Similarity-based mistakes in choice

We characterize the following choice procedure. The decision maker is endowed with two binary relations over alternatives, a preference and a similarity. In every choice problem she includes in her choice set all alternatives which are similar to the best feasible alternative. Hence she can, by mistake, choose an inferior option because it is similar to the best. We characterize this boundedly rational behavior by suitably weakening the rationalizability axiom of Arrow (1959). We also characterize a variation where the decision maker chooses alternatives on the basis of their similarities to attractive yet infeasible options. We show that similarity-based mistakes of either kind lead to cyclical behavior. Finally, we reinterpret our procedure as a method for choosing a bundle given a set of individual items, in which the decision maker combines the best feasible item with those that complement it.     

Timing of information in agency problems with hidden actions

This paper endogenizes in a standard hidden action model the point in time when a risk neutral and wealth constrained agent and the principal observe the realization of an additional signal: before the agent’s effort choice (ex ante information) or after (ex post information). In a decision problem, ex ante information does (weakly) better than ex post information because the decision maker can tailor efforts to the information. We show that this is not the case for incentive problems: a negative incentive effect arises under ex ante information that prevails even though the principal tailors the agent’s effort to the information.     

A model of stochastic choice from lists

We study stochastic choice from lists. All lists present the same set of alternatives albeit in different orders. Faced with a list, the decision maker makes her choice in two stages. In the first stage she searches through the list till she sees $k$ alternatives. In the second stage she chooses from the alternatives she has seen. Both $k$ and the choice rule governing her second stage behavior are random. We show that the underlying primitives of our model are revealed by the decision maker’s choice frequencies from lists. We characterize the model and two of its special cases. In the first special case the decision maker deterministically chooses the best observed alternative according to a given preference. In the second, the decision maker maximizes random preferences.     

Minimax regret treatment choice with covariates or with limited validity of experiments

This paper continues the investigation of minimax regret treatment choice initiated by Manski (2004). Consider a decision maker who must assign treatment to future subjects after observing outcomes experienced in a sample. A certain scoring rule is known to achieve minimax regret in simple versions of this decision problem. I investigate its sensitivity to perturbations of the decision environment in realistic directions. They are as follows. (i) Treatment outcomes may be influenced by a covariate whose effect on outcome distributions is bounded (in one of numerous probability metrics). This is interesting because introduction of a covariate with unrestricted effects leads to a pathological result. (ii) The experiment may have limited validity because of selective noncompliance or because the sampling universe is a potentially selective subset of the treatment population. Thus, even large samples may generate misleading signals. These problems are formalized via a “bounds” approach that turns the problem into one of partial identification.In both scenarios, small but positive perturbations leave the minimax regret decision rule unchanged. Thus, minimax regret analysis is not knife-edge-dependent on ignoring certain aspects of realistic decision problems. Indeed, it recommends to entirely disregard covariates whose effect is believed to be positive but small, as well as small enough amounts of missing data or selective attrition. All findings are finite sample results derived by game theoretic analysis.     

冒风险的决策者：聪明乎？糊涂乎？

决策是对行为的选择——选择做什么或者不做什么,是人类的高级认知活动之一。风险决策特指决策者在对未来情况不完全确定、但是确知各种决策后果以及各种后果出现概率情况下的决策。人们如何进行风险决策,一直是经济学和心理学非常关心的未解之谜。本文尝试从有限理性的角度,分析“齐当别”抉择模型是如何理解、解释人们的风险决策过程,并讨论了“最终的决策只能在单一维度上进行”这一决策规则在现实生活中的旨趣和寓意。     

An approach to the display and analysis of multiobjective problems

In all operations research applications, the problem of implementation rests on the information conveyed to the decision maker. The presentation of results is a critical link in the success of a project. An ineffective transfer of information will reduce the chance of decision maker acceptance. This final step in the analysis is particularly difficult in multiobjective analyses, where the amount of relevent information increases with each performance measure.This paper will describe an alternative to the typical graphical approach to multiobjective display, which is adaptable to any number of objectives. A real world example is given and some theoretical insights are developed.     

Randomly evolving tastes and delayed commitment

We consider a decision maker with randomly evolving tastes who faces dynamic decision situations that involve intertemporal tradeoffs, such as those in consumption savings problems. We axiomatize a recursive representation of choice that features uncertain consumption utilities, which evolve according to a subjective Markov process. The parameters of the representation, which are the subjective Markov process governing the evolution of utilities, and the discount factor, are uniquely identified from behavior. We relate the correlation of tastes over time and the desire to delay commitment to future consumption.     

基于二层规划的协同采购策略模型研究

在对协同采购理论阐述的基础上,提出了一种基于二层规划的协同采购决策模型;以企业为上层决策者,主要供应商为下层决策者,构建了上下层间通过合作契约而共享资源的二层规划模型;采用遗传算法求解二层规划问题,最后给出一个算例。     

Revenue forecasting, budget setting and risk

The budget-making process can be viewed as a problem of decision making under uncertainty since revenues are unknown at the time the budget is written. Revenue forecasts become information for describing the uncertainty in revenue receipts, thereby allowing the decision maker to trade off desire for a larger budget against the uncertainty in revenues. A chance constraint decision model is used to model this process. Application of the model to evaluate alternative forecasting techniques is then demonstrated for several revenue sources used in Kansas City, Missouri.     

The effects of ambiguity on entrepreneurship

We incorporate ambiguity (Knightian uncertainty) into a classic model of entrepreneurship to analyze, among other things, its effects on the optimal level of business startups, the relation between total assets and the size of the entrepreneurial investment, the effects of increasing ambiguity on developing new ventures, and the decision to self‐select into entrepreneurship for an indifferent decision maker. We first show that, under the monotone‐likelihood ratio property, the introduction of ambiguity negatively affects the optimal entrepreneurial investment, something that is consistent with most experimental evidence about entrepreneurial choice under ambiguity. Then, we show that the classical explanations for the positive correlation between total assets and business startups based on decreasing absolute risk aversion preferences and prudent behavior can be challenged when ambiguity is incorporated into the analysis, and we provide the conditions that guarantee that the traditional comparative static result under risk is replicated under ambiguity. We also show that increases in ambiguity aversion reduce entrepreneurial activities. Finally, we discuss our results under alternative ways of modeling ambiguity.     

基于投影模型的不确定语言型多属性决策方法及应用

研究了不确定语言型多属性决策评价结果与决策者对方案的偏好信息之间存在偏差的问题,提出了基于投影模型的不确定多属性决策方法。该方法通过建立与区间型语言标度对应的术语指标矩阵,及方案综合属性值与决策者主观偏好值之间的投影模型,确定属性的权重,然后运用加权法得到方案的综合属性值,利用已有的可能度矩阵排序公式得到决策方案的排序。最后,通过算例对该方法的实用性和有效性进行了说明。     

Robustness as a corporate objective under uncertainty

This paper examines the nature of the objective function of the firm when operating under conditions of uncertainty. Robustness is presented as a purposeful maximand for decision making both under conditions of certainty and uncertainty - a robust decision being one in which the decision maker retains the maximum flexibility with regard to future decisions after an initial decision has been made. Its incorporation within a managerial objective function provides a measurable scale for making choices between alternative courses of action, including under conditions of internal organizational conflict and environmental reaction by other decision makers.     

Decision making with imprecise probabilistic information

We develop an axiomatic approach to decision under uncertainty that explicitly takes into account the information available to the decision maker. The information is described by a set of priors and a reference prior. We define a notion of imprecision for this informational setting and show that a decision maker who is averse to information imprecision maximizes the minimum expected utility computed with respect to a subset of the set of initially given priors. The extent to which this set is reduced can be seen as a measure of imprecision aversion. This approach thus allows a lot of flexibility in modelling the decision maker attitude towards imprecision. In contrast, applying Gilboa and Schmeidler [J. Math. Econ. 18 (1989) 141] maxmin criterion to the initial set of priors amounts to assuming extreme pessimism.     

Learning about learning in games through experimental control of strategic interdependence

We report results from an experiment in which humans repeatedly play one of two games against a computer program that follows either a reinforcement or an experience weighted attraction learning algorithm. Our experiment shows these learning algorithms detect exploitable opportunities more sensitively than humans. Also, learning algorithms respond to detected payoff-increasing opportunities systematically; however, the responses are too weak to improve the algorithms' payoffs. Human play against various decision maker types does not vary significantly. These factors lead to a strong linear relationship between the humans' and algorithms' action choice proportions that is suggestive of the algorithms' best response correspondences.     

Stopping with anticipated regret

This paper analyzes a stopping problem where the decision maker is driven by anticipated ex-post regret. There are two sources of potential dynamic inconsistency, one is arrival of information and the other is changing choice opportunities over time—discarding the current stopping option may change how she stops the game in the future.First we consider a naive planner who prescribes a commitment solution, and illustrate the nature of the inconsistency problem. Then we consider a sophisticated planner who plays backward induction against her ‘successive selves’. The resolution of dynamic inconsistency does not in general allow the use of standard dynamic programming technique. We provide, however, a simple characterization of the backward induction strategy, which is given in a recursive formula.We also provide a behavioral implication, that larger indeterminacy of belief may lead to a more aggressive behavior, that is, continuing the gamble longer, which contrasts to the implication of ambiguity aversion.     

Biased recommendations from biased and unbiased experts

When can you trust an expert to provide honest advice? We develop and test a recommendation game where an expert helps a decision maker choose among two actions that benefit the expert and an outside option that does not. For instance, a salesperson recommends one of two products to a customer who may instead purchase nothing. Subject behavior in a laboratory experiment is largely consistent with predictions from the cheap talk literature. For sufficient symmetry in payoffs, recommendations are persuasive in that they raise the chance that the decision maker takes one of the actions rather than the outside option. If the expert is known to have a payoff bias toward an action, such as a salesperson receiving a higher commission on one product, the decision maker partially discounts a recommendation for it and is more likely to take the outside option. If the bias is uncertain, then biased experts lie even more, whereas unbiased experts follow a political correctness strategy of pushing the opposite action so as to be more persuasive. Even when the expert is known to be unbiased, if the decision maker already favors an action the expert panders toward it, and the decision maker partially discounts the recommendation. The comparative static predictions hold with any degree of lying aversion up to pure cheap talk, and most subjects exhibit some limited lying aversion. The results highlight that the transparency of expert incentives can improve communication, but need not ensure unbiased advice.     

A theory of stochastic choice under uncertainty

In this paper we propose a characterization of stochastic choice under risk and under uncertainty. We presume that decision makers’ actual choices are governed by randomly selected states of mind, and study the representation of decision makers’ perceptions of the stochastic process underlying the selection of their state of mind. The connections of this work to the literatures on random choice, choice behavior when preference are incomplete; choice of menus; and grades of indecisiveness are also discussed.     

A social procedurial approach to the Pareto optimization problematique

This article applies a variant of game theory to the Pareto multi-value problematique, that is situations where members of a group, community or society are faced with alternative allocations, institutional arrangements, or states of the world and may collectively choose an allocation, institutional arrangement or state of the world if they can agree on it. This type of multiple value decision situation is increasingly prevalent not only on the level of societal and political issues but on the level of many enterprises, particularly those advocating corporate social responsibility. Because actors hold and apply values from different perspectives, there are potential contradictory value judgments and incompatible equilibria. In a world of contradiction, incommensurability, and disequilibrium, to what extent can conflicts be resolved and social equilibrium accomplished? Force works but it is inherently unstable. Drawing on an extension of classical game theory, generalized game theory (GGT), this article addresses the multi-value problematique in terms of collective “resolution procedures.” These regulative procedures—or social algorithms—are applied to problems of conflict and suboptimality in a multiple value world such as Pareto envisioned. This paper (the first of two) outlines key elements of GGT, defines the Pareto multi-value problematique, pointing out several of the critical weaknesses, theoretical as well as empirical, of the Pareto approach. GGT is then applied in defining and analyzing several major procedures to realize improvements in a multi-value world characterized by conflict and sub-optimality. A second article conceptualizes a complex of societal games making up a social system with 2-phase multi-level game processes; it applies the conceptualization to the different societal procedures for multi-value choice under conditions of conflict. Procedures such as democratic voting, adjudication and administrative decision-making, and multi-lateral negotiation are capable of producing outcomes that in many cases are widely accepted as legitimate and become social equilibria (at least within some range of conditions). These procedures and the conditions for their activation and implementation are modelled and explicated through a generalized game approach.     

The Paternalistic Bias of Expert Advice

Using a theoretical model of noisy expert advice, I show that language inflation can be a rational response to the vagueness of language. Experts will tend to overstate their positions to a like‐minded decision maker (DM) and this constitutes a Pareto improvement over sending a sincere message. When the message space is bounded, overstatement may interfere with the DM's ability to aggregate the experts' information, because communication is less precise when the same message is sent for many states of the world. Despite this, I show that experts are willing to send either the most extreme message to the DM, or a partially overstated message, because by doing so the expert can decrease the likelihood the DM makes a suboptimal decision due to his subjective interpretation of the advice. Because the expert inflates his message toward the policy he believes the DM would be better off choosing, rather than sincerely revealing his information, I refer to this behavior as a paternalistic bias.