 12:45-13:45****OZGUR Onur**(Université de Montréal) :__Dynamic Linear Economies with Social Interactions__

**Jeudi 16 février 2012 12:45-13:45****TRUYTS Tom**(U. Leuven) :__Stochastic Signaling: Information Substitutes and Complements__

**Jeudi 9 février 2012 12:45-13:45****SPAETER Sandrine**(BETA, Univ. Strasbourg) :__The Prudent Principal__**Co-author(s): Bernard Sinclair-Desgagné**- Texte intégral [pdf]

**Jeudi 2 février 2012 12:45-13:45****KOLM Julian**(Université de Vienna) :__Learning Across Subgames: An Application to the Hold-up Problem"__- Texte intégral [pdf]

**Jeudi 12 janvier 2012 12:45-13:45****GUESNERIE Roger**(PSE) :__Eductive Stability in Real Business Cycle Models__**Co-author(s): G. Evans & B. McGough**- Texte intégral [pdf]

**Jeudi 5 janvier 2012 12:45-13:45****CASELLA Alessandra**(University Columbia) :__Vote trading with and without party leaders__**Co-author(s) Tom Palfrey & Sebastien Turban**- Texte intégral [pdf]

**Jeudi 15 décembre 2011 12:45-13:45****CHASSAGNON Arnold**(U. Tours & PSE) :__Multiple lenders, strategic default and the role of debt covenants__

**Jeudi 1er décembre 2011 12:45-13:45****GUERON Yves**(UCL ) :__Failure of Gradualism under Imperfect Monitoring__

**Jeudi 17 novembre 2011 13:00-13:45****MARCH Christoph**(PSE ) :__Do We Follow Others when We Should? A Simple Test of Rational Expectations__**Attention exceptionnellement ce séminaire commencera à 13h**- RésuméWeizsäcker (2010) suggests a simple counting technique to test the rational-expectations hypothesis in experimental datasets. The counting technique is applied to data on information cascade games and regression results strongly reject the rational-expectations hypothesis. Weizsäcker (2010) concludes that participants perform poorly when trying to extract the information that is contained in others' choices. We replace participants' choices with equilibrium choices in the cascade game dataset and we show that the counting technique sometimes leads to the erroneous conclusion that players follow their signal more often than they follow others even when the latter is the optimal decision. We correct the counting technique and we conclude that, in situations where it is empirically optimal to follow others, participants do so in the large majority of cases. We conrm that the rational-expectations hypothesis is rejected in cascade game experiments.

**Jeudi 10 novembre 2011 12:45-13:45****RUIZ-ALISEDA Francisco**(Ecole Polytechnique ) :__The Dynamics of Protection and Imitation of Innovations__**with Emeric Henry**- RésuméMost innovators choose not to patent their inventions. This is somewhat of a puzzle for those who view innovators as at the mercy of imitators in the absence of legal protection. In practice, innovators invest actively in making their products technologically hard to reverse-engineer. In this paper, we consider the dynamics of imitation and protection of innovations, and we find that innovators can still obtain high profits in the absence of legal protection. The protection technologies that yield the highest profits for the innovator are expensive and do not protect well. Our model also allows us to draw conclusions on the design of patent policy and on the dynamics of employment and mobility of researchers in innovative industries.

**Jeudi 20 octobre 2011 12:45-13:45****KOESSLER Frederic**(PSE) :__Extortion and Political Risk Insurance in Weak Governance Countries__

**Jeudi 13 octobre 2011 12:45-13:45****BIANCHI Milo**(U. Paris-Dauphine) :__Creative Accounting and Efficient Markets__

**Jeudi 6 octobre 2011 12:45-13:45****DE CORNIERE Alexandre**(PSE) :__Search Advertising__

**Jeudi 29 septembre 2011 12:45-13:45****COLLIARD Jean-Edouard**(PSE) :__Rational Blinders: Is it Possible to Regulate Banks Using their Internal Risk Models?__

**Jeudi 23 juin 2011 12:45-13:45****MARCH Christoph**(PSE) :__Individual Learning and Strategic Thinking in Laboratory Cascades__

**Jeudi 9 juin 2011 12:45-13:45****KORIYAMA Yukio**(Ecole Polytechnique) :__Optimal Apportionment__**Co-auteur(s) : Jean-François Laslier**- Texte intégral [pdf]

**Jeudi 19 mai 2011 12:45-13:45****BEHRINGER Stefan**(University of Bonn) :__Price Wars in Two-Sided Markets: The case of the UK Quality Newspapers__**Co-auteur(s) : Lapo Filistrucchi**- Texte intégral [pdf]

**Jeudi 5 mai 2011 12:45-13:45****LAMY Laurent**(PSE) :__On the use of absolute auctions and secret reserve prices__**Co-auteur(s) : Philippe Jehiel**

**Jeudi 28 avril 2011 12:45-13:45****SAND-ZANTMAN Wilfried**(Toulouse School of Economics) :__Contracting with Bypass in Bilateral Relationship__**with Bruno Jullien and Jerome Pouyet**

**Jeudi 7 avril 2011 12:45-13:45****LIMARDI Michela**(PSE) :__Nongovernmental Regulation: NGO Monitoring and Firm Compliance with Social Standards__

**Jeudi 31 mars 2011 12:45-13:45****BRAOUEZEC Yann**(Ecole Supérieure d’Ingénieurs Léonard de Vinci) :__Some theory of constrained third-degree price discrimination__

**Jeudi 24 mars 2011 12:45-13:45****BRAMOULLE Yann**(Université Laval) :__Favoritism__- Texte intégral [pdf]

**Jeudi 10 mars 2011 12:45-13:45****LEVY Raphaël**(University of Mannheim) :__Humouring both parties: a model of two-sided reputation__

**Jeudi 3 mars 2011 12:45-13:45****TUKIAINEN Janne**(University of Helsinki) :__The Effect of Minimum Bid Increment on Revenue in Internet Auctions: Evidence from a Field Experiment__- Texte intégral [pdf]

**Jeudi 10 février 2011 12:45-13:45****LEVY Raphaël**(U. Mannheim - TBA) :__*__

**Jeudi 3 février 2011 12:45-13:45****GOSSNER Olivier**(PSE) :__Entropy and the value of information for investors__**Co-auteur(s) : Antonio Cabrales & Roberto Serrano**- Texte intégral [pdf]

**Jeudi 20 janvier 2011 12:45-13:45****MORENO DE BARREDA Ines**(LSE) :__Cheap Talk With Two-Sided Private Information__- RésuméI investigate the strategic interaction between a perfectly informed expert and a decision maker when the latter has imperfect private information relevant to the decision. To analyse the effect of the decision maker's information, I extend the Crawford and Sobel (1982) canonical model of cheap talk by allowing the decision maker to access an unbiased and symmetric signal about the state of the world. I first show that for symmetric preferences partition equilibria exist in this more general environment. Second, for quadratic-loss preferences, I show that access to private information might hamper communication. Surprisingly, in a wide range of environments, the decision maker private information cannot make up for the loss in communication implying that the welfare of both agents decreases.
- Texte intégral [pdf]

**Jeudi 6 janvier 2011 12:45-13:45****BIANCINI Sara**(Thema, Cergy) :__Intellectual Property Rights Adoption in Developing Countries__

**Jeudi 9 décembre 2010 12:45-13:45****KOVBASYUK Sergei**(TSE and CREST) :__Optimal Certification Design__

**Jeudi 2 décembre 2010 12:45-13:45****GODEFROY Raphael**(PSE) :__Choosing Choices: Agenda Selection with Uncertain Issues__- Texte intégral [pdf]

**Jeudi 25 novembre 2010 12:45-13:45****DE CORNIERE Alexandre**(PSE & CREST-ENSAE) :__Online Advertising and Privacy__**with Romain de Nijs**

**Jeudi 28 octobre 2010 12:45-13:45****VIDA Peter**(University of Vienna) :__Collusion communication schemes in first price auction__

**Jeudi 14 octobre 2010 12:45-13:45****COLLIARD Jean-Edouard**(PSE) :__The Impact of Private Information about Positive Feedback Trading__

**Jeudi 7 octobre 2010 12:45-13:45****PAWLOWITSCH Christina**(PSE) :__Why evolution does not necessarily lead to an optimal signaling system__- Texte intégral [pdf]

**Jeudi 30 septembre 2010 12:45-13:45****MARCH Christopher**(PSE) :__Adaptive Social Learning__- Texte intégral [pdf]

**Mardi 8 juin 2010 12:45-13:45****BRESKY Michal**(CERGE-EI, Prague) :__Revenue and Efficiency in Multi-Unit Uniform-Price Auctions__

**Jeudi 3 juin 2010 12:45-13:45****PICARD Pierre**(Ecole Polytechnique: TBA) :__*__

**Jeudi 27 mai 2010 12:45-13:45****DE NIJS Romain**(PSE) :__Dynamic Targeted Pricing in Online Markets__

**Jeudi 20 mai 2010 12:45-13:45****CRAWFORD Vincent**(Oxford) :__New York City Cabdrivers’ Labor Supply Revisited: Reference-Dependent Preferences with Rational-Expectations Targets for Hours and Income__**with Juanjuan Meng**- Texte intégral [pdf]

**Jeudi 13 mai 2010 12:45-13:45****BRESKY Michal**(CERGE-EI, Prague) :__Revenue and Efficiency in Multi-Unit Uniform-Price Auctions__

**Jeudi 6 mai 2010 12:45-13:45****LASLIER Jean-François**(Ecole Polytechnique) :__Stubborn learning__**with Bernard Walliser**- Texte intégral [pdf]

**Jeudi 8 avril 2010 12:45-13:45****La séance a été annulée.****LAMY Laurent**(PSE : TBA) :__*__

**Jeudi 25 mars 2010 12:45-13:45****JANSEN Jos**(Max Planck Institute ) :__Share To Scare: Technology Sharing in the Absence of Intellectual Property Rights__- Texte intégral [pdf]

**Jeudi 11 mars 2010 12:45-13:45****GOSSNER Olivier**(PSE : TBA) :__A Folk Theorem in Robust Equilibrium Structures.__

**Jeudi 11 mars 2010 12:45-13:45**- *

**Jeudi 18 février 2010 12:45-13:45****FERAL Arnaud**(U. Cergy : TBA) :__Merger Control with Transfers from the Capital Gains Tax__

**Jeudi 4 février 2010 12:45-13:45****PERONA Mathieu**(PSE) :__How broadcasting quotas harm program diversity__

**Jeudi 28 janvier 2010 12:45-13:45****MARTIMORT David**(TSE) :__Public Contracting in Delegated Agency Games__**with Lars Stole - David Martimort will be at PSE-Jourdan on thursday (building E), please fill this form if you want to meet him**

**Jeudi 21 janvier 2010 12:45-13:45****TOXYAERD Flavio**(University of Cambridge) :__The Economics of Infectious Disease__

**Jeudi 10 décembre 2009 12:45-13:45****HARSTAD Ron**(University of Missouri) :__Two Examples of Why I've Worked with Mike Rothkopf : "Winner's Curse Corrections Magnify Adverse Selection" (joint with Robert Bordley) and "Auctioning Rights to Choose When Competition Persists"__- Texte intégral [pdf]

**Jeudi 26 novembre 2009 12:45-13:45****ROGERS Brian**(Northwestern) :__Diversity and Popularity in Social Networks (avec Yann Bramoulle).__**Brian Rogers sera à Jourdan la journée, contactez moi si vous voulez le rencontrer, ou complétez ceci: http://www.doodle.com/6paq5gc44vys3hzi**- Texte intégral [pdf]

**Jeudi 12 novembre 2009 12:45-13:45****TERCIEUX Olivier**(PSE) :__Common p-Beliefs and the Hold-Up Problem (avec Philippe Aghion, Drew Fudenberg, Richard Holden et Takashi Kunimoto)__- Texte intégral [pdf]

**Jeudi 15 octobre 2009 12:45-13:45****WALLISER Bernard**(PSE) :__La révision des croyances multi-agents (avec JC Vergnaud et A Billot)__

**Jeudi 15 octobre 2009 12:45-13:45****WALLISER Bernard**(PSE) :__La révision des croyances multi-agents (avec JC Vergnaud et A Billot)__

**Jeudi 1er octobre 2009 12:45-13:45****FLECKINGER Pierre**(PSE) :__The Incentive Value of Deadlines__

**Jeudi 1er octobre 2009 12:45-13:45****FLECKINGER Pierre**(PSE) :__The Incentive Value of Deadlines__

**Jeudi 24 septembre 2009 12:45-13:45****MENAGER Lucie**(Paris 2) :__Intimidation strategies in competitive bidding__

**Jeudi 24 septembre 2009 12:45-13:45****MENAGER Lucie**(Paris 2) :__Intimidation strategies in competitive bidding__

**Jeudi 4 juin 2009 12:45-13:45****COLLIARD Jean-Edouard**(PSE) :__Competition between limit order markets and trading fees__

**Jeudi 28 mai 2009 12:45-13:45****MAILTAH George**(University of Pennsylvania) :__Common Learning in the Presence of Dependence__

**Jeudi 14 mai 2009 12:45-13:45****KRANTON Rachel**(University of Maryland) :__Networked Teams and Public Goods: Who Does the Work?__

**Jeudi 23 avril 2009 12:45-13:45****La séance a été annulée.****ALKAN Ahmet**(Sabanci University) :__Stable Matching Through Stable Shortlists__

**Jeudi 9 avril 2009 12:45-13:45****RENY Philip**(University of Chicago) :__On the Existence of Monotone Pure Strategy Equilibria in Bayesian Games.__- Texte intégral [pdf]

**Jeudi 2 avril 2009 12:45-13:45****La séance a été annulée.****PERONA Mathieu**(PSE) :__*__

**Jeudi 26 mars 2009 12:45-13:45****MIERENDORFF Konrad**(University of Bonn) :__Optimal Dynamic Mechanism Design with Deadlines__

**Jeudi 19 mars 2009 12:45-13:45****PICCOLO Salvator**(University of Salerno) :__Multiple-Bank Lending, Creditor Rights and Information Sharing.__- Texte intégral [pdf]

**Jeudi 5 mars 2009 12:45-13:45****GORDON Sidartha**(University of Montreal) :__Iteratively Stable Equilibria in Cheap Talk Games__

**Jeudi 26 février 2009 12:45-13:45****KUNIMOTO Takashi**(McGill University) :__Robust Virtual Implementation with Incomplete Information: Towards a Reinterpretation of the Wilson Doctrine__

**Jeudi 19 février 2009 12:45-13:45****SEKIGUCHI Tadashi**(Kyoto University) :__Repeated Games with Costly Imperfect Monitoring (joint with Eiichi Miyagawa and Yasuyuki Miyahara)__- Texte intégral [pdf]

**Jeudi 12 février 2009 12:45-13:45****MATHIS Jérôme**(TSE) :__Rating the Raters: Are Reputation Concerns Powerful enough to Discipline Rating Agencies?__

**Jeudi 5 février 2009 12:45-13:45****LAMBERT-MOGILIANSKY Ariane**(PSE) :__TI-games I: An Exploration of Type Indeterminacy in Strategic Decision-making__- Texte intégral [pdf]

**Jeudi 29 janvier 2009 12:45-13:45****DE CORNIERE Alexandre**(PSE) :__On the Optimal Design of a Search Engine's Advertising Platform__

**Jeudi 22 janvier 2009 12:45-13:45****BLOCH Francis**(Polytechnique) :__Optimal assignment of reassignable objects__

**Jeudi 15 janvier 2009 12:45-13:45****La séance a été annulée.****DE CORNIERE Alexandre**(PSE) :__On the Optimal Design of a Search Engine's Advertising Platform__

**Jeudi 11 décembre 2008 12:45-13:45****LOVO Stefano**(HEC) :__Belief-free Equilibria in Games with Incomplete Information: The N-player Case__

**Jeudi 4 décembre 2008 12:45-13:45****ETTINGER David**(Univ. Cergy) :__Deception in a Repeated Expert/Agent Interaction: Theory and Experiment__

**Jeudi 27 novembre 2008 12:45-13:45****OYAMA Daisuke**(Hitotsubashi Univ.) :__On Ex Post Individually Rational Partnership Dissolution__

**Jeudi 13 novembre 2008 12:45-13:45****NUNEZ Matias**(Ecole polytechnique) :__A Study of Approval Voting on Large Poisson Games__

**Jeudi 23 octobre 2008 12:45-13:45****La séance a été annulée.****PICCOLO Salvatore**(Univ. of Salerno) :__The strategic value of quantity forcing contracts__

**Jeudi 16 octobre 2008 12:45-13:45****CASELLA Alessandra**(Columbia Univ.) :__Storable Votes and Agenda Control. Theory and Experiments__

**Jeudi 2 octobre 2008 12:45-13:45****LAMY Laurent**(PSE) :__Mechanism Design with Partially-Specified Participation Games__

**Jeudi 18 septembre 2008 12:45-13:45****FALLY Thibault**(PSE) :__Global Sourcing under imperfect capital markets__

**Jeudi 29 mai 2008 12:45-13:45****DENG X.**(City univ. of Hong Kong) :__Sales of Clicks and Auction Theory__- RésuméSponsored search has become a principal source of revenue for search engine companies. The business model relies on advertisement sales that provide advertisers opportunities to introduce their products directly to potential customers. Pricing of such advertising positions has been vitally important for the search engines. The generalized second price auction has turned out to be the primary protocol for such sponsored search markets. While several Nash equilibria may exist for the generalized second price auction, we propose the concept of the forward looking Nash equilibrium as the bidder's individual rational manipulation outcome. We further identify and focus on traffic arbitrage and click arbitrage possibilities crossing multiple markets as well as derive and characterize equilibria across markets.

**Jeudi 15 mai 2008 12:45-13:45****KOESSLER F.**(PSE) :__Multidimensional Communication Mechanisms: Cooperative and Conflicting Designs__**Co-auteur : D. Martimort**- Texte intégral [pdf]

**Jeudi 17 avril 2008 12:45-13:45****La séance a été annulée.****KOESSLER Frédéric**(PSE) :__Multidimensional Communication Mechanisms: Cooperative and Conflicting Designs__**David Martimort**

**Jeudi 10 avril 2008 12:45-13:45****VAN DER STRAETEN Karine**(PSE) :__A Communication Game on Electoral Platforms__**Gabrielle Demange**

**Jeudi 20 mars 2008 12:45-13:45****DESSEIN W.**(Univ. of Chicago) :__Organize to Compete__**R. Alonso and N. Matouschek**

**Jeudi 21 février 2008 12:45-13:45****MOIZEAU F.**(Toulouse School of Economics) :__Dynamic Regulation of Quality__

**Jeudi 14 février 2008 12:45-13:45****La séance a été annulée.****LAMBERT-MOGILIANSKY A.**(PSE) :__*__

**Jeudi 31 janvier 2008 12:45-13:45****GOSSNER O.**(PSE) :__Efficiency in repeated prisoner's dilemma without conditional independence__**Co-auteurs : K. Fong, J. Horner, et Y. Sannikov**

**Jeudi 24 janvier 2008 12:45-13:45****BIRAN O.**(Univ. de Paris 9) :__Efficiency and the Final Consumer in Resale Markets with Externalities__

**Jeudi 10 janvier 2008 12:45-13:45****OURY M.**(PSE) :__Continuous Implementation__**Co-auteur : O. Tercieux**- Texte intégral [pdf]

**Jeudi 20 décembre 2007 12:45-13:45****RAYO L.**(Univ. of Chicago) :__Status, Market Power, and Veblen Effects__**Co-auteurs : Miguel Diaz, Haresh Sapra**- Texte intégral [pdf]

**Jeudi 13 décembre 2007 12:45-13:45****SAURI ROMERO L.**(European univ. institute) :__Parallel trade and incentives to innovate when governments decide on prices__

**Jeudi 29 novembre 2007 12:45-13:45****RENAULT R.**(Univ. de Cergy-Pontoise) :__Gathering and disseminating information in discrete choice models__

**Jeudi 15 novembre 2007 12:45-13:45****HAGENBACH J.**(Univ. de Paris 1) :__Strategic Communication Networks__

**Jeudi 25 octobre 2007 12:45-13:45****COMOLA M.**(Univ. Pompeu Fabra) :__The Network Structure of Informal Arrangements : Evidence from Rural Tanzania__

**Jeudi 4 octobre 2007 12:45-13:45****MIERENDORFF K.**(Univ. Bonn) :__An efficient Intertemporal Auction__

**Jeudi 14 juin 2007 12:45-13:45****SCHULTZ N.**(PSE) :__Anticompetitive Vertical Mergers Waves__**Co-auteur (s) : J. Hombert et J. Pouyet**- RésuméWe develop an equilibrium model of vertical mergers. We show that, when a wave of mergers removes all upstream firms, the competitive forces on the upstream market may collapse. Indeed, when an integrated firm supplies the upstream market, it internalizes the fact that customers lost on the downstream market can be recovered via the upstream market. Thus, the upstream supplier charges higher downstream prices. Its integrated rivals benefit from this behavior, they may therefore not undercut on the upstream market. This mechanism leads to anticompetitive waves of mergers. Journal of Economic Literature Classification Number: L22, L13, L42. Keywords: vertical merger, vertical integration, foreclosure.
- Texte intégral [pdf]

**Jeudi 7 juin 2007 12:45-13:45****AUSUBEL L. M.**(Univ. of Maryland) :__Walrasian Tatonnement for Discrete Goods__

**Jeudi 24 mai 2007 12:45-13:45****TAKAHASHI S.**(Harvard univ.) :__Community Enforcement when Players Observe Partners’ Past Play__- RésuméI investigate whether a community can sustain cooperation in the repeated prisoner’s dilemma by having cheaters sanctioned not by their victims but by third parties. Motivated by systems of credit history recording, online feedback systems, and some experimental settings, I assume that players can access information about their partners’ past play for free, but that acquiring information about their partners’ past partners’ past play is prohibitively costly. In this setting, even though players cannot distinguish cheaters from those who punish cheaters, I show that any level of cooperation can be sustained by an equilibrium. The equilibrium I construct has the following two properties: every player chooses his actions independently of his own record of play, and he is indifferent between cooperation and defection at all histories. This equilibrium carries over to the finite-population setting and is robust to noise in the process of choosing actions or of recording past play. The technique of equilibrium construction is applied to more general stage games. I also analyze the possibility of cooperation either when players are required to have strict incentives to follow equilibrium strategies or when only summary statistics of records are stored in the community.
- Texte intégral [pdf]

**Jeudi 10 mai 2007 12:45-13:45****ROEMER J. E.**(Yale univ.) :__The political economy of income taxation in a two-party democracy__- RésuméParties compete on a large policy space. Only a fraction of each voter type will vote for each party, perhaps because of party reputations, or issues not modeled here. Each party’s policy makers comprise two factions, one concerned with maximizing the welfare of its constituency, the other with maximizing vote share. These factions are particularly concerned with the party’s base or core voters and the swing voters, respectively. All these concepts (constituency, core, swing) are endogenous to the policy choice. An application to competition over redistribution produces equilibria in which each party proposes a piece-wise linear tax schedule, and these schedules coincide in their treatment of a possibly large interval of middle-income voters. This appears to conform with recent tax history in the US.
- Texte intégral [pdf]

**Jeudi 26 avril 2007 12:45-13:45****BOS O.**(PSE) :__Charity Auctions for the Happy Few__- RésuméRecent literature has shown that all-pay auctions raise more money for charity than winnerpay auctions. We demonstrate that the first and second-price winner-pay auctions outperform first-price all-pay auctions when bidders are sufficiently asymmetric. To prove it, we consider a framework with complete information. This analysis is relevant for two main reasons. On the one hand, complete information is more realistic and corresponds to events which occur for instance in a local service club (like in a voluntary organization) or in a show business dinner. Potential bidders are acquaintances or know one another well. On the other hand, our model keeps the qualitative predictions of a private value model under incomplete information in which bidders are ex ante asymmetric that is to say different bidders’ values are drawn from different distributions. Furthermore, we also analyze second-price all-pay auction. Finally, we show that individual minimum bids could improve the relative revenue performance of first-price allpay compared to first-price winner-pay auction. Keywords: All-pay auctions, charity, complete information, externalities JEL Classification: D44, D62, D64

**Jeudi 5 avril 2007 12:45-13:45****LAURENT R. A.**(PSE) :__Duopole avec biens d'expérience et consommateurs raisonnant par similarité__- RésuméDeux firmes sont présentes sur le marché d’un bien d’expérience mais ne peuvent signaler leur "qualité". L’utilité nette attendue d’un bien par les consommateurs dépend de leur degré d’exigence et d’une information privée, plus ou moins précise. Ces consommateurs choisissent le bien qui maximise l’utilité nette pondérée par la similarité entre l’information privée et le bien considéré : ce critère de décision est une version simplifiée et statique de la "Case-Based Decision Theory" (Gilboa et Schmeidler, 1995). L’introduction des prix dans cette règle de décision permet d’établir des fonctions de demande dans un cadre spatial. Cet article montre qu’un équilibre de Nash en prix existe en fonction de la forme de l’utilité nette et du degré d’asymétrie des coûts. Lorsque les consommateurs sont peu exigeants, la différenciation est purement horizontale : chaque firme sert les consommateurs informés sur son bien. Si les consommateurs sont relativement exigeants, une dimension verticale est introduite dans la différenciation : la firme vendant le bien le moins apprécié choisit un prix faible et attire les consommateurs dont l’information est la moins précise. Les autres consommateurs achètent le bien concurrent. Toutefois, si les consommateurs sont très exigeants, cette dimension verticale disparaît. classification jel : D11, D43, L13. mots clés : différenciation des produits, oligopole, biens d’expérience, information privée, théorie de la décision à base de cas.
- Texte intégral [pdf]

**Jeudi 29 mars 2007 12:45-13:45****CHASSANG S.**(MIT) :__Fear of miscoordination and the robustness of cooperation in dynamic global games with exit__- RésuméThis paper develops a framework to assess the impact of miscoordination fear on agents' ability to sustain dynamic cooperation. Building on theoretical insights from Carlsson and van Damme (1993), it explores the effect of small amounts of private information on a class of dynamic cooperation games with exit. It is shown that lack of common knowledge creates a fear of miscoordination which pushes players away from the full-information Pareto frontier. Unlike in one-shot two-by-two games, the global games information structure does not yield equilibrium uniqueness, however, by making it harder to coordinate, it does reduce the range of equilibria and gives bite to the notion of local dominance solvability. Finally, the paper provides a simple crite- rion for the robustness of cooperation to miscoordination fear, and shows it can yield predictions that are qualitatively different from those obtained by focusing on Pareto effcient equilibria under full information. keywords: cooperation, strategic risk, miscoordination risk, global games, dy- namic games, exit games, rationalizability, local strong rationalizability, local domi- nance solvability. Jel classification codes: C72, C73
- Texte intégral [pdf]

**Jeudi 8 mars 2007 12:45-13:45****VERDIER M.**(ENST) :__Interchange fees and incentives to invest in the quality of a payment card system__- RésuméIn this paper, I analyse the optimal interchange fee in a payment card system where two monopoly banks have the possibility to invest in quality to raise the transaction volume. I model quality as a sum of specific investments from the cardholder's bank (the Issuer) and the merchant's bank (the Acquirer), and I assume that investments impact di¤erently the consumer and the merchant side. If the level of quality is exogenous, I prove that it is optimal for the payment platform to set an interchange fee equal to the Acquirer's margin. I extend Baxter (1983)'s model by proving that the optimal interchange fee depends on the level of quality of the payment system. However, if banks have the possibility to invest in quality, and if the perception of quality improvements is higher on the consumer side, the optimal interchange fee can be lower than the Acquirer's margin. This is because the interchange fee and the Acquirer's investments are strategic substitutes. If investments impact relatively more the merchant side, the optimal interchange fee remains equal to the Acquirer's margin. JEL Codes: G21, L31, L42. Keywords: Payment card systems, interchange fee, two-sided markets, investments in quality.
- Texte intégral [pdf]

**Jeudi 15 février 2007 12:45-13:45****LAMBERT-MOGILIANSKI A.**(PSE) :__Système non-classique et rationalité limitée__**Co-auteur (s) : V. I. Danilov**- RésuméIndividual choices often depend on the order in which decisions are made, i.e., they exhibit non-commutativity. In this paper, we expose a general theory of measurable systems (an example of which is an individual’s preferences) allowing for incompatible (non-commuting) measurements. The basic concepts are illustrated in an example of non-classical rational choice. We conclude with a discussion of some of the basic properties of non-classical systems in the context of social sciences. In particular, we argue that the distinctive feature of non-classical systems translates into a formulation of bounded rationality. JEL: D80, C65, B41 keywords: non-classical system, incompatible measurement, orthospace, state, properties.
- Texte intégral [pdf]

**Jeudi 8 février 2007 12:45-13:45****LEGER P. T.**(HEC) :__Provider Competition in a Dynamic Setting__**Co-auteur (s) : M. Allard & L. Rochaix**- RésuméIn this paper, we examine provider and patient behaviour where effort is non-contractible and competition between providers is modeled in an explicit way. More specifically, we construct a model where physicians repeatedly compete for patients and where patients' outside options are solved for in equilibrium. In our model, physicians are characterized by an individual-specific ethical constraint which allows for unobserved heterogeneity in the physicians market. Allowing for unobserved heterogeneity in the physicians market introduces uncertainty in the patient's expected treatment if he were in fact to leave his current physician to seek care elsewhere. We also introduce switching costs associated with moving from one physician to another and, un- certainty in the treatment-outcome relationships. Our model can generate equilibria which are consistent with real-world observations. That is, our model can generate treatment heterogene- ity, unstable physician-patient relationships and, over-treatment (a form of defensive medicine). Our model also suggests several avenues which may lead to more e¢ cient provision of care. JEL classification: I10, I18, J24, C30 Keywords: Physician Payment Mechanisms, Physician heterogeneity, Competition, Information Asymmetry, Insurance.
- Texte intégral [pdf]

**Jeudi 25 janvier 2007 12:45-13:45****CHASSAGNON A.**(PSE) :__On moral hazard and non exclusive contracts__

**Jeudi 11 janvier 2007 12:45-13:45****MENONI B.**(Centre de recherche en économie et statistique) :__Decision making with doubt regarding the consequences of an action__- RésuméAnscombe & Aumann (1963) improved the model developed by von Neumann & Morgenstern (1944) suggesting that the outcome of an act can be a lottery. They showed that if the preference relation of a decision maker obeys several axioms then the latter behaves as if they were maximizing some expected utility. We slightly depart from their definition of acts and consider that, in a given state of Nature, the result of an act is a possibility distribution over outcomes rather than a probability distribution. We then extend the work by ? to a more general setting. One can consider our contribution as a refinement of a model developed by Ghirardato (2001): the latter models the consequence of an act as a list of possible outcomes. We add to the list of possible outcomes some structure, namely a qualitative structure. We show that if the preference relation a decision maker may have obeys several restrictions, their choices ensue from the maximization of a – qualitative – expected utility. Keywords: Decision making, ordinal representation, possibility theory. JEL Classification: D81, D83, D89.
- Texte intégral [pdf]

**Jeudi 21 décembre 2006 11:00-12:00****WEYL G.**(Princeton univ.) :__The Price Theory of Two-Sided Markets__- RésuméI establish a number of baseline positive and normative results in the price theory of two-sided markets building on the work of Rochet and Tirole (2003). On the positive side, I introduce the notion of vulnerability of demand to separate previously confounded effects. I find that competition, price controls and subsidies always reduce the price level, defined as the sum of prices on the two sides of the market. However, price controls and competition that are “unbalanced” may raise prices on one side of the market. The normative analysis emphasizes the importance of externalities across the two sides of the market and their impact on socially optimal pricing. The socially optimal price level, which takes an intuitive Ramsey-pricing form, is always below cost. Subsidies may be desirable even if the profits of the firm are disregarded. In determining optimal price balance, seemingly similar welfare criteria generally conflict. Consumers on one side of the market may want to make transfers to the other side in order to thicken their pool of partners. Unbalanced competition that undermines such transfers may harm all parties. A number of implications for policy are discussed.
- Texte intégral [pdf]

**Jeudi 7 décembre 2006 12:45-13:45****LAMY L.**(Laboratoire d'économie industrielle) :__Order Independent Individual Rationality__- RésuméWe consider the implementation of an economic outcome under complete information when the principal cannot commit to a simultaneous participation game. From a class of sequential participation games, we introduce the concept of implementability under order independent individual rationality. We characterize the set of implementable mechanisms, which is possibly a non-convex set, and we solve the optimal design program: the principal raises a lower revenue but economic efficiency is not damaged.
- Texte intégral [pdf]

**Jeudi 23 novembre 2006 12:45-13:45****NIKOLOVA R.**(Laboratoire d'économie industrielle) :__Incentives, markets and knowledge based hierarchies__

**Jeudi 16 novembre 2006 12:45-13:45****DESGRANGES G.**(THEMA) :__Strongly Rational Expectations Equilibria, Endogenous Acquisition of Information and the Grossman-Stiglitz Paradox__**Co-auteur (s) : M. Heinemann**- RésuméStrongly Rational Expectations Equilibria, Endogenous Acquisition of Information and the Grossman-Stiglitz Paradox Abstract: This paper analyzes conditions for existence of a strongly rational expectations equilibrium (SREE) in models with private information where the amount of private information is endogenously determined and where the price transmits relevant information to market participants. It is shown that the conditions for existence of a SREE known from models with exogenously given private information have to be qualified if private information is endogenously determined. A SREE exists only if informativeness of the market price falls short of a specific lower bound, which depends on the properties of the cost function associated with the individual acquisition of information. This upper bound is generally lower than that valid in case of exogenously given information. An interpretation that links our results to the famous Grossman-Stiglitz Paradox is also given.
- Texte intégral [pdf]

**Jeudi 19 octobre 2006 12:45-13:45****LAURENT R.A.**(PSE) :__Differentiated Duopoly with Elimination By Aspects__- RésuméElimination by aspects (EBA) is a discrete model of probabilistic choice worked out by Tversky in 1972 which supposes that decision makers follow a particular heuristic during a process of sequential choice. Goods are described by their attributes and, at each decision stage, consumers eliminate all the products not having an expected specific attribute, until only one option remains. In this paper, probabilities resulting from the EBA model are used to construct demands of a differentiated duopoly with imperfectly rational consumers. These demands are also consistent with partial heterogeneity of preferences and may be linked with a spatial framework in which consumers have convex perception of distance. In this model, a price Nash equilibrium in pure strategies exists under two conditions on attributes level and unit costs. At the outcome, the “differentiation by attributes” constitutes a general framework which embodies both horizontal and vertical differentiation. When the equilibrium does not exist, the interaction of best response functions of the firms induces an Edgeworth cycle instead of an exit of the lowest attributes level firm. This result underlines the role of cost asymmetries in the existence of such a cycle.
- Texte intégral [pdf]

**Jeudi 12 octobre 2006 12:45-13:45****SIMULA L.**(PSE) :__Optimal Nonlinear Income Tax when Highly Skilled Vote with their Feet__**Co-auteur (s) : A. Trannoy**- RésuméThis paper examines how allowing individuals to emigrate to pay lower taxes changes the optimal non-linear income tax scheme in a Mirrleesian economy. Type-dependent participation constraints are borrowed from contract theory. An individual emigrates if his domestic utility is less than his utility abroad net of migration costs, utilities and costs both depending on productivity. Three social criteria are distinguished according to the agents whose welfare matters. Mobility significantly alters the closed-economy results qualitatively, but also quantitatively as verified by simulations. A curse of the middle-skilled occurs in the first-best. In the second-best, the middle-skilled can support the highest average tax rates and the marginal tax rates can be negative. Moreover, preventing emigration of the highly-skilled is not necessarily optimal.
- Texte intégral [pdf]

**Jeudi 21 septembre 2006 12:45-13:45****KETS W.**(Tilburg univ.) :__Learning to be Prepared__**Co-auteur : Mark Voorneved**- RésuméBehavioral economics provides several motivations for the common observation that agents appear somewhat unwilling to deviate from recent choices. More recent choices can be more salient than other choices, or more readily available in the agent’s mind. Alternatively, agents may have formed habits, use rules of thumb, or lock in on certain modes of behavior as a result of learning by doing. This paper provides discrete-time adjustment processes for strategic games in which players display precisely such a bias towards recent choices. In addition, players choose best replies to beliefs supported by observed play in the recent past, in line with much of the literature on learning. These processes eventually settle down in the minimal prep sets of Voorneveld (2004, 2005).
- Texte intégral [pdf]

**0000**- Salle R1-14 - Campus Jourdan 75014 PARIS
**HELIOS HERRERA prenom ...**(Warwick) :__Echo chamber elections__**Ravideep Sethi**- RésuméElections aggregate information of citizens who gather information, in part from mainstream news, but possibly to a large extent from their different echo chambers which filter news in particular ways. We study information aggregation and electoral outcomes in the presence of endogenous echo chambers. Citizens choose their echo chambers in a self-serving way, in part to shield them from the outside world (mainstream news) and thus preserve their political faith as often as possible. The key asymmetry is how isolated citizens are, namely how much they trust/distrust mainstream news to begin with. We study in several contexts the electoral advantage that the side with more isolated followers may have. This advantage can turn into a sure electoral victory regardless of the true state of the world, thereby failing information aggregation, in some instances. Information aggregation still fails if voters can abstain, don’t fully trust their echo chamber sources either, or gain utility from consuming favourable news.

**0000**- Salle R1-14 - Campus Jourdan 75014 PARIS
**HELIOS HERRERA prenom ...**(Warwick) :__Echo chamber elections__**Ravideep SETHI**- RésuméElections aggregate information of citizens who gather information, in part from mainstream news, but possibly to a large extent from their different echo chambers which filter news in particular ways. We study information aggregation and electoral outcomes in the presence of endogenous echo chambers. Citizens choose their echo chambers in a self-serving way, in part to shield them from the outside world (mainstream news) and thus preserve their political faith as often as possible. The key asymmetry is how isolated citizens are, namely how much they trust/distrust mainstream news to begin with. We study in several contexts the electoral advantage that the side with more isolated followers may have. This advantage can turn into a sure electoral victory regardless of the true state of the world, thereby failing information aggregation, in some instances. Information aggregation still fails if voters can abstain, don’t fully trust their echo chamber sources either, or gain utility from consuming favourable news.

**0000****LEVY John**:__*__

**0000**- campus Jourdan, 48 boulevard Jourdan, 75014 Paris
**BASTIANELLO Lorenzo**(Paris 1) :__Target-based solutions for Nash bargaining__- RésuméAbstract. We revisit the Nash model for two-person bargaining. A mediator knows agents' ordinal preferences over feasible proposals, but has incomplete information about their acceptance thresholds. We provide a behavioural characterisation under which the mediator recommends a proposal that maximises the probability that bargainers strike an agreement. Some major solutions are recovered as special cases; in particular, we o ffer a straightforward interpretation for the product operator underlying the Nash solution.

**0000**- Campus Jourdan, bâtiment G, 1er étage, salle de réunions
**REDLICKI Bartosz**(U of Cambridge) :__*__

**0000****RIVERA Thomas**(HEC) :__Incentives and the Structure of Communication__- RésuméThis paper analyzes the incentives that arise within an organization when communication is restricted to a particular network structure (e.g., a hierarchy). We show that restricting communication between the principal and agents may create incentives for the agents to misbehave when transmitting information and tasks throughout the organization. Such incentives can render the principal's most preferred outcome infeasible and therefore introduces a trade off between the cost of communication borne by the principal and the benefit of curbing incentives to deviate induced by the communication structure. To remedy this issue, we provide necessary and sufficient conditions on the topology of the network of communication such that restricting communication to a particular network does not restrict the set of outcomes that the principal could otherwise achieve. In this sense, we show that for any underlying incentives and any outcome available when communication is unrestricted, there exists a (finite) communication scheme restricted to a particular network that implements this outcome (i.e., does not induce agents to misbehave in the communication phase) if and only if that network satisfies our conditions.