Charles A. Holt

Risk Aversion and Incentive Effects: New Data without Order Effects

Holt and Laury (2002) used a menu of ordered lottery choices to make inferences about risk aversion under various payment conditions. The main results of that paper were: (a) subjects are risk averse, even for relatively small payments of less than

A theoretical analysis of altruism and decision error in public goods games

5; (b) risk aversion increases sharply with large increases in the scale of cash payoffs; and (c) there is no significant effect from increasing the scale of hypothetical payment. With a few exceptions noted in the paper, all treatments began with a low-payment choice, followed by a choice with hypothetical payments that had been scaled up (by 20 , 50 , or 90 ), followed by a real-cash decision with the same high payment scale (20 , 50 , or 90 ), followed by a single, final, low (1 ) real payment choice. Those in the 90 treatment could earn amounts ranging from

Risk Averse Behavior in Generalized Matching Pennies Games

9.00 to

Rent Seeking with Bounded Rationality: An Analysis of the All-Pay Auction

346.50 in this task. As Glenn W. Harrison et al. (2004) correctly note, this design confounds order and treatment effects since the high real payment choice was always completed after the low real and high hypothetical payment tasks. In a new experiment reported below, we first seek to replicate Harrison et al.’s finding that the order effect (participating in a low-payment choice before making a high-payment choice) magnifies the scale effect. In a second treatment, each subject completes the menu of lottery choices under just one payment condition (1 or 20 , real or hypothetical), thereby eliminating any order effects. I. New Data

A model of noisy introspection

Abstract We formalize an equilibrium model in which altruism and decision-error parameters determine the distribution of contributions for linear and quadratic public goods games. The equilibrium density is exponential for linear games, and normal for quadratic games. Our model implies: (i) contributions increase with the marginal value of the public good, (ii) total contributions increase with the number of participants, (iii) mean contributions lie between the Nash prediction and half the endowment. These predictions, which are not implied by a Nash analysis, are consistent with laboratory data. Maximum likelihood estimates of altruism and error parameters are significant and plausible.

An Experimental Test of the Consistent-Conjectures Hypothesis

In experimental studies of behavior in 2×2 games with unique mixed strategy equilibria, observed choice frequencies are systematically different from mixed-strategy Nash predictions. This paper examines experimental results for a variety of such games, and shows that a structural econometric model which incorporates risk aversion into a quantal response equilibrium explains the data very well. Moreover, risk aversion estimates are stable across the different games and are close to those obtained from laboratory and field auction data, as well as from individual lottery choice experiments.

An experimental study of costly coordination

The winner‐take‐all nature of all‐pay auctions makes the outcome sensitive to decision errors, which we introduce with a logit formulation. The equilibrium bid distribution is a fixed point: the belief distributions that determine expected payoffs equal the choice distributions determined by expected payoffs. We prove existence, uniqueness, and symmetry properties. In contrast to the Nash equilibrium, the comparative statics of the logit equilibrium are intuitive: rent dissipation increases with the number of players and the bid cost. Overdissipation of rents is impossible under full rationality but is observed in laboratory experiments. Our model predicts this property.

The Nash equilibrium: A perspective

We present a theoretical model of noisy introspection designed to explain behavior in games played only once. The model determines layers of beliefs about others’ beliefs about ..., etc., but allows for surprises by relaxing the equilibrium requirement that belief distributions coincide with decision distributions. Noise is injected into iterated conjectures about others’ decisions and beliefs, which causes the predictions to differ from those of deterministic models of iterated thinking, e.g., rationalizability. The paper contains a convergence proof that implies existence and uniqueness of the outcome of the iterated thought process. In addition, estimated introspection and noise parameters for data from 37 one-shot matrix games are reported. The accuracy of the model is compared with that of several alternatives.

Quantal Response Equilibrium

This paper compares the theoretical predictions of the consistent-conjectures hypothesis with data for individuals’ behavior in several laboratory experiments. Subjects simultaneously choose either price or quantity in a sequence of market periods, and are given payoff tables that provide complete information about the relationship between decisions and payoffs for all participants. The matching protocols include both fixed pairings for a random number of periods and a deterministic rotation that implements a series of one-shot games. The table has been expanded so that there is no decision that the consistent conjectures equilibrium decision is not dominated, and a constant added to all payoffs ensures a reasonable level of earnings in this proposed equilibrium. This equilibrium concept does not provide good predictions in these experiments. The data are more consistent with the Cournot/Nash prediction, especially in the later rounds of the deterministic rotation protocol. With fixed pairings, some pairs managed to reach tacitly collusive outcomes.

Facilitating Practices: The Effects of Advance Notice and Best-Price Policies

This paper reports data for coordination game experiments with random matching. The experimental design is based on changes in an effort-cost parameter, which do not alter the set of Nash equilibria nor do they alter the predictions of adjustment theories based on imitation or best response dynamics. As expected, however, increasing the effort cost lowers effort levels. Maximization of a stochastic potential function, a concept that generalizes risk dominance to continuous games, predicts this reduction in efforts. An error parameter estimated from initial two-person, minimum-effort games is used to predict behavior in other three-person coordination games.