Heuristics for Resource Allocation

资源分配中的启发式资源分配问题出现在政府和其他机构的决策中。例如，将医学研究资金分配给疾病，将减少风险的资金分配给风险来源，将开发资金分配给公司内部的项目，以及将大学资金分配给院系。一些分配因效率低下而受到批评，例如降低风险，有人认为太多的钱花在了降低一些风险上，而花在另一些风险上的钱太少。分配的部分原因是人类对如何分配资源的判断。判决可能由负责拨款的官员、受影响的人或其他团体做出。这项研究探索了人们在做出这样的判断时所使用的推理或启发式原理。研究对象对假想案例和真实案例做出判断。这些判断揭示了受试者的启发式。我们将这些判断的结果与相同案例的成本效益分析所确定的分配进行比较。例如，当最优分配是将所有资源放在一个类别中时，因为它们总是在那里更有效，所以人们仍然按照其有效性按比例进行分配，从而将一些资源分配给效率较低的类别。同样，当分配应该基于边际有效性(额外资源的有效性)时，人们可能会根据平均有效性进行配置，或者根据类别的大小进行配置，或者只是平等地分配给所有类别(完全忽略有效性)。当在每个类别中投入一些资源是最有效的时候，人们可能会从绝对优先的角度来考虑，不想为了在另一个类别上的较小投资而牺牲一个类别。例如，根据类别的定义，类别的平均分配可能会发生变化。(“癌症”应该算作一个类别，还是每种类型都应该有一个类别？)基于对人们如何做出分配判断的理解，我们挑战人们自己的不一致，并询问他们将如何解决这些不一致。这些挑战，加上成本效益基本原理的指导，可以导致即使在没有精确计算的情况下也能做出更有效的判断。这项研究涉及一种新的方法。研究对象阅读假想和真实的场景，并在万维网上做出回应。通过使用网络浏览器能够理解的编程语言--脚本，可以编写交互式实验程序，在这些实验中，问题基于受试者早先的回答。

Heuristics in resource allocationProblems of resource allocation arise in decisions by governments and other institutions. Examples are allocation of medical research funds to diseases, allocation of funds for risk reduction to sources of risk, allocation of development funds to projects within a company, and allocation of university funds to departments. Some allocations have been criticized for inefficiency, e.g., risk reduction, where it has been argued that too much money is spent reducing some risks and too little money on others.Allocations arise in part from human judgments about how resources should be allocated. The judgments may be made by officials responsible for the allocations, by those affected, or by some other group. This research explores the principles of reasoning, or heuristics, that people use when they make such judgments. Research subjects make judgments about hypothetical and real cases. These judgments reveal the subjects' heuristics. We compare the results of these judgments to allocations prescribed by cost-effectiveness analysis of the same cases.For example, when the optimal allocation is to put all the resources in one category, because they are always more effective there, people still allocate in proportion to their effectiveness, thus giving some to the less effective category. Similarly, when allocation should be based on marginal effectiveness (effectiveness of additional resources), people may allocate on the basis of average effectiveness, or on the basis of the size of the category or just giving equal allocation to all categories (completely ignoring effectiveness). When it is most effective to put some resources in each category, people may think in terms of absolute priorities, not wanting to sacrifice one category for a smaller investment in another.Many of these heuristics can lead to inconsistent judgments. For example, equal allocation to categories can change depending on how the categories are defined. (Should "cancer" count as a category, or should there be one category for each type?) On the basis of the understanding of how people make allocation judgments, we challenge people with their own inconsistencies and ask how they would resolve these inconsistencies. These challenges, plus instruction in basic principles of cost-effectiveness, can lead to judgments that result in greater effectiveness, even in the absence of precise calculation.The research involves a novel methodology. Research subjects read hypothetical and real scenarios and respond to them on the World Wide Web. Use of JavaScript, a programming language understood by web browsers, allows programming of interactive experiments, in which questions are based on subjects' earlier answers.***