Tacit Coordination

Introducing defect vacancies into nanostructures is a straightforward yet powerful technique employed by scientists to manipulate their properties. In this study, we used a tight-binding model to investigate the eects of an external electric eld and the creation of double vacancies at dierent sites on penta-graphene nanoribbons on the electronic and transport properties. Understanding the eects of external electric elds and vacancy creation on the electronic and transport properties of penta-graphene nanoribbons is important for the advancement of nanoelectronics and the development of innovative applications. After calculating the formation energy for all vacancy structures, it was determined that the maximum stability is achieved when the second vacancy is at the C2 site. By creating vacancies in the structure, in addition to tuning the energy gap, an indirecttodirect bandgap transition can be achieved in penta-graphene nanoribbons. The presence of a direct gap in the electronic properties of penta-graphene nanoribbons has signicant implications. Direct bandgap materials can absorb and emit photons with energies close to the bandgap energy and may be better suited for optical devices. It was also observed that the creation of vacancies in penta-graphene nanoribbons leads to a phase transition from a semiconductor to a metal. Next, the eect of these vacancies on the transport properties of penta-graphene nanoribbons was investigated. The results clearly show that the maximum current and threshold voltage can be controlled by creating vacancies at various sites on the nanoribbon. In general, by creating double vacancies in the structure or applying an external electric eld, an indirect transition to a direct band gap and a semiconductortometal phase transition have been observed. Additionally, when a double vacancy is introduced into the system and an external electric eld is applied simultaneously, at bands are observed in the band structure, as are the tunable band gap, semiconductortometal phase transition, and indirecttodirect bandgap transition. Additionally, to explore the cause of the change in electronic and transport properties with the creation of a vacancy in the structure, the charge density distribution of carbon atoms was analyzed using density functional theory calculations. Due to the dierence in charge density between the penta-graphene nanoribbon sites, signicant charge transfer occurs in the structure after a double vacancy is created in the structure. This charge transfer leads to the generation of an electric current in the nanoribbon. Because of these unique characteristics, penta-graphene nanoribbons are promising candidates for the development of solar cells.

This thesis develops a new theoretically-driven model that predicts the anticipation of upcoming words in a sentence. This model draws from a number of academic disciplines, and it incorporates different frameworks and tools from theoretical linguistics, cognitive science, computational linguistics, and computational models of analogical reasoning.
Using a highly interdisciplinary perspective regarding the nature of linguistic prediction and the kinds of cognitive processes involved therein, I present a set of cognitive desiderata that linguistic theories must
consider: incrementality, nonmonotonicity, and interpretability of sub-propositional content. I differentiate two kinds of contributions when deriving a linguistic prediction: those coming from different levels of semantic granularity and those coming from the coordination of linguistic interaction, and I present a language model that marries these two contributions.

It has long been known that cryptographic schemes o ering provably unbreakable security exist,
namely the One Time Pad (OTP). The OTP, however, comes at the cost of a very long secret key- as long as the plain-text itself. In this paper we propose an encryption scheme which we (boldly)
claim o ers the same level of security as the OTP, while allowing for much shorter keys, of size
polylogarithmic in the computing power available to the adversary. The Scheme requires a large
sequence of truly random words, of length polynomial in the both plain-text size and the logarithm of
the computing power the adversary has. We claim that it ensures such an attacker cannot discern the
cipher output from random data, except with small probability. We also show how it can be adapted
to allow for several plain-texts to be encrypted in the same cipher output, with almost independent
keys. Also, we describe how it can be used in lieu of a One Way Function.

This paper presents a preliminary, numerical what-if analysis of selected geometrical parameters of the tip seal of the last stage of an aircraft engine turbine and the impact of the parameters on the leakage mass ow. The analysis is a part of the task of improving the rotor tip seal con guration in aircraft turbines. Calculations were carried out using the commercial computational uid dynamics code. A straight-through seal with two leaning ns and a honeycomb land was analysed. The computational model was prepared based on some simpli cations introduced to improve the e ciency of the calculations. The rotor entire blade-to-blade channel was analysed, while the seal pitch was reduced to the dimensions of two honeycomb cells. The geometry of the ns themselves was simpli ed too. In the original variant, the n tips had caps to protect them from wear (shape deformation) due to rubbing. In the simpli ed model, the caps were omitted. The simpli cations did not change the leakage mass ow si...

Cette thèse a été réalisée dans le laboratoire Heudiasyc, au sein de l'équipe RO. Son nancement a été assuré par le laboratoire d'excellence Labex MS2T. Je tiens à exprimer mes plus sincères remerciements à mes directeurs de thèse, Pr. Jacques CARLIER et Pr. Aziz MOUKRIM pour la conance qu'ils m'ont accordée pour réaliser ce travail, pour leurs conseils et leur soutient. Je leur suis très reconnaissant de m'avoir permis de poursuivre mes recherches tout en étant toujours présent pour m'aider. Je remercie M. Emmanuel NERON, Professeur à l'Université François Rabelais de Tours, pour avoir présidé mon jury de thèse. Mes remerciements vont également à Eric PINSON Professeur à l'Université Catholique de l'Ouest et Alain QUILLIOT Professeur à l'Université Université Blaise Pascal pour le temps passé à relire cette thèse en tant que rapporteurs et pour leurs remarques pertinentes vis à vis de mon travail. Je remercie également Antoine JOUGLET Maitre de conférence à l'Université de Technologie de Compiègne et Philippe LABORIE ingénieur de recherche chez IBM qui m'ont fait l'honneur d'accepter de participer au jury de ma thèse. Je voudrais également remercier plus particulièrement mes parents et mes soeurs pour leur soutient et leur conance en moi. Le parcours que j'ai fait jusqu'à aujourd'hui n'aurait pas été possible sans eux. Enn, je remercie mes collègues du laboratoire Heudiasyc pour la convivialité ainsi que la bonne ambiance qui régnait au laboratoire. Je remercie particulièrement mon ami Kaci BADER pour son aide et ses conseils.

The all scores matrix of a grid graph is a matrix containing the optimal scores of paths from every vertex on the rst row of the graph to every vertex on the last row. This matrix is commonly used to solve diverse string comparison problems. All scores matrices have the Monge property, and this was exploited by previous works that used all scores matrices for solving various problems. In this paper, we study an extension of grid graphs that contain an additional set of edges, called bridges. Our main result is to show several properties of the all scores matrices of such graphs. We also give an O(r(nm + n 2)) time algorithm for constructing the all scores matrix of an m × n grid graph with r bridges.

Y 2 TiO 5 and Y 2 Ti 2 O 7 are very important materials for nuclear applications and hence, a complete understanding of their radiation response as well as their physical properties is highly desired. Molecular dynamics (MD) is one of the most ecient techniques for such studies, but the reliability of the data obtained from MD studies depends heavily on the robustness of the interatomic potentials used. Here we report the development of novel pair-wise interatomic potentials suitable for molecular dynamics studies of highly energetic processes in Y 2 TiO 5 and Y 2 Ti 2 O 7 crystal systems. An extensive set of rst principle calculations data is generated for a reliable tting of the potential parameters and the generated potentials are used for studying various physical properties. The potentials reproduce melting points of both the systems fairly accurately. The cation antisite pair formation energy is found to be quite less in both the systems, which indicates towards the tendency of pyrochlores to form cation antisite defects upon irradiation. A study of diusion coecients and Zachariasen criteria shows that Y 2 Ti 2 O 7 is more resistant to amorphization than Y 2 TiO 5 .

We examine the eciency and environmental consequences of assigning species-specic common-property rights, considering a Lotka-Volterra model in which sheries are specialized in the harvesting of a single species. We show that the fragmentation of the ecosystem implies the tragedy of the anticommons even when sheries compete for the resource. Indeed, contrasting the private exploitation equilibrium with the socially optimal solution, we demonstrate that the predator stock is too high while the prey stock is too low under private property rights. A puzzling result is that the "abundant" species is actually underused because of insucient economic incentives; however, the scarce and high-priced species does not necessarily suer from overexploitation. Biological interactions are consequently the main driver of stock depletion. Finally, we investigate how to simultaneously solve both the tragedy of the commons and that of the anticommons and analyze the economic costs of regulating only the tragedy of the commons.

Scalable masks for the selection of angular structures in three-dimensional (3D) digital images are considered, which are used in processing with a 3D window sliding over the image and convolved with image fragments. The model of scalable 3D mask was developed based on expanding smaller mask along its sides and edges. In this case, the submatrices remain unchanged, and new elements are added by repeating the elements of the submatrix, preserving the structure of the corner. This approach helps to design the hierarchical computations of 3D data.

This paper studies how welfare outcomes in centralized school choice depend on the assignment mechanism when participants are not fully informed. Using a survey of school choice participants in a strategic setting, we show that beliefs about admissions chances differ from rational expectations values and predict choice behavior. To quantify the welfare costs of belief errors, we estimate a model of school choice that incorporates subjective beliefs. We evaluate the equilibrium effects of switching to a strategy-proof deferred acceptance algorithm, and of improving households' belief accuracy. We find that a switch to truthful reporting in the DA mechanism offers welfare improvements over the baseline given the belief errors we observe in the data, but that an analyst who assumed families had accurate beliefs would have reached the opposite conclusion. Adam Kapor Department of Economics Princeton University 280 Julis Romo Rabinowitz Building Princeton, NJ 08544 and NBER akapor@pr...

We present a new model for the estimation of positional uncertainty of petroleum wells. The model uses a heavy tailed normal inverse Gaussian distribution for the errors in Earth's magnetic field reference values. These references are required for calculating the position of a well using magnetic directional surveying. The results show that the normal inverse Gaussian distribution gives a more realistic fit to the Earth's magnetic field reference values than the normal distribution, which is the current state of art. Errors in surveying sensors and reference errors are propagated along the well path to obtain the positioning uncertainty. The positional uncertainty is important for the probability of collisions between wells and for drilling accurate relief wells when an eventual collision has resulted in a blow-out situation. We compare results of the normal model used in the petroleum industry with our proposed model. The comparison is made for anti-collision calculations between drilled wells and a planned well. The results indicate a higher risk of collisions for certain well geometries when using the normal distribution, while a collision is avoided in the normal inverse Gaussian situation. We find clear differences between an approximate normal test and a Monte Carlo test.

A series of experiments was designed and implemented to investigate cross-cultural differences in preferences for contributing to local public goods. The research investigates differences between contributions made by participants from the United States, Russia and Kazakhstan. In these experiments each participant has three options: keep money for herself, contribute to a public good that benefits a small group (the local good), or contribute to a public good that benefits the entire group (the global good). The researchers find significant differences in contribution patterns across the three cultures, and find that all participants contribute significantly more to the small group public good than to the large group public good.

Many people have I met, or that have followed me throughout this journey, that contributed to this work in one way or another. First and foremost I want to express my deep gratitude to my advisor Morten S. Riishuus. His guidance, insightful conversations, enthusiasm, and above all, friendship, was exceptional in all ways. Besides this, he will be remembered for his excellent outdoor cooking skills during field work. I'm also greatly thankful to Þorvaldur Þórðarson, for his mentorship throughout the years and inspiring enthusiasm about volcanology, which has resulted in many fruitful conversations. Christian Tegner is also thanked for his interest and support to this project.

Patterns of animal movement data are analysed to understand animal behaviour. Movement patterns on the population level can be described with respect to environmental conditions, albeit we are not able to understand the ultimate motivation of animal movement. New conceptual frameworks in the eld of movement ecology seem to be promising to support in understanding the behavioral process of movement. In particular, individual-based movement models (IBMM) that focus on the individual and models its internal state have been capable of linking patterns on the population level to behavioral processes on the individual level. In this study, we set up a discrete non-linear IBMM that combines the framework of the traditional method of resource selection functions and IBMM. GPS-based position data of roe deer (Capreolus capreolus) and red deer (Cervus elaphus) recorded from 2002 to 2012 in the Bavarian Forest National Park, Germany, were used to parameterize the model with generalized mixed eect models. Forwardsimulation of the parameterized model were used to validate the model, by comparing observed with simulated patterns on the population, such as home range size and distribution of step length. We further followed a pattern-oriented approach by testing whether diurnal variation of attraction to and repulsion from environmental factors like buildings, roads, home range centre, food and cover availability are analogous for empirical and simulated data. The forward simulation provided realistic pathways with respect to the patterns under study. Our approach illustrates a feasible combination of forwardand reverse-modelling.

Cette thèse explore quelques facettes du rôle des infrastructures de transport urbain dans la fluidification des frictions du marché du travail causées par la dissociation spatiale entre les lieux de résidence et les opportunités d'emploi. Pour ce faire, elle est construite en trois chapitres.Dans le premier, nous construisons une série de fonctions d'appariement spatiales entre les travailleurs au chômage et les postes vacants qui tiennent compte de la dimension spatiale de la recherche d'emploi, qui dépendent des paramètres de l'infrastructure de transport et qui ne sont pas conditionnées par la structure de la ville dans laquelle l'appariement se déroule.Dans le second chapitre nous présentons un outil d'évaluation de politiques publiques novateur, utilisé ici pour mesurer l'impact de la performance de l'infrastructure et des services de transports publics franciliens sur les taux de chômage locaux et pour prédire l'impact du Grand Paris Expres...

s'adapte en continu lorsque les noeuds se déplacent. L'ecacité des algorithmes ainsi que leur capacité de passage à l'échelle ont été évaluées au moyen de simulations dans des réseaux de tailles et de congurations diverses. Finalement, nous présentons Adcast, le logiciel où sont implémentés nos algorithmes. Il s'exécute sur des PDAs fonctionnant sous Windows Mobile et des ordinateurs sous Windows. Adcast a rendu possible des tests dans des environnements réseaux ad hoc réels et simulés.

This paper presents an approach for modeling and associating A-Policies to services' based applications. It proposes to extend the SOD-M model driven method with (i) the π-SCM an A-Policy services' composition meta-model for representing non-functional constraints associated to services' based applications; (ii) the π-Pews meta-model providing guidelines for expressing the composition and the policies; and, (iii) model to model and model to text transformation rules for semi-automatizing the implementation of reliable services' compositions.

This thesis consists of two separate studies of models of public good production. The rst study is a baseline cross-cultural replication of the Group Based Meritocracy Mechanism (GBM) by Gunnthorsdottir et al. (2010a). Experimental data collected using Icelandic subjects is compared to data previously collected using US subjects. The results show that despite cultural di erences, subjects from both countries respond similarly to the incentive structure of the GBM and possess the same ability to coordinate its near-e cient equilibrium (NEE). Without communicating and within very few trials, subjects from both countries are successfully able to coordinate this complex asymmetric Nash equilibrium. In the second study a new model, based on the GBM, is proposed where players are segregated into two or more endowment classes and each endowment class has the possibility to form uniform collaborative groups. This extended mechanism and the implications of its novelties are analyzed in detai...

We test, via cross-cultural comparison, the robustness of complex tacit coordination in an endogenous-grouping mechanism. The mechanism requires coordination along two dimensions: 1) equilibrium selection and 2) the tacit coordination of an asymmetric equilibrium -a somewhat puzzling phenomenon commonly referred to as "Magic" (Kahneman, 1988, p. 12). In 2010, Gunnthorsdottir et al. introduced and experimentally tested a dual-level extension of a standard Social Dilemma game, which they call "Group-based Meritocracy Mechanism" (GBM). The GBM incorporates group formation, which a standard Social Dilemma Game does not address. The GBM is, to our knowledge, the first formal model of contribution-based team formation. Formally incorporating group formation into a team game increases a model's external validity since under naturally occurring circumstances, most groups are consistently on the lookout for high-contributing members. Incorporating this feature into a team game however, complicates its equilibrium structure: A standard Social Dilemma's sole equilibrium of non-contribution by all still persists in the weaker form of a best-response equilibrium. There is now however also a second, very efficient equilibrium, which is payoff dominant, asymmetric, and not particularly obvious without a formal analysis. For this equilibrium to occur, experimental subjects must first select it over its alternative. If indeed selected, subjects must coordinate its asymmetric profile, via simultaneous decisions and without communication. Kahneman (1988, p. 12) coined the term "Magic" for when experimental subjects coordinate an asymmetric equilibrium via simultaneous decisions, without communication. While Kahneman observed that the Nash Equilibrium is an excellent, albeit surprising, predictor of aggregate behavior in such a setting, he also noted that individual choices over trials were unpredictable. This fact adds to the puzzle of how such a high degree of aggregate coordination can be achieved at all. Kahneman initially observed the phenomenon in a Market Entry Game (MEG; Cary-Bobo, 1990;, a simple game with a binary strategy space; its asymmetric equilibrium is both obvious and unique. Extensive experimental tests of "Magic" in MEGs show that it is a very robust phenomenon (for overviews, see, e.g., Camerer, 2003, Ch. 7.3;. In experimental tests of the GBM, Gunnthorsdottir et al. (2010a; 2010b) report more striking "Magic" than previously reported: 1) since the GBM has two pure strategy equilibria, subjects first of all need to tacitly coordinate which equilibrium to select -the payoff dominant asymmetric one, or the payoff-inferior one of non-contribution by all. 2) The GBM's asymmetric equilibrium is not easy to detect. 3) While the MEGs strategy space is binary, in Gunnthorsdottir et al.'s version of the GBM and without loss of generality, the strategy space contains 101 choices, of which only two are part of the asymmetric payoff-dominant equilibrium. subjects, US university students, reliably selected the payoff dominant equilibrium and tacitly coordinated its asymmetric strategy profile. The current series of experiments test the robustness of these findings with additional tests in Iceland, a society with a distinctly different culture. Our results show that both the aggregate and individual behavioral patterns in both locations are very similar. While it is not yet clear how this phenomenon occurs, the findings here indicate that very efficient tacit coordination in contribution-based group formation is a robust occurrence.

In a laboratory experiment we compare voluntary cooperation in Iceland and the US (N=156 in a convenience sample of university students). We furthermore compare the associated thought processes across cultures. The two countries have similar economic performance, but survey measures show that they differ culturally. Our hypotheses are based on two such measures, Inglehart’s cultural world map and Knack & Keefer’s scale of civic attitudes toward large-scale societal functioning. We prime the participants with different social foci, emphasizing in one a narrow grouping and in the other a larger social unit. In each country we implement this using two different feedback treatments. Under group feedback, participants only know the contributions by the four members of their directly cooperating group. Under session feedback they are informed of the contributions within their group as well as by everyone else in the session. Under group feedback, cooperation levels do not differ between t...

We compare coordination success of individuals and teams in the minimum effort coordination game. The game is played by groups of either five individuals or five two-person teams with either fixed or random re-matching protocols. When groups are fixed, teams perform at least as well as individuals, if not better, in terms of coordinating to the payoff dominant outcome. But with random re-matching, teams experience pervasive coordination failures. A public recommendation to a strategy or a performance bonus exhorting players to coordinate to the payoff-dominant equilibrium has similar impact on coordination for both individuals and teams playing with fixed matching. However, coordination is far more difficult to achieve with teams playing under random re-matching. Our results have implications for the design of work-groups in organizations.

Souvent apparenté à la théorie de l’acteur rationnel, à la théorie des jeux et au contexte de la guerre froide, Thomas Schelling n’en a pas moins produit une œuvre iconoclaste et occupé une position originale dans le champ scientifique états-unien, ne serait-ce que par son penchant affirmé pour l’interdisciplinarité. Soucieux de scruter les « impératifs de l’action », ses travaux permettent d’ouvrir des pistes de recherche stimulantes pour peu qu’on les croise avec ceux d’autres auteurs et types de démarche. Son concept de point focal peut par exemple renouveler l’approche du charisme. Enfin, sa théorie de la décision interdépendante et son insistance sur l’importance de la coordination tacite mettent en cause les conceptions exclusives (ascendante ou descendante) de la causalité, nous incitant à penser en termes de causalité spirale et de jeux spéculaires, ces jeux de miroirs dans lesquels sont pris les acteurs et où s’originent les processus d’auto-amplification.

A new mechanism that substantially mitigates social dilemmas is examined theoretically and experimentally. It resembles the voluntary contribution mechanism (VCM) except that in each decision round subjects are ranked and then grouped according to their public contribution. The game, particularly in a discrete implementation, has multiple mostly asymmetric, Pareto-ranked pure-strategy equilibria which are rather counterintuitive, yet experimental subjects tacitly coordinate the payoff-dominant equilibrium reliably and quite precisely. In the VCM grouping is random which, with its arbitrary relation to contribution corresponds to any grouping unrelated to output, for example grouping based on race or gender. The new mechanism resembles a meritocracy since based on how much they contribute participants are assigned to strata that vary in payoff. The findings shed light on the nature of merit-based social and organizational grouping and provide guidelines for future research and application.

This thesis consists of two separate studies of models of public good production. The first study is a baseline cross-cultural replication of the Group Based Meritocracy Mechanism (GBM) by Gunnthorsdottir et.al. (2010). Experimental data collected using Icelandic subjects is compared to data previously collected using US subjects. The results show that despite cultural differences, subjects from both countries respond similarly to the incentive structure of the GBM and possess the same ability to coordinate its near-efficient equilibrium (NEE). Without communicating and within very few trials, subjects from both countries are successfully able to coordinate this complex asymmetric Nash equilibrium.
In the second study a new model, based on the GBM, is proposed where players are segregated into two or more endowment classes and each endowment class has the possibility to form uniform collaborative groups. This extended mechanism and the implications of its novelties are analyzed in detail and an overall Nash equilibrium solution, called the multi-tier equilibrium (MTE), is proposed. The MTE describes possible combinations and types of social strata that players form depending on the society class structure. An experimental test of the extended mechanism reveals that even with a stratified society, subjects in a GBM social stratum continue to coordinate an NEE. However, subjects in a top tier VCM stratum collaborate considerably more than what is typically observed in VCM experiments.

this paper we experimentally examine the effect that keeping free-riders fromcooperators has on the latters" public contributions. We use the "Goodwill Accounting" modelof self-interested reciprocity, first advanced by McCabe and Smith (1999) and reviewed in thenext section, to organize our analysis. One important reason is that the Goodwill Accountingtheory provides a simple and testable procedure to classify each subject as

The personality - situation debate in psychology repeats itself in behavioral economics as the rational actor (equilibrium) versus type debate. This paper examines under which condition psychology overrides rational behavior and vice versa. If experimental subjects have insufficient information about the game they project their personality into the situation.

We test, via cross-cultural comparison, the robustness of complex tacit coordination in an endogenous-grouping mechanism. The mechanism requires coordination along two dimensions: 1) equilibrium selection and 2) the tacit coordination of an asymmetric equilibrium – a somewhat puzzling phenomenon commonly referred to as “Magic”.

Anna Gunnthorsdottir, Australian Graduate School of Management Roumen Vragov, City University of NY Kevin McCabe, George Mason University Stefan Seifert, Technical University Karlsruhe Keywords: social dilemmas, Nash equilibrium, non-cooperative games, coordination, mechanism design, experiment.

It is well known that property rights can privatize incentives for the provision of goods that provide public benefits. Unfortunately, it is often infeasible to implement a property rights system which truly privatizes incentives by making benefits proportional to investments. 1 Nevertheless, there is substantial evidence, including Ronald Coase's (1974) historical research on Great Britain's lighthouse system, that suggests most public goods can be provided privately even without such a system. 2 So how are public goods provided?

According to theory a pure meritocracy is efficient because individual members are competitively rewarded according to their individual contributions to society. However, purely individually based meritocracies seldom occur. We introduce a new model of social production called “team-based meritocracy”(TBM) in which individual members are rewarded based on their team membership.

We examine theoretically and experimentally how competitive contribution-based group formation affects incentives to free-ride. We introduce a new formal model of social production, called a “Group-based Meritocracy Mechanism”(GBM), which extends the single-group-level analysis of a Voluntary Contribution Mechanism (VCM) to multiple groups. In a GBM individuals are ranked according to their group contributions. Based on this ranking, participants are then partitioned into equal-sized groups.

Abstract: We examine theoretically and experimentally how a society's grouping and stratification rules affect incentives and efficiency, and compare meritocratic and ascriptive grouping. We present a multi-level model of social production, and extend the usual single-group-level analysis of cooperation to a broadly defined system, in which individuals compete for inclusion in stratified groups based on the contributions they make. Group members share their collective output equally amongst themselves.

A new mechanism that substantially mitigates social dilemmas is examined theoretically and experimentally. It resembles the voluntary contribution mechanism (VCM) except that in each decision round subjects are ranked and then grouped according to their public contribution.

Novel voluntary contribution mechanism experiments are used to investigate how individuals' experience (history) and cooperative disposition and interact. We find that a subject's initial public contribution is a useful measure of cooperative disposition. History effects are isolated by comparing treatments with random group assignment to treatments where subjects are grouped based on their contributions. The latter treatment increases the frequency with which cooperative subjects interact.

It is well known that property rights can privatize incentives for the provision of goods that provide public benefits. Unfortunately, it is often infeasible to implement a property rights system which truly privatizes incentives by making benefits proportional to investments. 1 Nevertheless, there is substantial evidence, including Ronald Coase's (1974) historical research on Great Britain's lighthouse system, that suggests most public goods can be provided privately even without such a system. 2 So how are public goods provided?

Abstract: We design a novel sequential public goods experiment to study reciprocity, or conditional cooperation. In contrast to the standard simultaneous contribution game, our sequential design provides direct evidence on how subjects condition their own contributions on the contributions of other subjects in the experiment. We develop a simple but useful behavioral-type classification procedure and use it to analyze the data from this design.

We introduce a formal account of the efficiency-enhancing effects of contribution-based grouping. We find that a somewhat complicated equilibrium, rather than off-equilibrium psychological motives, account for the effect in our case. Experimental subjects are able to coordinate this asymmetric equilibrium via simultaneous decisions, without communicating, and apparently without learning.

We focus on a class of market entry games in which a newly emergent market opportunity may be fruitfully exploited by no more than a commonly known, exogenously determined number of firms. Our results show significant effects of the parameters manipulated in the study, namely, the market capacity, entry fee, and method of subject assignment to groups (fixed vs. random). In contrast to previous market entry games with linear payoff functions, we find no evidence of convergence to equilibrium play on the aggregate level. Shifting the focus of the analysis from the aggregate to the individual level, four clusters of subjects are identified. The patterns are: (1) choice of the same action that is independent of the parameters of the game or the outcome of previous presentations of the same game; (2) random choices with probabilities prescribed by the equilibrium solution for risk-neutral players;

We test, via cross-cultural comparison, the robustness of complex tacit coordination in an endogenous-grouping mechanism. The mechanism requires coordination along two dimensions: 1) equilibrium selection and 2) the tacit coordination of an asymmetric equilibrium -a somewhat puzzling phenomenon commonly referred to as "Magic" (Kahneman, 1988, p. 12).

"This work received an Emerald Literati Best Author Award 2011.

Purpose and approach –
We examine theoretically and experimentally how unequal abilities to contribute affect incentives and efficiency when players compete for membership in stratified groups based on the contributions they make. Players have either a low or a high endowment. Once assigned to a group based on their group contribution, players share equally in their group’s collective output. Depending on the parameters, the mechanism has several distinct equilibria that differ in efficiency.
Findings –
Somewhat counter to conventional expectation our theoretical analysis indicates that as long as certain assumptions are satisfied, efficiency increases rather than decreases the more abilities to contribute differ. The analysis also suggests various follow-up experiments about equilibrium selection, tacit coordination, and the effect of unequal abilities in systems with endogenous grouping. We conduct an experiment that shows that subjects tacitly coordinate the mechanism’s asymmetric payoff-dominant equilibrium with precision; this precision is robust to a change in the structure and complexity of the game.
Implications –
The results suggest that people respond to merit-based grouping in a natural way and that competitive contribution-based grouping encourages public contributions even when abilities to contribute differ, which is the case in all communities and societies."