ABSTRACT. In second-price auctions with interdependent values, bidders do not necessarily have dominant strategies. Moreover, such auctions may have many equilibria. In order to rule out the less intuitive equilibria, we define the notion of distributional perfect equilibrium for Bayesian games with infinite type and action spaces. We prove that every Bayesian game has a distributional perfect equilibrium provided that the information structure of the game is absolutely continuous and the payoffs are continuous in actions for every type. We apply distributional perfection to a class of symmetric second-price auctions with interdependent values and demonstrate that a specific type of equilibrium is perfect, while many of less intuitive equilibria are not.

ABSTRACT. In explaining the winner’s curse, recent approaches have focused on one of two cognitive processes: conditional reasoning and belief formation. We provide the first joint experimental analysis of the two obstacles. First, we observe significant differences in behavior between a simple common-value auction and a transformed version of this auction that differs in not requiring conditional reasoning. Second, assistance in belief formation leads to significant behavioral changes in both games. The two effects are of similar magnitude and amplify each other when present jointly. We conclude that the combination and the interaction of the two cognitive processes in auctions lead to relatively low strategic sophistication compared to other domains. The study’s focus on games’ objective cognitive challenges is potentially useful for predictions across games and complements the common focus on behavioral models and their explanatory power.

ABSTRACT. We explore existence and properties of Bayesian Nash equilibrium in simultaneous first-price auctions for complementary goods. We introduce a novel partial order on types and apply methodology in Athey (2001), McAdams (2003) and Reny (2011) to demonstrate existence of monotone equilibria when the bid space is finite. We then apply this finding to demonstrate two new results on monotone equilibria in continuous bid spaces. The first one establishes existence of monotone equilibria in a setting similar to Krishna and Rosenthal (1996), where a single global bidder bids for multiple objects against a set of local bidders bidding for single objects only. The second considers equilibria with communication as in Jackson, Simon, Swinkels and Zame (2002), showing that in our setting their statements on distributional strategies can be refined to be monotone. Finally, we discuss conditions under which these results can be extended to all-pay, second-price and other standard simultaneous auctions.

ABSTRACT. We analyze the effect of industrial espionage on limit-pricing models. We consider an incumbent monopolist engaged in R&D trying to reduce his cost of production and deter a potential entrant from entering the market. The R&D project may be successful or not and its outcome is a private information of the incumbent. The entrant has an access to an Intelligence System (IS) of a certain precision that generates a noisy signal on the outcome of the R&D project, and she decides whether to enter the market based on two signals: the price charged by the incumbent and the signal sent by the IS. Assuming that the precision of the IS is exogenous and common knowledge, for intermediate values of the IS precision, the set of pooling equilibria is non-empty even with profitable entry and the entrant enters if the IS tells her the R&D project was not successful.

ABSTRACT. We consider an information acquisition game between duopolists, who can learn about unobservable demand from two sources: market research and industrial espionage. The latter provides a noisy signal of the competitor's private information. In equilibrium firms pick an action which is a weighted average of their signals. The weight on espionage increases in the complementarity of the players' actions, but is lower than the weight placed on the market research. If the coordination motive is not strong enough, or the market research technology is sufficiently precise, the players will not engage in espionage. Spying under price competition is beneficial for the industry, and the economy as a whole: aggregate profit and social welfare increase as the precision of the spying device improves. The same is true under quantity competition if and only if the espionage technology is already more accurate than the alternative market research technology.

ABSTRACT. This talk has been cancelled.

ABSTRACT. We consider optimal nonlinear income taxation in a Stiglitz two type framework with a single period where the government cannot commit to announced policies. Individuals announce their types to the government which then sets taxes. Two crucial factors are the degree to which the government desires to carry out redistribution and how much it discounts the welfare of individuals it believes may have misrevealed their type. These traits are fixed attributes of the individual in charge of the government who is chosen in a prior stage election. If the redistributive intentions of the government are too strong, its ability to redistribute may be weak. In the initial voting stage, this may create an extremist’s dilemma between purity and pragmatism where a voter who has a strong preference for redistribution may have to vote for a candidate whose preferences for redistribution are much less in order for any redistribution to occur.

ABSTRACT. We study dynamics of risky innovations when the firms need constantly update the product to be able to sell new versions to the consumers. The firms face uncertainty about the consumers' valuation for the product, and the consumers' willingness to buy depends on the firms' product development efforts. Product development is costly and the expected net benefit is only positive if the consumers like the update. The firms have the opportunity to wait until the others introduce new features in the product and to copy them only if the consumers like them. The possibility to wait and observe the other firms' success creates incentives to free-ride on their development efforts. We find that when the firms' are sufficiently pessimistic about the consumers' valuation for the product, free-riding leads to an inefficiently low development effort. The innovation is retarded, leading to an inefficiently short lifetime when compared to the efficient benchmark.

ABSTRACT. We propose a theoretical model to explain two salient features of the U.S. stock exchange industry: (i) the proliferation of stock exchanges offering identical transaction services; and (ii) sizable dispersion and frequent changes in stock exchange fees, highlighting the role of discrete pricing.

ABSTRACT. Rational agents might choose to invest in a certain kind of capital in a period in the hope of making higher returns from their investments made in consecutive periods. We examine the impact of such an interaction on the incidence of coordination failure and accordingly social welfare. In our set-up, investment complementarities are present both within periods (call vertical complementarity) and between periods (call horizontal complementarity). In particular, other than the underlying economic fundamental, the return on investment depends on its aggregate level in that period as well as the aggregate investment made in the previous period. The results suggest that full transparency is optimal at the social level as long as agents have an access to relatively more precise private information and complementarities are sufficiently low. More transparency otherwise reduces social welfare as the gain from better vertical coordination is outweighed by the loss resulted from lesser horizontal coordination.

ABSTRACT. Nowadays, corporate websites have become a major communication channel between the management team of corporations and the existing or potential investors. However, the corporations’ internet information disclosure can be a double-edged sword that some may lead drastic fluctuations in stock prices and have an impact on information users and stakeholders. In fact, the corporate internet information disclosure is a game in the capital market. This paper tries to research a way to analyze the advantages and disadvantages of internet information disclosure. This research will contribute to improve the information disclosure level in internet environment and strengthen the management of corporate information disclosure.

ABSTRACT. We consider Baron-Ferejon type majoritarian bargaining model with recognition probabilities determined by Tullock contest. While there is no symmetric perfect equilibrium with pure investment strategies, we show that there are asymmetric perfect equilibria with pure investment strategies (when discount factor is close to 1). One difficulty is caused by the possibility that there are players who receive less payoffs than the payoff of the critical player whose recognition probability is the q-th largest with q being the quota for the majority. To avoid this difficulty, we make use of the limit game as the discount factor tends to 1, even though the convergence of the game is not uniform, since one can verify that for these equilibria approximate the equilibria of discounted games. In this asymmetric equilibrium, there is always a player who has higher probability and not always invited to be a member of the majority coalition.

ABSTRACT. As is known, while introducing games in extensive form in their basic book of 1944, von Neumann and Morgenstern also supplied a method for transforming such games into normal form. Once more in that book, the same authors provided a method for transforming games from characteristic function form into normal form, although limited to constant-sum games. In 2007 Gambarelli proposed a generalization of this method to variable-sum games. Gambarelli introduced a solution for the game in characteristic function form, made up of the set of Pareto-Optimal payoffs generated by Nash Equilibria of the transformed game. In this paper, the above transformation method is generalized to the case in which each player’s requests vary according to the coalition being addressed. Theorems regarding the existence of a solution are proved. Software for the automatic generation of the solution is supplied.

ABSTRACT. To analyze players' strategic alliance behavior, we introduce a new noncooperative coalitional bargaining model, in which each player can buy out other players with upfront transfers. We find that delay in bargaining generically occurs and we uncover the role of an essential player in a transferable utility game, or a veto player in a simple game, in preventing efficient outcomes.
In an application to legislative bargaining with vote buying, if a veto player and a non-veto player coexist, then a non-winning coalition forms as an intermediate bargaining step and the final winning coalition is not necessarily minimal. We fully characterize the set of the equilibrium outcomes in three-player simple games and compare it with well-known cooperative power indices.

ABSTRACT. We develop a model of college assignment as a large contest wherein students with heterogeneous abilities compete for seats at vertically differentiated colleges through the acquisition of productive human capital. We use a continuum model to approximate the outcomes of a game with large, but finite, sets of colleges and students. By incorporating two common families of affirmative action mechanisms into our model, admissions preferences and quotas, we can show that (legal) admissions preference schemes and (illegal) quotas are outcome and incentive equivalent. We also assess the welfare costs of using human capital accumulation to compete for college admissions. While competition is necessary to generate a (welfare enhancing) assortative match, the welfare losses from the accumulation of human capital solely to compete for a better college seat may be larger on average than the gains from assortativity.

ABSTRACT. The Kelso and Crawford (1982) model of job matching has been a basis for the analysis of two-sided many-to-one markets. They proved the existence of a stable matching in both the market with discrete transfers and the market with continuous transfers. The cumulative offer process can be applied to the former market, but not to the latter market. Thus, the algorithm provides no explicit shape of a competitive price vector. The purpose of this paper is to introduce a new algorithm that can be directly applied to the market with continuous transfers. We show that our new algorithm, the cumulative offer process with continuous transfers, finds the maximal competitive price vector in finite steps. Our result offers a clear insight into how the maximal competitive price vector is determined.

ABSTRACT. This talk has been cancelled.

ABSTRACT. Many cooperative games, especially ones stemming from resource pooling in queueing or inventory systems, are based on situations in which each player is associated with a single attribute (a real number representing, say, a demand) and in which the cost to optimally serve any sum of attributes is described by an elastic function (which means that the per-demand cost is non-increasing in the total demand served). For this class of situations, we introduce and analyze several cost allocation rules: the proportional rule, the serial cost sharing rule, the benefit-proportional rule, and various Shapley-esque rules. We study their appeal with regard to fairness criteria such as coalitional rationality, benefit ordering, and relaxations thereof. After showing the impossibility of combining coalitional rationality and benefit ordering, we show for each of the cost allocation rules which fairness criteria it satisfies.

ABSTRACT. Each agent holds a discrete resource (object). A governing authority wishes to reshuffle (possibility in a probabilistic fashion) these objects amongst the agents taking into account which agent is currently holding which object. We propose the x-endowment lower bound : a parameterized family of properties which weakens the endowment lower bound (also known as individual rationality ). We consider probabilistic rules which are decomposable : convex combinations of efficient and group strategy-proof rules. Theorem 1 shows when efficiency is preserved when we mix only Serial Priority rules. As a corollary to this, we identify each value of x for which we can construct rules that satisfy efficiency, the x-endowment lower bound, and can be decomposed into Serial Priority rules. Theorem 2 states that for most values of x, efficiency, the x-endowment lower bound, and decomposability characterize Gale’s Top Trading Cycles.

ABSTRACT. This talk has been cancelled.

ABSTRACT. Many cartels, especially in the industry where monitoring is imperfect and private, involve third-party facilitators, who organize meetings, distribute agreed market shares, collect and verify data, detect deviations, and moderate tensions. Motivated by this observation, we study repeated games with general private monitoring and mediator, and characterize the limit set of sequential communication equilibrium as the discount factor goes to one. As a corollary, we prove the folk theorem with a weak identifiability condition.

ABSTRACT. Equilibrium payoffs corresponding to subgame perfect equilibria in pure strategies are characterized for infinitely repeated games with discounted payoffs. The equilibrium payoff set of a game is a fixed-point of set-valued operators introduced in the paper. The new operator formalism is utilized in showing the folk theorem for repeated games with unequal but constant discount rates. When the players become more patient, the equilibrium payoff set converges to a particular fixed-point of an asymptotic operator. The limit sets for constant discount rates can be used in analyzing the outer limit of equilibrium payoffs when the discount factors increase but discount rates are not fixed.

ABSTRACT. The set of equilibrium payoffs of an infinitely repeated game is the greatest fixed point of the generation map, B, as shown by Abreu, Pearce and Stacchetti (1990). Generally regarding a fixed point, there are several familiar questions of stability: Is the fixed point attractive? Is it essential, in the sense that nearby maps have nearby fixed points? This paper takes up these questions regarding stability of the equilibrium payoff set in repeated games. The payoff set is attractive from above and varies upper semicontinuously in the parameters of the game. Under some conditions it is attractive from below and lower semicontinuous, and under other conditions not. Under some conditions, appending a small continuation payoff set to the finitely repeated game yields approximately the same initial payoff set as that of the infinitely repeated game, because the latter's payoff set is widely attractive.

ABSTRACT. We study a class of evolutionary game dynamics under which the population state moves in the direction that agrees most closely with current payoffs. This agreement is defined by means of a Riemannian metric which imposes a geometric structure on the set of population states. By supplying microfoundations for our dynamics, we show that this metric provides a state-dependent but payoff-independent specification of the saliences and similarities of available strategies. Similarly to the replicator and the projection dynamics (the archetypal examples of this class), all Riemannian dynamics satisfy certain basic desiderata (such as global convergence in potential games). Moreover, when the underlying metric satisfies a Hessian integrability condition, the dynamics preserve many further properties of the replicator and projection dynamics. We also examine the close connections between these Hessian dynamics and reinforcement learning in games, extending and elucidating a well-known link between the replicator dynamics and exponential reinforcement learning.

ABSTRACT. This paper studies imitative dynamics for population games with continuous strategy space. We define imitative dynamics---which include the replicator dynamic as a special case---as evolutionary dynamics that satisfy the imitative property and payoff monotonicity. Our definition of payoff monotonicity is different from the one defined in Oechssler and Riedel (2002). We find that our definition is better at capturing the notion of payoff monotonicity for the finite strategy case (cf. Weibull (1995)), and Oechssler and Riedel (2002)'s definition is closer to aggregate monotonicity in the sense of Samuelson and Zhang (1992). We provide sufficient conditions for imitative dynamics and general evolutionary dynamics to be well-defined. Finally, we extend general properties of imitative dynamics as well as global convergence and local stability results in potential games from finite strategy settings to continuous strategy settings.

ABSTRACT. This paper provides sufficient conditions under which aggregate monotonic selection dynamics selects against strategies that do not survive iterated elimination of weakly dominated strategies. The conditions are satisfied by e.g. Bertrand duopolies, first price auctions, the finitely repeated Prisoners' Dilemma, and the p-beauty contest. The condition implies evolutionary selection against iteratively strictly dominates strategies.

ABSTRACT. We consider the decision of placing two public bads in a region, modelled by a line segment. Residents of the region have single-dipped preferences over different locations in the region. Our set of alternatives are collection of pairs of locations. Each agent can have any strict preference over the set of alternatives which can be written as a separable extension of some single-dipped preference. A social choice function or rule takes a profile of reported preferences over the set of alternatives as input and selects a pair of locations. All rules satisfying strategy-proofness and Pareto optimality are characterized. These rules pick only boundary locations.

ABSTRACT. It is shown that in a model where agents have single-peaked preferences on the sphere, every Pareto optimal social choice function that is strict or coalitional strategy-proof, is dictatorial.

ABSTRACT. The paper provides consistent mathematical framework for the seminal Tiebout free-mobility model (1956). Our setting supports continuum of consumers with multidimensional preferences and finite number of strategic public good providers. We accommodate the most general assumptions: providers' production function may have variable returns to scale, our framework is rich enough to incorporate possible externalities, spillovers, scale economies, network effects, etc.; consumers utility may depend on choice of other agents in almost arbitrary way. We focus on equilibrium existence, however the questions of efficiency and stability are not left behind.

ABSTRACT. Before group members individually decide their efforts in a contest to set a policy, groups are allowed to make some concessions to their opponent by choosing a less controversial policy to lobby for. When valuations over the set of policies follow a linear function, we show that concessions are never profitable when the contest success function is homogeneous of degree zero but they are when it is of difference form. Surprisingly, concessions might be detrimental for the members of the group that does not make them. Comparing this situation with another where efforts are decided collectively at a group level allows us to remark the effect of positive externalities of effort as the key cause of this damage.

ABSTRACT. In a framework of contest design, we study capacity-constrained competition and analyze its resulting revenue shares. Our analysis contrasts the near-symmetric case, where firms have similar supply sizes, and the extremely asymmetric case, where one large firm dominates the market. In particular, we show that while in the near-symmetric case simple contests provide near optimal equilibrium revenues for all sellers, in the asymmetric case a large firm can design more complicated contests that result in disproportionally low equilibrium revenues for her smaller opponents.

ABSTRACT. Consider a symmetric common-value Tullock contest with incomplete information in which the players' cost of effort is the product of a random variable and a deterministic real function of effort, d. We show that the Arrow-Pratt curvature of d; Rd; determines the effect on equilibrium efforts and payoffs of the increased flexibility/reduced commitment that more information introduces into the contest: If Rd is increasing, then effort decreases (increases) with the level of information when the cost of effort (value) is independent of the state of nature. Moreover, if Rd is increasing (decreasing), then the value of public information is non-negative (non-positive).

ABSTRACT. We study the effect of verification control in a communication game with partially verifiable information. In this game, an informed party (the sender) sends a two-dimensional message to a decision-maker (the receiver), but only one dimension of the sender's claim can be verified. We use laboratory experiments to compare strategic behaviour in situations where the receiver chooses which aspect to verify and cases where the sender has this verification control. We find that both parties benefit when the receiver is the one who decides which dimension to verify. This is not due to more informative messages but rather to lower skepticism on the receiver's part when the verified dimension is in accord with the message sent. The findings of this study provide guidance for potential employers, judges or investors as to which decision-making environment promotes fewer erroneous choices.

ABSTRACT. We investigate the reasons why a political leader (manager) may be reluctant to abandon her chosen project, when she is concerned both about social welfare and the electorate's (shareholders') beliefs about her ability. The quality of the project is correlated with her ability, and there is ex ante symmetric information. However, news regarding the project arrives gradually over time and is privately observed by the leader, who may choose whether and when to cancel the project, so that we have a dynamic signalling model. We find that the inefficiency caused by career concerns varies with the information structure.

ABSTRACT. This paper studies how rumours spread by word of mouth. I develop a model of diffusion of unverifiable information in a population of two types of biased agents. Each agent cares about his own action and the actions of others. One agent observes a signal of the state of the world and communicates it by cheap talk to an agent whom he meets, thus starting a chain of communication. I show that truthful diffusion of information is possible under certain conditions on, among others, the preferences of agents, the distribution of types in the population, and the assortativity of meetings. Depending on the characteristics of the population, a widely spread rumour follows one of two types of diffusion: (1) with a large number of informed agents and a small proportion of uninformative messages, or (2) with a small number of informed agents and a large proportion of uninformative messages.

ABSTRACT. We consider a network game with production and externalities, which describes a situation typical for many economic, social, and political systems. In the first period of time each of the agents in a network receives endowment and distributes it between consumption and investment. In the second period the agent’s consumption depends on her own investment as well as on investments of her neighbors in the network. The agent’s payoff is determined by her consumption in the two periods. We introduce adjustment dynamics into the model and study the problem of stability of the game equilibrium. An important fact which we find is a special role of conditions, which are referred here as the presence and the absence of productivity, both in the static and in the dynamic frameworks. We find instability of the inner equilibrium and convergence to a new corner equilibrium and, correspondingly, stability of the latter.

ABSTRACT. We study the effect of potential entry on the formation and stability of R&D networks considering farsighted firms. In particular, we seek to understand whether the incentives for forming R&D collaborations may be altered by the presence of a potential entrant, as well as which are the factors that affect these incentives. We find that an incumbent firm might be willing to sustain an otherwise undesirable collaboration in order to prevent the entry of a new competitor, as well as to refrain from establishing an otherwise desirable collaboration, expecting to form a more profitable collaboration with the entrant. We also find that entry may lead an inefficient incumbent firm to exit the market.

ABSTRACT. We analyze the stability of networks when two firms strategically form costly links to customers. We interpret these links as customer relationships that enable trade to sell a product. Equilibrium prices and equilibrium quantities are determined endogenously for a given network of customer relationships. We investigate in how far the degree of substitutability of the firms' products and the costs of link formation influence the equilibrium profits and thus have an impact on the incentives to strategically form relationships to customers. For networks with n customers we analyze local stability regions for selected networks and determine their limits when n goes to infinity. We establish the existence of locally stable networks for complementary or independent products. We also relate local and Nash stability for selected networks with n customers. For networks with three customers we characterize locally stable networks. In particular, existence is guaranteed for any degree of substitutability.

ABSTRACT. Decision makers often face uncertainty about the ability and the objectives of their advisors. If an advisor is sufficiently concerned about establishing a reputation for being skilled and unbiased, the advisor truthfully reports his private information about the decision-relevant state. However, truthful revelation may not necessarily maximize the expected payoff of the decision maker. There is a trade-off between the amount of information revealed about the decision-relevant state and what the decision maker learns about the advisor's type. While in a truth-telling equilibrium the decision maker learns only about the ability of the advisor, in an equilibrium with some misreporting the decision maker learns about both the ability and the objectives of the expert. In other words, while reputational concerns improve discipline, they nevertheless reduce sorting. Therefore, if a decision maker places enough weight on future choices relative to present ones, equilibria with some lying may be preferred to truth-telling.

ABSTRACT. Economists focus on Dictator Game giving in monetary domain rather than in non-monetary domain although investigation of giving in non-monetary domain has also economic value. This study aims to make a systematic comparison of generosity and rationalizability of individual preferences in a modified dictator game across monetary and non-monetary domains with positive and negative framing using laboratory experiments. The rationalizability of preferences is measured using the severity of Generalized Axiom of Revealed Preferences (GARP) violations as in previous studies (e.g. Andreoni and Miller, 2002; Davis et al. 2012). Results suggest that generosity and GARP violations (in amount and severity) are larger in non-monetary domain than in monetary domain and that the high level of generosity in non-monetary treatments cannot be fully explained by the high level of GARP violations.

ABSTRACT. Sequential rationality requires that players maximize their conditional expected payoff at all information sets, including those that are not reached during any given play of the game, and those the player does not expect to be reached. However, planned choices and conditional beliefs at such information sets cannot be verified or elicited in an incentive-compatible way. Thus, there is a tension between sequential rationality and the revealed-preference methodology. This paper addressed this tension by proposing a new choice criterion, "sequential preference." A strategy that is maximal for sequential preference is sequentially rational in the traditional sense. However, if players conform to the sequential-preference model, their planned choices and conditional beliefs can be elicited using a variant of the strategy method (Selten, 1967). These results thus demonstrate that dynamic games can be analyzed consistently with the tenets of revealed preference.

ABSTRACT. The behavior of strongly dynamically consistent agents who follow through with any ex ante optimal plan, whether it involves mixed strategies or not, cannot be distinguished from the behavior of Bayesian agents. This observational equivalence does not hold for weakly dynamically consistent agents who only follow through on ex ante optimal pure strategies. Applying these results to games and mechanism design with ambiguity averse agents, I show that the predictions of such models may differ substantially from the predictions of models with Bayesian agents if we only assume weakly dynamically consistent behavior. If we assume strongly dynamically consistent behavior the equilibria of games and the social choice functions implementable by mechanisms with ambiguity averse agents do not differ from their counterparts with Bayesian agents.

ABSTRACT. We study incomplete information games involving players who perceive ambiguity about the types of others and may be ambiguity averse as modeled through smooth ambiguity preferences (Klibanoff, Marinacci and Mukerji, 2005). Our focus is on multi-stage games with observed actions and on equilibrium concepts satisfying sequential optimality -- each player's strategy must be optimal at each stage given the strategies of the other players and the player's conditional beliefs about types. We show that for the purpose of identifying strategy profiles that are part of a sequential optimum, it is without loss of generality to restrict attention to beliefs generated using a particular generalization of Bayesian updating. We propose and analyze two strengthenings of sequential optimality. Examples illustrate new strategic behavior that can arise under ambiguity aversion. Our concepts and framework are also suitable for examining the strategic use of ambiguity.

ABSTRACT. This paper revisits cooperative games with externalities, i.e. games where the payoff of each coalition depends on the partition of the outside players. We construct the characteristic function of a coalition assuming its members have probabilistic expectations over the coalitional behavior of the outsiders, i.e., they assign various probability distributions on the set of partitions they can form. We apply this framework to the class of aggregative normal form games and we derive conditions on the payoff functions and the probability distributions that guarantee the non-emptiness of the core of the corresponding cooperative games.

ABSTRACT. Shapley (1955) introduced the model of abstract market game. Players are partitioned into two groups such that any two players in the same group are substitutes and those from different groups are complements. The model is a generalization of several familiar models of games including the assignment game. Shapley conjectured that all abstract market games possess cores. We show that the conjecture is true if the number of players is no bigger than 4 or one side is monopoly, but in general the structure of abstract market games is not strong enough to guarantee the nonemptiness of the core. In addition, the stable set may touch both the seller facet and the buyer facet of the imputation set at infinitely many rather than single points. We establish necessary and sufficient additional conditions for symmetric abstract market games to possess cores.

ABSTRACT. We present a new model for cost-sharing in minimum cost spanning tree problems to allow planners to identify how many agents merge. Under this new framework, in contrast to the traditional model, there are rules that satisfy the property of merge-proofness. Furthermore, strengthening this property and adding some others, such as population monotonicity and solidarity, makes it possible to define a unique rule that coincides with the weighted Shapley value of an associated cost game.

ABSTRACT. In U.S. presidential elections, voters in noncompetitive states seem not to count—and have zero power, according to standard measures of voting power—because they cannot influence the outcome in their states. But the electoral votes of these states are essential to a candidate’s victory, so they do count, but in a different way.

We propose a simple model that enables us to measure the setup power of voters in noncompetitive states by modeling how these states structure the contest in the competitive states, as illustrated by the 2012, 2008, 2004, and 2000 presidential elections. We define three measures of setup power—winningness, vulnerability, and fragility—and show how they pinpoint the advantages of the candidate ahead in electoral votes in the noncompetitive states. In fact, this candidate won in all four elections.

ABSTRACT. This paper investigates experimentally the effects of different polling systems on the structure of electoral competition. We study a divided majority problem with a unique Condorcet loser and a unique Condorcet winner. The electoral process is modeled through a repeated game with a plurality voting election as the stage game. Voters have no information about other voters' preferences but can observe past election outcomes. We show that the polling system employed to conduct pre-election polls can impact the electoral competition. An approval voting poll preceding plurality voting elections increases the viability of the Condorcet winner in comparison to a plurality voting poll or having no-polls preceding elections. The approval voting poll increases the vote share of the Condorcet winner in the elections and increases the number of election the Condorcet winner wins. Voting groups also reach higher welfare levels.

ABSTRACT. We study efficiency and information aggregation in common value elections with continuous private signals and informative priors. We show that small elections are not generally efficient and that there are equilibria where some voters vote against their private signal even if it provides useful information and abstention is allowed. This is not the case in large elections, where the fraction of voters who vote against their private signal tends to zero. In an experiment, we then study how informativeness of priors and private signals impact efficiency and information aggregation in small elections. We find that there is a substantial amount of voting against the private signal. Moreover, while most experimental elections are efficient, voting behaviour reveals little about voter's private information.

ABSTRACT. We investigated the processes underlying the violation of social norm expectations in the Ultimatum Game (UG) by using measurements of electrocortical activity in the human brain. Participants' EEGs were recorded while they took part in a series of one-shot UGs in the role of the responder. Prior to the task, they reported their expectations about proposer offers. Results showed that expectations significantly determined acceptance behavior in the UG. The medial frontal negativity (MFN), an event-related potential associated with conflict monitoring, reward prediction errors, and social pain, was significantly associated with expectations. The higher participants' initial expectations about the offers they were to receive, the larger was their MFN amplitude following unequal offers. Our findings imply that the MFN reflects the degree of social norm violation, and might serve as an emotional signal to bias behavior to enforce the social norm.

ABSTRACT. Group formation is a fundamental activity in human society. Humans often exclude others from a group and divide the group benefit in a fair way only among group members. Such an allocation is called in-group fair. Does natural selection favor an in-group fair allocation? We investigate the evolution of fairness and group formation in a three-person Ultimatum Game (UG) in which the group value depends on its size. In a stochastic model of the frequency-dependent Moran process, natural selection favors the formation of a two-person subgroup in the low mutation limit if its group value exceeds a high proportion (0.7) of that of the largest group. Stochastic evolutionary game theory provides theoretical support to explain the behavior of human subjects in economic experiments of a three-person UG.

ABSTRACT. We experimentally test the prediction of the perfect equilibrium concept in the three-player game Selten’s Horse (Selten 1975). At first sight, our data show little support of the perfect equilibrium and rather favor the intuitively unreasonable equilibrium. The observed dynamics of the game, however, are in line with learning direction theory whose impulse response dynamics converge on the perfect equilibrium. We tentatively conclude that the observed dynamics are in line with a convergence on the perfect equilibrium, but convergence takes time.

ABSTRACT. We analyze response times in an experimental Cournot oligopoly and obtain evidence suggesting that two prominent behavioral rules codetermine behavior in this setting. On the one hand, imitation of successful behavior is an intuitive behavioral rule corresponding to a more automatic process. On the other hand, myopic best reply is a deliberative rule corresponding to a more controlled process. While the previous literature has indirectly deduced the relevance of imitation from convergence properties observed in the lab, our evidence rests on the analysis of process data of individual decisions, independently of convergence. Our response-time predictions, derived from a formal dual-process model, point to a non-trivial asymmetry which is readily found in the data. We conclude that a dual view of behavior incorporating both rational optimization and evolutionary ideas on bounded rationality might provide a useful synthesis and help improve our understanding of how economic decisions are made.

ABSTRACT. Schelling proposed that payoff-irrelevant cues may systematically affect the outcome of tacit bargaining games. Early evidence supports this hypothesis, but suggests that payoff asymmetries may hurt focal point reasoning. We extend the investigation to tacit bargaining games in which players have imperfect information about each other’s payoffs, and this is common knowledge. When players know only their own payoffs, they fail to ignore unhelpful payoff information, and the effects of payoff-irrelevant cues on coordination and efficiency are relatively small. When no exact information about payoffs is provided, payoff-irrelevant cues have larger effects, but potential payoff asymmetries are still detrimental.

ABSTRACT. The computational nature of the platforms of key economic transactions, as well as the rising importance of incentives in networked computation, have led the two fields of Game Theory and the Theory of Computation to look closer into each other and often to engage in joint research. As a result, two decades later both fields have grown richer in insights and methodology, in wisdom and power. This talk will focus on certain key problems in this unexpectedly fertile research interface, in which significant progress was recently made: How are learning algorithms affected if the data providers - or the data itself - are strategic? Can we compute an approximate Nash equilibrium fast? And what are the principles of the optimal design of multidimensional mechanisms and online ad auctions?

ABSTRACT. The elegance and simplicity of Gale and Shapley's deferred acceptance algorithm (DA) has made it the algorithm of choice for determining stable matchings in a variety of settings. Each setting has imposed new demands on the algorithm. Among them are to how to handle complementarities and distributional constraints. However, the simplicity of the DA makes it difficult to accommodate these new considerations except in special cases. In this talk I will outline an alternative approach based on Scarf's lemma for tackling such problems.

Based on joint work with Thanh Nguyen.

ABSTRACT. We study a dynamic matching environment where individuals arrive sequentially. There is a tradeoff between waiting for a thicker market, allowing for higher quality matches, and minimizing agents' waiting costs. The optimal mechanism cumulates a stock of incongruent pairs up to a threshold and matches all others in an assortative fashion instantaneously. In decentralized settings, a similar protocol ensues in equilibrium, but expected queues are inefficiently long. We quantify the welfare gain from centralization, which can be substantial, even for low waiting costs. We also evaluate welfare improvements generated by transfer schemes and by matching individuals in fixed time intervals.

Joint work with Mariagiovanna Baccara and SangMok Lee.

ABSTRACT. We study mediated many-to-many matching in dynamic two-sided markets in which agents' private valuations for potential partners evolve stochastically over time, either as the result of exogenous shocks, or as the result of learning through experimentation. In many environments, the platform's profit-maximizing mechanism takes the form of a dynamic matching auction in which, in each period, agents submit bids for each possible partner and where the matches with the highest joint virtual score are implemented. The virtual score of each match is either myopic or takes the form of an index, accounting for the endogenous changes in the partners' expected match values. Each score depends only on the partnerns' current bids, the number of past interactions between the partners, and the partners' membership status (which captures vertical differentiation in preferences). Despite the complexity of the environment and the discrepancy between the agents' and the platform's value of experimentation, under the profit-maximizing equilibrium, bidding is straight-forward and myopic. We also contrast matching dynamics under profit maximization with their counterparts under welfare maximization. When all agents benefit from interacting with all other agents from the opposite side, profit maximization involves fewer and shorter interactions. This conclusion, however, need not extend to environments where certain agents dislike certain interactions.

Joint with Daniel Fershtman.

ABSTRACT. This paper analyzes a bilateral trade model where the buyer's valuation for the object is uncertain and she observes only a signal about her valuation. The seller gives a take-it-or-leave-it offer to the buyer. Our goal is to characterize those signal structures which maximize the buyer's expected payoff. We identify a buyer-optimal signal structure which generates (i) efficient trade and (ii) a unit-elastic demand. Furthermore, we show that every other buyer-optimal signal structure yields the same outcome as the one we identify, in particular, the same price.

Based on joint work with Daniele Condorelli and Anne-Katarin Roesler.

ABSTRACT. More than half a century after the first experiment on the finitely repeated prisoner’s dilemma, evidence on whether cooperation decreases with experience, as suggested by backward induction, remains inconclusive. We will explore a meta-analysis of prior experimental research and the results of a new experiment to elucidate how cooperation varies with the environment in this canonical game. We will describe forces that affect initial play (formation of cooperation) and unraveling (breakdown of cooperation). First, contrary to the backward induction prediction, the parameters of the repeated game have a significant effect on initial cooperation. We will identify how these parameters impact the value of cooperation - as captured by the size of the basin of attraction of Always Defect - to account for an important part of this effect. Second, despite these initial differences, the evolution of behavior is consistent with the unraveling logic of backward induction for all parameter combinations. Importantly, despite the seemingly contradictory results across studies, we will establish a systematic pattern of behavior: Subjects converge to using threshold strategies that conditionally cooperate until a threshold round; and conditional on establishing cooperation, the first defection round moves earlier with experience. Simulation results generated from a learning model estimated at the subject level provide insights on the long-term dynamics and the forces slowing down the unraveling of cooperation.

ABSTRACT. We consider a game between a non-benevolent incumbent politician and a bureaucrat, where both can illegally appropriate public resources. The corrupt politician uses the resources to finance political campaigns and the bureaucrat to buy goods and services. Any illegal withdraw from the treasury diminishes the politician's capability to improve social welfare. Politicians care about votes, bureaucrats care about money, and voters care about social welfare and are influenced by campaigns. We analyze the role of society's educational level and of different punishing structures. We find that: i) the bigger the appropriation by the bureaucrat, the smaller the optimal appropriation by the politician; ii) the more educated the politician's constituency, the less lenient to corruption he will be; and iii) the more sensitive to marginal increases in corruption the punishing structure is, the smaller the players' optimal appropriation.

ABSTRACT. 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 offer a straightforward interpretation for the product operator underlying the Nash solution.

ABSTRACT. In a seminal paper, Meyer [David Meyer, Phys. Rev. Lett. 82, 1052 (1999)] showed that a player using a quantum strategy can win against a classical player 100% of the time in the penny flip game. Here we make a slight modification to this game, with the two players sharing an entangled state to begin with. We have the quantum player making unitary transformations to his qubit while the classical player makes use of a mixed strategy wherein he either flips or not his qubit with some probability "p". The 100% win record of a quantum player is drastically reduced and for a particular probability "p" the classical player can even win against the quantum player. This is of possible relevance to the field of quantum computation as we show that in this quantum game of preserving versus destroying entanglement a particular classical algorithm can beat the quantum algorithm.

ABSTRACT. This model applies the model for team performance developed by Holm- strom (1982) to a setting in which the good produced by the team is a public good. It shows that when the team good is a public good, for example an essay mark then the team members are inclined to exert the efficient level of effort and no free-riding occurs. Contrary to this results, group members are not willing to acknowledge each others contribution since this diminishes the value they can assign to their own contribution.

ABSTRACT. The paper describes the development of financial markets and changes in the nature of economic growth using the theory of cooperative games. It turned out that various applications and contexts correspond to numerous possible solutions of standard tasks, e.g. Nash (S, d) bargaining problem. We show that under conditions of sufficiently effective financial markets the question of the relationship between efficiency and equality, which is typical of the theory of social welfare, may be replaced by the question of making full utilization of investment opportunities associated with the acquisition, preservation and application of human capital. Sufficiently efficient functioning of financial markets (in the above mentioned sense) assumes also good functioning of such financial market instruments, e.g. human capital contracts associated with the use of transferred prices and mediated utilization of transferred prices. The development of financial markets in the above direction is prerequisite to economic growth.

ABSTRACT. We consider multi–leader–follower games where the actions chosen by any leader are observed by only one “exclusive” follower. Many real–world situations can be modeled in such a way, for example see Pagnozzi and Piccolo (2012), where competing manufacturers delegate retail decisions to exclusive retailers offering a private contract. Such games, called multi–leader–follower games with vertical information, may have an infinity of Nash equilibria but it is not possible to refine using the concept of subgame perfect Nash equilibrium since the associated extensive forms have no proper subgames. This motivated the introduction of selections of Nash equilibria based on the beliefs that each follower has about the actions chosen by the leaders rivals of his own leader. We illustrate the effectiveness of the concept for a general model and we investigate the existence of such a selection for significative classes of problems satisfying conditions of minimal character on possibly discontinuous data.

ABSTRACT. The competitive behavior of a given number of rational agents who simultaneously invest their resources in varying proportions in a number of different environments can be modeled by a new class of games called Multi-Games (MGs). In particular, multi-games model the investment in various continental and national markets in a global economy. Different environments are denoted as basic games. The total payoff for each player is assumed to be a convex combination of the rewards with the respect to the investment weights of basic games. When the players' investment weights in the basic games are, with a given conditional probability distributions, their private information or types, then a class of Bayesian games is obtained. In this paper, we derive an efficient algorithm to find a Bayesian Nash equilibrium for a class of multi-games called completely mixed regular double games.

ABSTRACT. This paper investigates in a theoretical framework the determinants of the "diabolic loop" between sovereign and bank credit risk that was the distinctive feature of the financial instability which spread out from 2009 to 2012 in the periphery of the euro area. In the proposed model there are strategic complementarities within and between the group of bank creditors and the group of sovereign creditors. The negative feedback loop originates from the fact that banks' holdings of sovereign bonds, being liquid assets, may be sold or not rolled-over in case of liquidity shortage. In the unique equilibrium of this model, as the private information that agents receive becomes very precise, the correlation between the probability of a banking crisis and a sovereign debt crisis approaches one and depends solely on bank balance sheet variables.

ABSTRACT. In this paper, an interactive multi-portfolio optimization framework is proposed from the perspective of self-interested clients. A key distinction of this approach from the classical methods is a built-in factor in the expected return related to client preferences on the bunched trading, which is termed as information pool and helps prompt the perceived fairness. It allows the clients, who should be the ultimate evaluator of fairness, to be enrolled in the interactive optimization process, differing from the approach centrally decided by the portfolio manager. Instead of the traditional share-based pro rata fashion, the manager split the market impact cost in accordance to a novel concept called relative contribution, which correctly captures the market impact of each individual account. The numerical study suggests that this interactive approach outperforms the relevant prominent solutions in literature, and information pooling does improve the portfolio performances.

ABSTRACT. The paper considers the market of software products. Regularly this markets are suffering from dealing with a problem of counterfeit or pirate products that are sold. Taking this into account the paper is solving the problem of price equilibrium on this market. The software company set the price and the quality of the software product while the counterfeit or pirate company suggest the consumers the product of the lower quality. First the general model is analyzed and price equilibrium is defined. Second, the monopoly case is considered separately and optimal software price is defined. Finally, it is supposed that there are two companies that produces original software on the market who differentiate in quality, and there are two pirate companies who produces this software. The duopoly case is analyzed and equilibrium prices for competing companies are obtained in the explicit form.

ABSTRACT. Consider two politicians who decide whether to follow what they believe the public wants, choose the option that secures their private gain. Laws are passed when the politicians reach a unanimous decision. The public only rewards a politician when a law is passed, or when the politician is the only one whose action coincides with the public decision. We find that if the politicians are good enough decision-makers, a sufficiently high public regard in policy implementation given moderate private agenda payoffs pushes the politicians to take the action that generates a public benefit, implementing a socially optimal law. For very low decision-making skills, at sufficiently high policy rewards, we find that they vote for the same action to pass a law regardless of what the public wants. This gives rise to politicians converging to a decision that neither provides them with a private benefit nor follows exactly the public decision.

ABSTRACT. I consider an enrichment of the framework of extensive form games, where players can make during the game unilaterally binding oers for transfer of utilities to the player who is to make a move, in order to provide an incentives for that player to deviate from a Backward Induction prescribed strategy for a common benefit. I characterise the solutions of such games for the 2-player case with perfect information and then extend the framework to coalitional interplay oers in N-player games, where the problem reduces in general to solving multi-party bargaining and coalitional games, requiring deeper analysis for which this paper is just a starting point.

ABSTRACT. An assignment game is a cooperative game with its player set divided into two groups, where a coalition receives a positive payoff only if it contains at least one player from both groups. In other words, it concerns a bipartite matching, with a payoff defined for each pair (payoffs for larger coalitions are determined additively). The 1994 paper by Solymosi and Raghavan gives an algorithm for finding the nucleolus (an optimal solution) of such a game. In this paper, we examine the special case in which the payoffs for each pair take binary values, an indicator of whether or not the pair is compatible. For this case, we present two new, simpler algorithms which capitalize on the graphical aspect of the previous one. These algorithms require O(n^4) operations. We also discuss sociological implications of this solution, considering each pair's split of the payoff to be their relationship's balance of power.

ABSTRACT. Evidence in psychology and experimental games suggests that people tend to overweight personal experience relative to descriptive information and observations. However, little is known about the relative importance of experience in individual economic decisions. The present paper investigates this issue using an experimental investment task with choice feedback and belief elicitation. The experiment creates tension between the Bayesian strategy and experience-based learning. We document the double-channeled effects of experience: controlling for beliefs, participants significantly relied on experience, behaving consistently with the law of effect, i.e. repurchasing more assets from which they gained than lost; additionally, experience also distorted beliefs, inducing more optimism on assets from which they gained and pessimism on unowned assets. We then structurally estimated the parameters of the reinforcement-learning model and found it significantly added predictive power to expected utility models. These results can inform theories of individual sequential decisions and the design of behavioral interventions.

ABSTRACT. This paper addresses the problem of determining when coordination is beneficial. I describe a negative externality game containing a ``worsening parameter” and develop a framework linearizing this parameter for tractable examination. The worsening parameter can be classified according to ``own effect'' -- changing the marginal utility of a player’s own action, ``opponent effect'' -- altering the marginal externality, or ``submodular effect'' – strengthening the game’s submodularity. Using this framework, I examine the sufficient conditions for parameter changes to move non-cooperative and cooperative solutions in opposite directions. In a symmetric game, an increase in own effect will increase the distance between utility and action level of the non-cooperative and cooperative solutions. In a non-symmetric game, there are sufficient conditions on the second derivatives which give this pattern as well. I argue that situations behaving in this manner have more benefit to coordination through the increased range in actions.

ABSTRACT. The epistemic program in game theory aims at making explicit the assumptions that underlie the basic concepts of the discipline.\ Thus far, this literature has not had much to say about the so-called ``trembling hand mistakes'' in games.\ This paper aims to fill in this lacuna.\ We argue that the idea of players who can make and reason about mistakes fits well with the overall direction of the epistemic program.\ We present an extension of Robert Stalnaker's epistemic models of extensive games with simultaneous moves, in which the players can make trembling hand mistakes.\ In particular, this means that when a player observes an unexpected move, she has to figure out whether it is a result of a deliberate choice or simply a mistake. We show how our models of beliefs in games sheds new light on the relationship between backward and forward induction.

ABSTRACT. The paper defines a non-cooperative simultaneous game with a number of potential deviators being a parameter of the game. A definition of the game embeds a coalition structure partition mechanism. The game has an equilibrium in mixed strategies. The equilibrium  encompasses intra and inter group externalities and an individual payoff allocation that make it different from a strong Nash, coalition-proof equilibrium and some other equilibrium concepts. We offer a non-cooperative stability criterion to describe a robustness of an equilibrium strategy profile to an increase in a number of deviators. The criterion may serve as a way to measure self-enforcement or a trust for the equilibrium in terms of a number of potential deviators and the mechanism design.

ABSTRACT. We study local equilibria resulting from perfect competition among a fixed number of firms in a screening market. A condition that guarantees zero profits to all firms at the given local equilibrium prices is the existence of a linear ordering on the available alternatives that dictates the preferences of the firms in those cases in which consumers display indifference; this condition is shown to hold in classical models of insurance markets. A partial converse, giving robust conditions under which this condition is violated and positive profits result, is also presented.

ABSTRACT. Cooperative games with coalitions’ values represented by intervals, called interval cooperative games for short, are discussed in this paper. The main purpose of this paper is to develop fast and effective quadratic programming methods for computing interval least square prenucleolus and nucleolus of interval cooperative games. Firstly, based on the square excess which can be interpreted as a measure of the dissatisfaction of coalitions, we construct an auxiliary quadratic programming model for interval least square prenucleolus of any interval cooperative game and derive its analytical solution, which is used to determine players’ interval payoffs. Then, two iteration algorithms are developed to compute the interval least square nucleolus of interval cooperative games based on the derived analytical solution of the interval least square prenucleolus. Finally, we study some useful and important properties of the interval least square prenucleolus and nucleolus such as existence and uniqueness, efficiency, additivity, symmetry, and anonymity.

ABSTRACT. This poster presentation has been cancelled.

ABSTRACT. Proportionality is a primary principle generally accepted when dividing a commodity between a set of agents who are evaluated with respect to a certain characteristic. The proportional rule provides a result in which all the agents obtain the same proportion with respect to their references.

When multiple characteristics are considered, there is no obvious way to define proportionality. In this paper we incorporate the extensions of two crucial properties which are inherent to proportionality: the proportions obtained with respect to the different references cannot be improved simultaneously, and the result does not depend on the scale in which each of the characteristics is measured.

In our approach, the choice of a single allocation for the division problem is understood as a negotiation between parties, each one "supporting" one of the characteristics. This fact pemits the design of specific rules for multiple reference division problems based on the axiomatic bargaining solutions.

ABSTRACT. This paper deals with the equilibria of games when the agents have multiple objectives and therefore, their utilities cannot be represented by a single value, but by a vector containing the various dimensions of the utility. Our approach is based on the incorporation into the model of information about the preferences of the agents, together with adittional decision rules. This allows us to identify the equilibria according to this preferential information. The potential application of the theoretical results is shown with an analysis of a mixed oligopoly in which the agents value additional objectives other than their own benefit. These objetives are related to social welfare and to the profit of the industry. The flexibility of our approach provides a general theoretical framework to analyse a wide range of strategic economic models.

ABSTRACT. We construct foundations for utility function and models of consumer behavior, voter’s behavior, and social choice which are based on model of attitude choice. The key idea of our approach is that the consumer’s decision in any situation of choice has dual nature: it relates both (1) choice of bundle of characteristics of goods from set of choice which characterizes the particular situation, and (2) choice of an attitude (vector of weights of characteristics) from behavior menu. We show that in the Nash collective bargaining model, arbitrator possesses a social behavior menu and acts in a similar way as an individual. Social behavior menu consists of vectors of average attitudes of the members of the collective. We provide examples of formation of the individual behavior menus as results of an evolutionary process and of spreading of attitudes in network. By use of our approach we develop Hillman’s political economy games.

ABSTRACT. In this paper, we use the VFJS model proposed by Vickers (1985), Fersthman and Judd (1987) and Sklivas (1987). We find that even an inefficient firm can obtain higher profits than its rival if it compensates its manager with a market share or relative profit incentive whereas its rival is just owner-led. We consider duopolies where the efficient firm remunerates its manager with a market share incentive and the rival firm uses a sales, relative profit or market share incentive. We find that the efficient firm could achieve Stackelberg Leadership profits if the difference between marginal costs of two firms exceeds a certain benchmark. Finally, we determine that the less efficient firm owner always prefers to use a managerial bonus based on relative profits. If the difference between marginal costs is large enough then the more efficient firm owner always prefers to use a managerial bonus based on market share.

ABSTRACT. Abstract I use a normal quadratic game à la Calvo-Armengol and al (2015) to find a theoretical explanation to buzzes on internet. More precisely I solve a game with strategic complementary, public and private information and information acquisition. I find the existence of the unique symmetric linear equilibrium and derive some comparative static with respect to it: an increase of the size of the population leads individuals to accord more importance to public information relatively to private information. I compare this equilibrium to a planner equilibrium and find that in the competitive equilibrium there is an over-weighting on the public information, following our interpretation of the model this over-weighting is a Buzz.

ABSTRACT. The Brazilian electricity sector adopted the auction mechanism to allocate the assets of transmission lines. The results are impacted by interdependence and the capacity constraint. We develop a theoretical model to evaluate these issues in a procurement auction sequence. The theoretical model shows that the joint action of interdependence and of capacity restriction decreases the competitiveness of the participants and the payoff of auctioneers. Using switching regression model, we found that the interdependence affected the probability of winning and the behavior of the bidding. In addition, plans for rapid expansion of the grid can be frustrated if these effects are not considered

ABSTRACT. The aim of this paper is to develop an effective methodology for solving matrix games with payoffs of intuitionistic fuzzy sets (IFSs). In this methodology, a new ranking order relation of IFSs is proposed and the concept of Pareto Nash equilibrium solutions of matrix games with IFS payoffs is firstly defined. It is proven that the solutions of matrix games with IFS payoffs are equivalent to those of a pair of bi-objective programming models. The models and method proposed in this paper are illustrated with a numerical example and compared with other methods to show the validity, applicability and superiority.

ABSTRACT. Governments of different countries try to lower the entry cost in public procurement in order to decrease public spending. The purpose of this paper is to examine how the entry cost influences favoritism and procurement prices in the corrupt environment. We adapt the model of selective entry and find that lower entry cost always reduces the contract price paid by the benevolent procurer, but at the same time may make favoritism more stable. Thus the entry cost does not affect the contract price paid by the corrupt procurer or increase it. We illustrate this result using case study on gasoline procurement in Russia where the entry cost of companies was decreased by e-procurement reform. This allows us to examine how changes in entry costs influence competition of companies and procurement prices in auctions.

ABSTRACT. This paper concerns property value diminution (PVD) of residential properties contaminated by a commercial / industrial source or a spill. Although long-term, negative price effects might seem inevitable, transaction data fail to confirm negative price effects when properties have been cleaned up. We develop an economic model which shows how this can occur.

In our model individuals are free to purchase units of housing in just one of two neighbourhoods, and all units of housing are identical. A pollution event occurs which partially contaminates of one of the two neighbourhoods. Some individuals face life event such as a job change. Moving is optional and individuals may commute. Owners are better informed about the contamination of their house than buyers. Buyers can, however, determine the information by paying a search cost.

We present a worked example in which, depending on search and commute costs, property value diminution may not occur.

ABSTRACT. In the paper the existing microeconomic/Game Theory models (such as John Morgan’s) of financing public good by means of lotteries are analysed. Several other models are proposed so as to compensate certain shortcomings of the above-mentioned models. Those models are based on the assumption that lottery organiser knows the welfare function of each participant of a lottery except the moral component.

The analysis of the classical and newly introduced models allows one to make the following conclusions (the major ones).

1. It is proved that the socially optimal level of providing public good is achievable.

2. The newly introduced notion of ‘relative expectations’ along with taking into consideration the influence of advertising and the moral utility factor for the participants of a lottery provides a new approach to the ‘free-rider’ problem.

ABSTRACT. This poster has been cancelled.

ABSTRACT. The aim of this paper is to develop a nonlinear programming method for solving interval multiobjective cooperative games in which importance of objectives and coalitions are taken into consideration. Firstly, we define the concepts of interval cores of interval multiobjective cooperative games and satisfactory degrees of comparing intervals with inclusion and/or overlap relations. Then, we construct an auxiliary nonlinear programming model for computing interval cores of interval multiobjective cooperative games in which importance of objectives and coalitions are considered. The derived auxiliary nonlinear programming model can be directly and easily solved. The proposed ranking method of intervals is an important development of existing interval ranking methods. Furthermore, the proposed model and method of interval cores for interval multiobjective cooperative games is an extension of that for classical cooperative games. The effectiveness, feasibility and applicability of the proposed model and method are illustrated with a real example.

ABSTRACT. The paper provides a reformulation of Nash equilibrium based on optimization approach. The set of Nash equilibria, if it is nonempty, is identical to the set of optimizers of a real-valued function, which connects the equilibrium problem to the optimization problem. Using this characterization, we study the interchangeability of Nash equilibria, and show that existing results on two-person (i) zero-sum games and (ii) supermodular games can be derived, in a unified fashion, by the lattice structure on optimal solutions.

ABSTRACT. We apply a simulation model to examine evolutionary population dynamics based on random interactions between individuals, within global and local neighbourhoods. Interactions occur in a continuous two-dimensional space. These interactions yield fixed numbers of offspring that are based on a given fitness matrix. These offspring are randomly placed within a certain spatial radius from the parents. Besides, we use a control mechanism to make sure that the neighbourhood population density does not grow too high. We compare our simulation results with theoretical population development as predicted by the replicator dynamics, and also compare it with practical results derived in an experimental set-up. Our results support the theoretical analysis of models with global interactions and also imitate real-life situations with local interactions in a realistic way. Next to this, we study the influence of several parameters on population development, like individual lifetime, radius of interaction, radius of placing offspring, etc.

ABSTRACT. We investigate how concerns for social status may a¤ect individuals’preferences for redistribution and give rise to a non monotonic relation between an agent’s income and his preferred tax rate. We postulate that agents are heterogeneous across two dimensions, productivity and social class, and we de…ne an individual’s social status as his relative standing in the distribution of a weighted average of these two components. Redistribution simultaneously a¤ects labor supply as well as the weights that determine social status. As such, taxation becomes a way to preserve or modify the current social ranking. In equilibrium, individuals with similar productivity but di¤erent social class support di¤erent tax rates and the actual tax rate is supported by a coalition of individuals who belong to di¤erent classes.

ABSTRACT. The standard mechanism design approach to collusion in auctions posits a disinterested third party who designs and implements the mechanism. However, in many environments, collusion agreements are likely to be both proposed and executed by the involved parties. When this is the case, the standard approach may be inadequate, as it neglects potential information leakages. We model collusion in a second-price auction by having one informed bidder propose to another. We find that all-inclusive cartels will be unable to extract all of the surplus from the seller, and equilibrium outcomes will generally be (socially) inefficient. While both bidders are better off than without colluding, we exhibit a first-mover (relative) disadvantage: information leakages and a corresponding “winner’s curse” cause the proposer to be worse off than the receiver. Introducing outside competition compounds this winner’s curse and reduces the potential gains to collusion, resulting in substantially higher revenues for the seller.

ABSTRACT. Consider a long-term auctioneer who repeatedly sells identical or similar items and who might disclose selective information about past bidding. We present a theory that yields different predictions about bidding depending on the information bidders are provided with. We focus on the disclosure of all bids and of winning bids only. Our theory is based on a selection bias: some bidders who are presented with historical winning bids mistakenly best-respond to that distribution, failing to realize that winning bids are not representative of all bids. In the steady state, this bias results in higher bids and auction revenue than when all bids are presented. Our experimental test confirms the qualitative predictions of the theory. On the theory side, our findings challenge the predictive power of Bayesian Nash Equilibrium based on rational bidders. On the market design side, they underline the role of historical market information as a key design choice.

ABSTRACT. This paper examines the feasibility of collusive mechanisms in single-unit auctions. A model is built up with private and common value information asymmetry and continuous type space. We show that an incentive compatible bid coordination mechanism (BCM) does not exist if common value uncertainty is present. This result contradicts actual antitrust cases where common effects or resale opportunities created uncertainty about valuations, but a price-fixing cartel was formed. We solve the puzzle by relaxing the assumption that all bidder types truthfully reveal their private information. The introduced Bayesian bid coordination mechanism (BBCM) exists if the main source of information asymmetry is private value. In that case, a designated ring member can signal high valuation and suppress competition. Our results demonstrate the rationale behind cartel mechanisms with pre-auction knockouts.

ABSTRACT. We analyze the contracting problem of a principal who faces an agent with private information and cannot commit to not renegotiate a chosen contract. We propose an infinite horizon negotiation protocol in which renegotiation is frictionless and there are no restrictions on how many times the contracts can be renegotiated. We provide a general characterization of renegotiation-proof outcomes and show that those outcomes are supported by a Perfect Bayesian Equilibrium of the negotiation game. The general characterization of renegotiation-proof outcomes provides a powerful and simple to use tool for finding such outcomes in specific environments. We proceed by applying the results to adverse selection environments with private and common values. We show that with private values and common values of the 'Spence' type only fully efficient and separating contracts can be renegotiation proof. With common values of the 'Rothschild-Stiglitz' type inefficient and (partial) pooling contracts may constitute renegotiation-proof outcomes.

ABSTRACT. This talk has been cancelled.

ABSTRACT. Horizontal subcontracting—outsourcing activities to competitors—stifles competition to the extent that subcontractors capture the subcontracting surplus. In our framework, price-competing firms choose the subcontracting terms to maximize profits. A merger does not create cost savings at the industry level because of efficient subcontracting before the merger. It does, however, reduce the merged parties’ in-house costs. Merging is therefore (i) profitable because it credibly limits the amount paid to subcontractors, and (ii) price-reducing as it decreases outsiders’ profits from subcontracting.

ABSTRACT. The effect of a treatment in a multiagent economy is causal if the treated economy would be different relative to the control economy. Causal effects measured in an equilibrium of the economy, the long- term causal effects, are more representative of the value of such treatments. However, the statistical estimation of long-term causal effects is difficult because it has to rely, for practical reasons, on short-term experimental data. We propose a methodology to estimate long-term causal effects, which relies on a model of agent behavior that plays a two-fold role. First, it predicts how agents would behave under different assignments, and, second, it predicts how agents would behave in a long-run equilibrium. These two prediction tasks enable the estimation of long-term causal effects under suitable assumptions, which we state explicitly.

ABSTRACT. This paper focuses on oligopolistic markets in which indivisible goods are sold by multiproduct firms to a continuum of homogeneous buyers, with measure normalized to one, who have preferences over bundles of products. By analyzing a kind of extended contract schedules that discriminate on exclusivity, the paper shows that pure strategy efficient equilibria always exist in such settings. There may also exist inefficient equilibria in which the agent chooses a suboptimal bundle Inefficient equilibria can be ruled out by either assuming that firms are pricing unsold bundles at the same profit margin as the bundle sold at equilibrium, or imposing the solution concept of subgame perfect strong equilibrium, which requires the absence of profitable deviations by any subset of principals and the agent. We provide a characterization of the equilibrium pure strategies and show that each principal's set of equilibrium contracts of minimum cardinality may contain at least three offers.

ABSTRACT. Should principals explain and justify their evaluations? Suppose the principal’s evaluation is private information, but she can provide justification by sending a costly unverifiable message. If she does not provide justification, her message space is restricted, but the message is costless. I show that the principal justifies her evaluation to the agent if the evaluation indicates bad performance. The justification assures the agent that the principal has not distorted the evaluation downwards. In equilibrium, the wage increases in the agent’s performance, when the principal justifies her evaluation. For good performance, however, the principal pays a constant high wage without justification.

ABSTRACT. This paper studies optimal information disclosure under the threat of collusion. A principal seeks to procure a good from one of two agents who can collude against the principal. The first agent has a publicly known cost of production and the second agent’s cost is his private information. The principal decides how much information the first agent receives about the costs of the second agent. In the choice of the optimal disclosure policy, the principal faces a trade-off: More information disclosure makes the elicitation of private information easier but facilitates collusion at the same time. It is optimal for the principal to partially disclose information. Under the optimal information structure, none of the agents receives a positive information rent.

ABSTRACT. We analyze a principal-supervisor-multiple agent hierarchy, in which information is soft and the supervisor may make a mistake that observes a wrong signal on the agents' effort levels. The supervisor and the agents can collude against the principal. We show that supervision is still valuable in the hierarchy, and the principal sends the supervisor according to both the output level and the effectiveness of the supervision technology. More interestingly, the characterization of optimal contract involves two features that collusion-proof implementation is always optimal with low output. If, however, high output is realized, tolerating a certain scope of collusion between the supervisor and a subset of the agents to take place would be beneficial, as this can `correct' potential mistakes from the supervisor and help in providing incentives to the agents.

ABSTRACT. An investor has some funds invested through investment firms. She also has additional funds to allocate among these investment firms according to the firms' performance. While the investor tries to maximize her total expected earnings, each investment firm tries to maximize the overall amount of funds it will be allocated to manage. A reward scheme is a rule that determines how funds are to be allocated among the investment firms based on their past performance. A reward scheme is optimal if it induces the (self-interested) firms to act in accordance with the interests of the investor. We show that an optimal reward scheme exists under quite general conditions.

ABSTRACT. This article sets forth a model of multilateral negotiations with alternating offers and voting in which members bargain over the distribution of equity shares prior to engaging in joint production. We endogenize the origin of the fund to be divided among members of a group and analyze how the committee size, voting rule, and impatience of players affect total production via the equity agreements reached. We focus on the trade-off between rent-sharing and rent-generating incentives faced by bargaining partners, an aspect that is absent in the standard divide-the-dollar games but that is certainly present in multiple economic settings.

ABSTRACT. A planner wants to carry on a project involving several firms. In many cases the planner, for instance the Spanish Administration, includes in the contract with the firms some penalties to be paid by the firms in case the project is delayed. We discuss two ways of include penalties. In the first one penalties are applied only when the whole project is delayed. In the second one is applied always. We compare both. The optimal penalty (for the planner) is larger in the second one. The utility of the planner is always lager or equal in the second one. The utility of the firms is always larger or equal in the first one.

ABSTRACT. We study one-sided matching problem, also known as roommate problem, where a group of people need to be paired in order to be assigned to certain location. We assume that number of rooms are limited and thus no student can live by himself. Each student has strict preferences over only their roommates. Central notion in matching is stability. We consider exchange-stability of Alcalde~\cite{alcalde85}, which is immune to group of students exchanging their rooms/roommates with each other. He shows that exchange-stable matching may not always exist. We define sufficiency conditions to guarantee (i) existence of exchange-stable matching and (ii) that starting from an arbitrary matching, there exist a sequence of matchings, each being obtained from the previous one by ``satisfying" a blocking group of students, such that the final matching is exchange-stable.

ABSTRACT. Motivated by our experience running the computational side of a large nationwide kidney exchange, we present FutureMatch, a framework for learning to match in a general dynamic model. FutureMatch takes as input a high-level objective decided on by experts, then automatically (i) learns based on data how to make this objective concrete and (ii) learns the "means" to accomplish this goal---a task, in our experience, that humans handle poorly. It learns the potentials of elements of the current input graph offline, translates these to weights, and performs a computationally feasible batch matching that incorporates dynamic, failure-aware considerations through the weights. We validate FutureMatch on real fielded kidney exchange data. It results in higher values of the objective. Furthermore, even under economically inefficient objectives that enforce equity, it yields better solutions for the efficient objective (which does not incorporate equity) than traditional myopic matching that uses the efficiency objective.

ABSTRACT. We study the allocation of indivisible objects (e.g. school seats) by lotteries. Agents have preferences over different objects and have different priorities at different objects. For this model, a lottery is ex-ante stable if there does not exist an agent-object pair such that we can increase the probability of matching this pair at the expense of agents who have lower priority at the object and of objects which are less preferred by the agent. We show that ex-ante stable lotteries have small support. The number of pairs in the support depends on how many indifferences in the priorities the lottery exploits. In the extreme case where no object is matched with positive probability to two equal priority agents the lottery is almost degenerate. Otherwise the size of the support is determined by the number of equal priority agents that are matched to the respective objects.

ABSTRACT. We study the existence and uniqueness (or order independence) of iterated elimination procedure from a choice-theoretic viewpoint. We show a general existence result of iterated elimination procedure on an abstract reduction system. We identify a sufficient condition of "Monotonicity*" for the order independence and, in (in)finite games, we provide a full characterization of Monotonicity*. Our approach is applicable to any form of iterated elimination processes in arbitrary strategic games, e.g., iterated strict dominance, iterated weak dominance, rationalizability, etc. As a by-product, we offer a solution to Dufwenberg and Stegeman's (2002) puzzle on order dependence of iterated strict dominance in infinite games.

ABSTRACT. Maximin strategies disregard mutually beneficial or harmful situations in general, which relates to most economic interactions. To capture this, we introduce a new concept, maximin equilibrium, that extends the maximin strategy solution to non-strictly competitive games by incorporating individual rationality constraint--players would not harm others unless it's in their self-interest. In particular, the maximin equilibrium value coincides with the minimax value in zero-sum games. We show that every finite game possesses a maximin equilibrium. As an example, maximin equilibrium selects the two Pareto dominant Nash equilibria in the battle of the sexes game; however, it may select a Pareto dominated solution such as the Nash equilibrium in the prisoner's dilemma. Both concepts may, nevertheless, give different predictions, as we illustrate in the so-called inheritance game.

ABSTRACT. A game is solvable if the set of Nash equilibria is nonempty and interchangeable. A pairwise solvable game is a two-person symmetric game in which any restricted game generated by a pair of strategies is solvable. We show that the set of equilibria in a pairwise solvable game is interchangeable, that a pairwise solvable game is solvable if it is quasiconcave at the diagonal, and that if in addition the game is finite then an iterated elimination of weakly dominated strategies converges precisely to the set of all Nash equilibria. In particular, the game is dominance solvable. Applying these results to relative payoff games, we establish simultaneous existence of Nash equilibrium and evolutionary equilibrium in a class of pairwise solvable games. All the results are applicable to symmetric contests, such as the rent-seeking game and the rank-order tournament, which are shown to be pairwise solvable.

ABSTRACT. The existing repeated games models (e.g. limit-of-means, discounted sum, and overtaking) presume that the preferences in the repeated game are determined, in one way or another, by the sum of the utilities of the stage games. The utilities in the stage game, however, are von Neumann Morgenstern (NM) utilities, for which (we argue) the conceptual meaning of adding utilities across separate games is unclear. Furthermore, we show that by assuming such additive utilities the classic folk theorems fail to account for some natural equilibria. We develop an ordinal framework for the study of repeated games, using the players' preferences alone (without the additivity assumption). Within this framework, we establish the associated folk-theorem, which generalizes the classic ones. In particular, the generalized theorem establishes the existence of equilibria unaccounted for by the classic theorems.

ABSTRACT. In this article we study the Nash equilibrium in infinitely repeated bimatrix games where payoffs are determined by reactive strategies; players’ payoff functions are defined as players’ benefits averaged over game rounds. Reactive strategies are stochastic 1-memory strategies such that a probability of players’ actions depends only on the opponent’s preceding move. We provide a characterization of all Nash equilibria in the class of reactive strategies. Then we derive a characterization for all symmetric stage games admitting Nash equilibria in the class of reactive strategies. Obtained results are compared with the available ones for 1-memory strategies. Extensively using examples, we illustrate new effects connected with Nash equilibria in the class of reactive strategies.

ABSTRACT. We consider repeated zero-sum games with incomplete information on the side of Player 2 with the total payoff given by the non-normalized sum of stage gains. In the classical examples the value V_N of such N-stage game is of the order of N or square root of N for large N.

We deal with another asymptotic behavior of V_N observed for the discrete version of the financial market model introduced by De Meyer and Saley. For this game Domansky and independently De Meyer with Marino found that V_N remains bounded as as N tends to infinity. This game is almost-fair, i.e., if Player 1 forgets his private information the value becomes zero.

We describe a class of almost-fair games having bounded values in terms of an easy-checkable "trigger" property of the auxiliary non-revealing game.

ABSTRACT. We introduce evolutionary competition between adjustment processes in the Cournot oligopoly model. Our main focus is on rational play versus a general short-memory adaptive adjustment process. We find that, although rational play has a stabilizing influence, a sufficient increase in the number of firms in the market tends to make the Cournot-Nash equilibrium unstable. Moreover, the interaction between adjustment processes naturally leads to the emergence of complicated endogenous fluctuations as the number of firms increases, even when demand and costs are linear.

ABSTRACT. This paper investigates the evolution of Cournot and Bertrand firms in a replicator dynamic. At each time period, every firm is randomly matched with a Bertrand or a Cournot firm to play a duopoly game. The distribution of Cournot and Bertrand firms is determined by their relative average profits. When homogeneous products are produced with identical cost, we find that Bertrand and Cournot firms will co-exist in the long run with Cournot firms being the majority. By contrast, if homogeneous products are produced by different costs, then both-type firms co-exist or only Cournot firms survive in the long run. When differentiated products are produced, both-type firms co-exist, Cournot firms survive alone, or Bertrand firms survive alone in the long run. For the co-existence equilibrium, the distribution of both-type firms will depend on the sizes of market demand, firms' marginal costs, and the product differentiation levels.

ABSTRACT. This paper presents a non-Walrasian equilibrium for an evolutionary model in the asymmetric oligopoly setup where firms have different cost functions to produce a homogenous good. Then, using revealed preference approach introduced by Carvajal, Deb, Fenske, and Quah [Econometrica 2013], I derive the testable conditions of the evolutionary oligopoly model. Therefore, without making any parametric assumption regarding to the demand curve and the cost function, this approach characterizes a set of conditions (restrictions) on observational dataset to be consistent with the non-competitive evolutionary equilibrium. An empirical application to crude oil market with main producers is presented and I compare the rejection rates of both Cournot and evolutionary hypotheses.

ABSTRACT. We study heterogeneous player rationality in games on networks with strategic substitutes. In the local public good game we distinguish two types of players, myopic (one-shot optimizers), and sophisticated (long-run optimizers). Starting from a static equilibrium, the game advances in two ways: by the sophisticated players' deviations to progress the game into a state where they earn higher payoffs, and by the myopic players' reactions to the deviations. We call Nash equilibria with no possible profitable deviations sophisticatedly stable. We derive conditions of stability in some networks and find that only experts and free riders equilibria can be stable. Stable equilibria also conflict with principles of stability and equality.

ABSTRACT. We contrast and compare three ways of predicting efficiency in a forced contribution threshold public good game. The three alternatives are based on ordinal potential, quantal response and impulse balance theory. We report an experiment designed to test the respective predictions and find that impulse balance gives the best predictions. A simple expression detailing when enforced contributions result in high or low efficiency is provided.

ABSTRACT. Share equilibrium was introduced in van den Nouweland and Wooders (2011) as an extension of ratio equilibrium to local public good economies. In that paper, we took an axiomatic approach to motivate share equilibrium. In the current paper we consider questions related to the existence of share equilibrium and we derive a necessary and sufficient condition for the existence of share equilibrium in symmetric economies. Along the way, we develop a deeper understanding of the possible variation in share equilibrium by considering when symmetric players necessarily have the same share indices in equilibrium.

ABSTRACT. A contestant's effort depends on her knowledge of her rival's type. This knowledge is often limited in real-life contests. I propose a model where the principal of a contest has commitment power to verifiably disclose contestants' types. What is the optimal disclosure policy to stimulate contestants' efforts? Full disclosure spurs more (less) efforts than full concealment if the distribution of types is skewed towards high (low) types. However, the optimal disclosure policy is a particular partial disclosure, regardless of the skewness of the distribution of types; it consists in disclosing the signal which is best for the principal (i.e., all contestants are high types) and concealing the rest.

ABSTRACT. This paper models conflict as a contest within a network of friendships and enmities. We assume that each player is either in a friendly or in an antagonistic relation with every other player and players compete for a fixed prize by exerting costly efforts. We axiomatically characterize a success function which determines the share of each player given the efforts and the network of relations. This framework allows for the study of strategic incentives and friendship formation under conflict as well as the application of stability concepts of network theory to contests.

ABSTRACT. Economic contests describe situations in which a group of individuals incur irretrievable expenditure of costly resources to win valuable rewards. We consider contests with more than one winning positions where any individual contestant can win at most one position. Different mechanisms have been suggested in theory for selecting the final set of winners from a larger set of contestants. However, no applied study is known of to experimentally validate those theoretical predictions so far. We consider three such mechanisms which are best-response equivalent and implement them in the laboratory using a lottery metaphor and a novel feedback. We find the data supporting the mathematical equivalence of the three mechanisms. Lastly, the observed bidding behaviour in the multi-winner contest mechanisms are compared with that from a standard single-winner Tullock contest with comparable equilibrium prediction. We find the simultaneous multi-winner contest success function to be behaviourally robust.

ABSTRACT. We propose a new approach to analyzing signaling games that is particularly useful if the signal space is large. Applications include informed principal problems in which a privately informed party can design any mechanism or can choose a mechanism from a restricted, but large set. Our approach has two main ingredients: sender-payoff reduced form and separation of on and off path analysis. We represent signals (mechanisms, contracts) in terms of correspondences that for each be- lief about the sender’s private information describe the set of sender’s equilibrium payoffs. We introduce a notion of composition to describe on-path equilibrium conditions and a notion of admissibility to describe off-path equilibrium conditions. Our main result is a criterion for a sender’s payoff to be admissible. We show that virtually all known equilibrium constructions in the literature on the informed-principal problem are applications of our criterion.

ABSTRACT. The sender, who has private information about the state of the world, and the receiver, who has to take an action, engage in a multi-stage face-to-face cheap talk. For the case of intermediate sender bias, we construct a new class of equilibria that result in higher ex ante payoffs of the players than the equilibria that are known in the literature. The information is revealed gradually in equilibrium, and at every stage there is a positive chance that the communication ends. We show that the greater the number of communication stages in equilibrium, the higher are the players' payoffs that can be achieved in equilibria of this class. We also characterize the structure of equilibria and the payoffs in the limiting case as the number of communication stages increases without bound.

ABSTRACT. This paper analyzes strategic communication between a sender and a receiver when their conflict of interest is small. In this instance, communication is smooth: the sender chooses close messages for close states of the world. Equilibrium strategies can be approximated by simple differential equations, allowing a neat visualization of how the payoff structure shapes communication. In particular, we characterize how the distribution of states influences the precision of the information transmission and which changes in the (state-dependent) bias improve communication.

ABSTRACT. We provide a model of network-formation where the quality of a link depends on the amount invested in it and is determined by a link-formation "technology", an increasing strictly concave function which is the only exogenous ingredient. The revenue from the investments is the information received through the network. First, assuming that the investments are made by a planner, the basic question is that of the efficient investments, either relative to a given infrastructure or in absolute terms. An "almost" complete characterization is achieved, being instrumental for this a class of core-periphery hybrid star-complete structure of which the star and complete networks are extreme cases. Second, assuming that links are the result of investments of the nodes, there is the question of stability in the network-formation game, either restricted to a given infrastructure or unrestricted. Stable complete and star networks are characterized, and examples of stable hybrid core-periphery networks provided.

ABSTRACT. We study two platforms that compete for users in providing additional functionalities on their networks. The stand alone values are identical for both platforms but the marginal utility from network size depends on functionality. We assume that, at the beginning of the game, both platforms have an exogenously given installed base of users. These old users incur switching costs if choosing another platform. New users have not used either platform and can choose freely. We find that, in equilibrium, platforms do not choose their functionality deterministically but randomize over two disconnected intervals. These correspond to competing for either the entire population or only new users. The value of remaining passive and not improving functionality depends on a platform's installed base. The platform with the smallest base bids more aggressively to compensate for its lower outside option and achieves a monopoly network with higher probability than its competitor.

ABSTRACT. We consider a principal-agent game of network formation with asymmetric information and we address the following questions: (1) Is it possible for the principal to design a mechanism that links the reports of agents about their private information and the set of connections allowed and recommended by the principal so that players truthfully reveal their private information and follow the recommendations specified by the mechanism. (2) Is it possible for the principal to achieve the same outcome by instead choosing a profile of catalogs of allowable ways to connect and then delegating connection choices to each pair of agents. We call this approach to network formation delegated networking and we show, under relatively mild conditions on our game-theoretic model of network formation, that strategic network formation with incomplete information, implemented via a mechanism and centralized reporting is equivalent to implementation via delegated networking with monitoring.

ABSTRACT. Many activities we tend to procrastinate, we prefer to do with others. In this paper, I develop a strategic model of procrastination in which present-biased people faced with an onerous task prefer to do it when someone else does. This turns their decision of when to do the task into a procrastination game -- a dynamic coordination game between present-biased players. The model characterises the conditions under which interaction mitigates or exacerbates procrastination. Interestingly, a procrastinator matched with a worse procrastinator may do her task earlier than she otherwise would, so as to avoid the increased temptation that her peer's company would generate. Procrastinators can thus use bad company as a commitment device to mitigate their self-control problem. Principals can improve people's welfare and reduce inefficient delay by simply matching them with each other, but who is matched with whom matters, and the efficient matching is not always stable.

ABSTRACT. A long tradition of thought identifies two distinct but inter-related dimensions of well being - gratification and flourishing. Relying on insights from relational psychoanalytic theory, we devise a modeling approach based on an analogy with electric circuits, where gratification is represented by electric current and flourishing by electric power. We explore how the latter is influenced non-monotonically by challenges and their perceived values, as represented in the model by loads and resistance. Further psychoanalytic conceptualizations suggest how to extend the electric circuits metaphor for modeling mutual influence among related individuals. This extension gives rise to a definition of a social equilibrium. We prove existence, and show that there might exist 'flourishing traps' - social equilibria dominated by other social equilibria with enhanced flourishing for everybody and with no decrease in gratification. We discuss how 'flourishing traps' differ from classical coordination failures, and the implied nature of guidance for public policy.

ABSTRACT. Limited reasoning about rationality is well understood to lead to important behavioral consequences. The literature has typically viewed such limited reasoning as an artifact of cognitive bounds - loosely, an inability to reason about how other players reason, etc. However, subjects may not be willing to believe their opponent is ``rational, believes rationality, etc…" even if the subject is capable of reasoning in this way. In this paper, we allow for the possibility that a subject's rationality bound may be lower than her cognitive bound: so two subjects with the same cognitive bound may have different rationality bounds. We develop an identification strategy that allows us to investigate the extent to which bounds on rationality are driven by bounds on cognition. We analyze the experimental data from Kneeland (2015). Our results show that rationality bounds are tighter than cognitive bound, especially for subjects whose cognitive bounds are not binding.

ABSTRACT. Centipede games represent a classic example of a strategic situation, where the unique theoretical prediction is at odds with human behavior. This study is explicitly designed to discriminate among the proposed explanations for initial responses in Centipede games. Using multiple and different Centipede games, our approach determines endogenously whether one or more explanations are empirically plausible. We find that non-equilibrium behavior is too heterogenous to be explained by one unique model. However, most non-equilibrium choices can be fully explained by level-k thinking and quantal response equilibrium, roughly in equal proportions. Preference-based models play negligible role in explaining the non-equilibrium play.

ABSTRACT. Minimax and its generalization to mixed strategy Nash equilibrium is the cornerstone of our understanding of strategic situations that require decision makers to be unpredictable. We examine whether the behavior of professional tennis players is consistent with the minimax hypothesis, using a dataset of approximately half a million serves from over 3000 matches. The large number of matches in our dataset necessitates the development of a novel statistical test, which we show is more powerful than the tests used in prior studies. We find that win rates conform remarkably closely to theory for males, but conform far less closely for females. Both males and females exhibit serial correlation in the direction of the serve, but males exhibit less serial correlation than females, and higher-ranked males exhibit less correlation that lower-ranked males.

ABSTRACT. We propose a cooperative solution concept for games in extensive form that incorporates both cooperation and subgame perfection. The new concept, which we label the subgame-perfect core, is a refinement of the core of an extensive game in the same sense as the set of subgame-perfect Nash equilibria is a refinement of the set of Nash equilibria. Also, the subgame-perfect core can be obtained as a subgame-perfect Nash equilibrium of a modified game. We establish several properties of the subgame-perfect core and demonstrate its applicability by studying three applications: the centipede game, the two-player infinite bargaining game of alternating offers, and a dynamic game of climate change. In addition, we motivate and introduce a concept of subgame-perfect strong Nash equilibrium of an extensive game and show that it is coalition proof.

ABSTRACT. In this paper we consider coalition formation in the presence of externalities and restricted cooperation. Previous papers in the literature define various solution concepts for games in partition function form (PFF), see Myerson (1977b), Bloch (1996) and Kóczy (2007,2009) among others. However, up to our knowledge there is no paper about PFF games with restricted cooperation. The case of restricted cooperation is the one when some coalitions cannot be formed. There are numerous examples for restricted cooperation in politics, economic theory and industrial organizations. In this paper we introduce the notion of PFF games with restricted cooperation, and the concept of the recursive core (Kóczy,2007) of PFF games with restricted cooperation, and discuss its properties. We also provide an example for a possible empirical application of our model.

ABSTRACT. In coalition formation games with arbitrary collection of permissible coalitions, a condition on preference profiles is sufficient for the existence of a unique core partition, strengthening uniqueness results both in Pycia (2012) and in Banerjee, Konishi, and Sonmez (2001). An algorithm is presented for finding the unique core partition when preference profiles satisfy that condition. The algorithm terminates after a finite number of iterations bounded by the number of agents.

ABSTRACT. We study information aggregation with a biased election organizer who recruits voters at some cost. Voters are symmetric ex-ante and prefer policy a in state A and policy b in state B, but the organizer prefers policy a regardless of the state. Each recruited voter observes a private signal about the unknown state but does not learn the size of the electorate. In contrast to existing results for large elections, there are equilibria in which information aggregation fails: As the voter recruitment cost vanishes, the organizer can ensure that policy a is implemented independently of the state.

ABSTRACT. We consider a simple judgment aggregation problem where a group of voters wants to make collective judgments over multiple issues. Voters share a preference for true collective judgments but hold private information in the form of types. Departing from the previous literature where a voter’s private information takes a binary value (‘true’ or ‘false’) for each issue, we assume that types are distributed from a state- dependent continuous distribution. We analyse efficient information aggregation and strategic voting in a Bayesian voting game setting. In contrast to the results when types take binary values for each issue, we find that with continuum of types efficient information aggregation is possible only under a strong condition which leads to the continuous analogue of the binary model.

ABSTRACT. The well-known jury paradox -- the more demanding the hurdle for conviction is, the more likely it is that a jury will convict an innocent defendant -- heavily relies on Bayesian updating. However, with ambiguous information (e.g., a forensic test with accuracy of 60%, or more), standard Bayesian updating becomes invalid, challenging the existence of this paradox. By developing a novel theoretical model of voting under ambiguity, this study advances our understanding of how individuals process more realistically imprecise measures of information reliability and how this impacts on information aggregation for the group decision-making. Hence, our findings inform the institutional design of collective deliberation, for small to large group decision-making.

ABSTRACT. In this paper we introduce an axiom, referred to as claims-separable consistency, that is satisfied by several classical division rules defined for claims problems. We determine the entire family of division rules that satisfy this new axiom. We introduce the concept of potential for claims problems and we relate it to claims-separable consistency. In addition, we use claims-separable consistency to characterize axiomatically the minimal overlap rule given by O-Neill (1982).

ABSTRACT. This article addresses linear sharing rules on TU-games with various structures, namely communication structures and conference structures as defined by Myerson in two papers (1977, 1980). Using matrix expressions, we rewrite those sharing rules. That presentation makes it possible to identify the tight relationship between the fairness property and a weak necessary players axiom. Moreover, we show that the latter is implied by the equal treatment of equals, linking the fairness property to the notion of equality.

ABSTRACT. We study the division of limited resources among agents with uncertain needs. The division must be made before uncertainty resolves. Waste is inevitable whenever the allocated amount exceeds the realized demand. An efficient rule minimizes the expected waste induced by an allocation. An egalitarian rule minimizes the gaps of the induced individual welfare. In general, efficiency and fairness are conflicted. We propose an axiom, No Possible Improvement, which should be respected by a planner with both concerns. It asks that no other allocation yields less expected waste than the selected one, and meanwhile achieves smaller gaps of individual welfare. We characterize the family of balanced rules satisfying No Possible Improvement, Recourse Monotonicity and Consistency. The balanced rules accommodate both criteria to some extent, with the two extreme cases being exactly the efficient rule and egalitarian rule. We also characterize the subclass of rules with linear trade-off between efficiency and fairness.

ABSTRACT. We consider auctions where bidders have external incentives and focus on the case where their valuations in the auction are positively correlated with their productivity which matters in a second stage job market. We study how this a¤ects bidding behavior and wages in the job market and proceed to test the model’s implication in an experiment where treatments differ according to which bids are disclosed. Our results broadly con…rm the theoretical prediction that bidders tend to overbid, and their bidding behavior and wages are in‡uenced by the disclosure rule. The data also suggests that the dispersion in worker wages is a¤ected by the disclosure rule, suggesting the importance of reputational bidding.

ABSTRACT. In a Diamond-Dybvig model of financial intermediation, we study how decision making changes if depositors can reveal at a cost to subsequent depositors that they keep their funds in the bank. Theoretically, without signaling, multiple equilibria exist, while under signaling, the unique equilibrium outcome is no bank run. We test the theoretical predictions in a lab experiment and find that indeed when signaling is available, bank runs are less likely to arise though, contrary to our theoretical predictions, signaling is extensively used.

ABSTRACT. To measure the prevalence of social preferences in a competitive environment, we conduct a natural field experiment on German eBay. Acting as a seller, we offer Amazon gift cards with values from 5 to 500 euros. The randomly arriving buyers, unaware of the experiment, make us price offers according to the platform’s rules. Using a novel estimation technique, we extract the underlying distribution of proposed sharing rules from the observed distribution of offers. Once we net out the effect of competition this way, we observe that the prevalence of social preferences in the field is comparable to the laboratory. We obtain three additional results: (i) the amount of money at stake has no significant effect on offers; (ii) the subjects make poor use of public information; (iii) the behaviour of East and West German subjects follows two separate patterns, likely due to the remaining cultural differences.

ABSTRACT. Long term strategic interactions in a stationary environment have usually been modeled as repeated games. At each stage of the process, the moves of the players determine the joint law of the new state and signals to the players and a stage-specific payoff. An evaluation that assigns a (for example, discounted) weight to each stage induces a total weighted payoff, hence a game, with a value that depends on the evaluation.

Longer games, when the duration associated to the evaluation increases, correspond to vanishing stage weight, and the associated limit of the game values is the asymptotic value. We will describe recent advances involving new approaches and results (including cases of existence and non-existence of the limit).

Another alternative approach for studying multistage interactions considers a continuous time process that the players observe and control at discrete times, corresponding to a partition. The asymptotic approach is the analysis of the game as the mesh of the partition decreases, thus with vanishing stage duration. We will present new developments in this direction and discuss the relation with differential games or more generally games in continuous time.

In both frameworks the main tool is the recursive structure and the associate operator that extend the initial Shapley formula for finite discounted stochastic games.