Risk dominance selects the leader. An experimental analysis

Coordination games arise very often in studies of industrial organization and international trade. This type of games has multiple strict equilibria, and therefore the identification of testable predictions isvery difficult. We study a vertical product differentiation model with two asymmetric players choosing first qualities and then prices. This game has two equilibria for some parameter values. However, we apply the risk dominance criterion suggested by Harsanyi and Selten and show that it always selects the equilibrium where the leader is the firm having some initial advantage. We then perform an experimental analysis totest whether the risk dominance prediction is supported by the behaviour oflaboratory agents. We show that the probability that the risk dominance prediction is right depends crucially on the degree of asymmetry of the game. The stronger the asymmetries the higher the predictive power of the risk dominance criterion.

Pre-play communication and credibility: A test of Aumann's conjecture

The effectiveness of pre-play communication in achieving efficientoutcomes has long been a subject of controversy. In some environments,cheap talk may help to achieve coordination. However, Aumannconjectures that, in a variant of the Stag Hunt game, a signal forefficient play is not self-enforcing and concludes that an "agreementto play [the efficient outcome] conveys no information about what theplayers will do." Harsanyi and Selten (1988) cite this example as anillustration of risk-dominance vs. payoff-dominance. Farrell and Rabin(1996) agree with the logic, but suspect that cheap talk willnonetheless achieve efficiency. The conjecture is tested with one-waycommunication. When the sender first chooses a signal and then anaction, there is impressive coordination: a 94% probability for thepotentially efficient (but risky) play, given a signal for efficientplay. Without communication, efforts to achieve efficiency wereunsuccessful, as the proportion of B moves is only 35%. I also test ahypothesis that the order of the action and the signal affects theresults, finding that the decision order is indeed important. WhileAumann s conjecture is behaviorally disconfirmed when the signal isdetermined initially, the signal s credibility seems to be much moresuspect when the sender is known to have first chosen an action, andthe results are not statistically distinguishable from those whenthere is no signal. Some applications and issues in communication andcoordination are discussed.

Endogenous price leadership

We consider a linear price setting duopoly game with di®erentiatedproducts and determine endogenously which of the players will lead andwhich will follow. While the follower role is most attractive for each firm, we show that waiting is more risky for the low cost firm so that,consequently, risk dominance considerations, as in Harsanyi and Selten(1988), allow the conclusion that only the high cost firm will choose towait. Hence, the low cost firm will emerge as the endogenous price leader.

On some geometry and equivalence classes of normal form games

Equivalence classes of normal form games are defined using the geometryof correspondences of standard equilibiurm concepts like correlated, Nash,and robust equilibrium or risk dominance and rationalizability. Resultingequivalence classes are fully characterized and compared across differentequilibrium concepts for 2 x 2 games. It is argued that the procedure canlead to broad and game-theoretically meaningful distinctions of games aswell as to alternative ways of viewing and testing equilibrium concepts.Larger games are also briefly considered.

Endogenous price leadership

We consider a linear price setting duopoly game with differentiated products and determine endogenously which of the players will lead and which will follow. While the follower role is most attractive for each firm, we show that waiting is more risky for the low cost firm so that, consequently, risk dominance considerations, as in Harsanyi and Selten (1988), allow the conclusion that only the highcost firm will choose to wait. Hence, the low cost firm will emerge as the endogenous price leader.

Endogenous Stackelberg leadership

We consider a linear quantity setting duopoly game and analyzewhich of the players will commit when both players have the possibility todo so. To that end, we study a 2-stage game in which each player caneither commit to a quantity in stage 1 or wait till stage 2. We show thatcommitting is more risky for the high cost firm and that, consequently,risk dominance considerations, as in Harsanyi and Selten (1988), allowthe conclusion that only the low cost firm will choose to commit.Hence, the low cost firm will emerge as the endogenous Stackelberg leader.

Equilibrium selection through incomplete information in coordination games: An experimental study

We perform an experiment on a pure coordination game with uncertaintyabout the payoffs. Our game is closely related to models that have beenused in many macroeconomic and financial applications to solve problemsof equilibrium indeterminacy. In our experiment each subject receives anoisy signal about the true payoffs. This game has a unique strategyprofile that survives the iterative deletion of strictly dominatedstrategies (thus a unique Nash equilibrium). The equilibrium outcomecoincides, on average, with the risk-dominant equilibrium outcome ofthe underlying coordination game. The behavior of the subjects convergesto the theoretical prediction after enough experience has been gained. The data (and the comments) suggest that subjects do not apply through"a priori" reasoning the iterated deletion of dominated strategies.Instead, they adapt to the responses of other players. Thus, the lengthof the learning phase clearly varies for the different signals. We alsotest behavior in a game without uncertainty as a benchmark case. The gamewith uncertainty is inspired by the "global" games of Carlsson and VanDamme (1993).

Stochastic dominance and absolute risk aversion

In this paper we propose the infimum of the Arrow-Pratt index of absolute risk aversion as a measure of global risk aversion of a utility function. We then show that, for any given arbitrary pair of distributions, there exists a threshold level of global risk aversion such that all increasing concave utility functions with at least as much global risk aversion would rank the two distributions in the same way. Furthermore, this threshold level is sharp in the sense that, for any lower level of global risk aversion, we can find two utility functions in this class yielding opposite preference relations for the two distributions.

Stochastic dominance and absolute risk aversion

In this paper we proose the infimum of the Arrow-Pratt index of absoluterisk aversion as a measure of global risk aversion of a utility function.We then show that, for any given arbitrary pair of distributions, thereexists a threshold level of global risk aversion such that all increasingconcave utility functions with at least as much global risk aversion wouldrank the two distributions in the same way. Furthermore, this thresholdlevel is sharp in the sense that, for any lower level of global riskaversion, we can find two utility functions in this class yielding oppositepreference relations for the two distributions.

Emotion and reason in everyday risk perception

Although research has documented the importance of emotion in risk perception, little is knownabout its prevalence in everyday life. Using the Experience Sampling Method, 94 part-timestudents were prompted at random via cellular telephones to report on mood state and threeemotions and to assess risk on thirty occasions during their working hours. The emotions valence, arousal, and dominance were measured using self-assessment manikins (Bradley &Lang, 1994). Hierarchical linear models (HLM) revealed that mood state and emotions explainedsignificant variance in risk perception. In addition, valence and arousal accounted for varianceover and above reason (measured by severity and possibility of risks). Six risks were reassessedin a post-experimental session and found to be lower than their real-time counterparts.The study demonstrates the feasibility and value of collecting representative samples of data withsimple technology. Evidence for the statistical consistency of the HLM estimates is provided inan Appendix.

Subjective income expectations, canonical models and income risk

Expectations are central to behaviour. Despite the existence of subjective expectations data, the standard approach is to ignore these, to hypothecate a model of behaviour and to infer expectations from realisations. In the context of income models, we reveal the informational gain obtained from using both a canonical model and subjective expectations data. We propose a test for this informational gain, and illustrate our approach with an application to the problem of measuring income risk.

Investor protection, risk sharing and inequality

This paper studies the relationship between investor protection, financial risk sharing and income inequality. In the presence of market frictions, better protection makes investors more willing to take on entrepreneurial risk while lending to firms. This implies lower cost of external finance and better risk sharing between financiers and entrepreneurs. Investor protection, by boosting the market for risk sharing plays the twofold role of encouraging agents to undertake risky enterprises and providing them with insurance. By increasing the number of risky projects, it raises income inequality. By extending insurance to more agents, it reduces it. As a result, the relationship between the size of the market for risk sharing and income inequality is hump-shaped. Empirical evidence from a cross-section of sixty-eight countries, and a panel of fifty countries over the period 1976-2000, supports the predictions of the model.

Improving the risk concept: a revision of Arrow-Pratt theory in the context of controlled dynamic stochastic environments

In the literature on risk, one generally assume that uncertainty is uniformly distributed over the entire working horizon, when the absolute risk-aversion index is negative and constant. From this perspective, the risk is totally exogenous, and thus independent of endogenous risks. The classic procedure is "myopic" with regard to potential changes in the future behavior of the agent due to inherent random fluctuations of the system. The agent's attitude to risk is rigid. Although often criticized, the most widely used hypothesis for the analysis of economic behavior is risk-neutrality. This borderline case must be envisaged with prudence in a dynamic stochastic context. The traditional measures of risk-aversion are generally too weak for making comparisons between risky situations, given the dynamic �complexity of the environment. This can be highlighted in concrete problems in finance and insurance, context for which the Arrow-Pratt measures (in the small) give ambiguous.

Optimal feedback control rules sensitive to controlled endogenous risk-aversion

The objective of this paper is to correct and improve the results obtained by Van der Ploeg (1984a, 1984b) and utilized in the theoretical literature related to feedback stochastic optimal control sensitive to constant exogenous risk-aversion (see, Jacobson, 1973, Karp, 1987 and Whittle, 1981, 1989, 1990, among others) or to the classic context of risk-neutral decision-makers (see, Chow, 1973, 1976a, 1976b, 1977, 1978, 1981, 1993). More realistic and attractive, this new approach is placed in the context of a time-varying endogenous risk-aversion which is under the control of the decision-maker. It has strong qualitative implications on the agent's optimal policy during the entire planning horizon.

Testing for poverty dominance: an application to Canada

The paper proposes and applies statistical tests for poverty dominance that check for whether poverty comparisons can be made robustly over ranges of poverty lines and classes of poverty indices. This helps provide both normative and statistical confidence in establishing poverty rankings across distributions. The tests, which can take into account the complex sampling procedures that are typically used by statistical agencies to generate household-level surveys, are implemented using the Canadian Survey of Labour and Income Dynamics (SLID) for 1996, 1999 and 2002. Although the yearly cumulative distribution functions cross at the lower tails of the distributions, the more recent years tend to dominate earlier years for a relatively wide range of poverty lines. Failing to take into account SLID's sampling variability (as is sometimes done) can inflate significantly one's confidence in ranking poverty. Taking into account SLID's complex sampling design (as has not been done before) can also decrease substantially the range of poverty lines over which a poverty ranking can be inferred.