The garbage can account of organizations where problems, solutions, and people chase each other is often invoked but rarely studied since its publication 44 years ago. It has been critiqued for being a metaphor rather than a model, and offering a deterministic rather than stochastic account. We reline the garbage can model of organizational decision-making by modeling the arrival of problems, people, and solutions as queues that get matched randomly. We show that queuing models allow us to understand the effect of using either experts, supervisor approval, teams, and deviation from supervision on problem resolution and oversight. Our approach shows that manager approval increased the standard deviation of problem resolution, whereas queues are processed faster and have lower variance when there is oversight or teamwork, or when manager approval is bypassed due to independent action by problem solvers. It also shows the costs of using an organizational hierarchy to address problems with different levels of difficulty, or specialization to address a mixture of fundamentally different problems. Thus, a stochastic garbage can model provides insights into why organizations make many decisions but often fail to resolve problems!

中文翻译：

---

调整组织决策的垃圾箱：将问题和解决方案的到达建模为队列

垃圾可以说明组织，问题，解决方案和人们互相追逐的组织，因为该组织自44年前发布以来就很少被研究。人们批评它是一个隐喻而不是模型，并提供了确定性而非随机性的解释。通过将问题，人员和解决方案的到达建模为随机匹配的队列，我们​​对组织决策的垃圾桶模型进行了重新排列。我们证明，排队模型使我们能够了解使用专家，主管批准，团队和偏离监督对问题解决和监督的影响。我们的方法表明，经理的批准增加了问题解决方案的标准偏差，而在进行监督或团队合作时，队列的处理速度更快，差异也较小，或由于问题解决者的独立行动而忽略了经理的批准。它还显示了使用组织层次结构来解决具有不同难度级别的问题或使用专业化来解决根本不同问题的混合所带来的成本。因此，随机垃圾箱模型可以洞察组织为何做出许多决策却常常无法解决问题！