Dynamics of Plan Choice and Prescription Drug Utilization in Medicare Part D

描述（由申请人提供）：医疗保险D部分中计划选择和处方药使用的动态。医疗保险D部分处方药福利开创了使用多个私人保险提供商提供公共保险产品的先机，因此提出了许多重要的政策问题。消费者有能力在这么多的私人保险选择中做出明智的选择吗？一旦参加了一项计划，消费者是否对他们全年面临的非线性价格表做出了适当的反应？在我们的工作中，关键的方法论创新是构建模型，在这些模型中，选择揭示了消费者偏好的信息，即使这些选择不一定是最优的。这建立在我们早期工作的基础上，该工作使用不完整的数据集来记录老年人选择D部分计划不一致的证据。我们将使用CMS提供的20%的完整医疗保险D部分索赔样本来复制和扩展这项工作。特别是，我们将评估消费者是否倾向于随着时间的推移而改善，因为受益人在市场上获得了经验，或者消费者是否倾向于留在次优计划中，要么是因为他们试图选择一个更好的计划而失败，要么是因为他们惯性地留在同一个计划中。我们将开发一个结构模型，它允许我们分解驱动计划选择随时间变化的需求和供给方面的因素。参加医疗保险D部分的个人面临着一个复杂的动态消费决策：今天的药品价格取决于迄今为止的自付额。有远见的消费者在决定一年中的每个时间点的药物消费时，应该考虑他们全年的整个预期支出路径，而时间不一致的消费者可能不会这样做。我们将针对消费者面临的非线性预算集建立一个药物消费的动态模型，该模型包含不确定性和潜在的折扣不一致性。围绕药物计划定价表变化的药物消费变化使我们能够确定选择是否随着时间的推移是一致的，以及个人对价格不确定性的反应。我们还将扩展我们的分析，以考虑处方药使用对下游医生和医院支出的影响。我们的经验策略将使我们能够估计医院和医生支出相对于处方药支出（和共同保险）的总体弹性，以及不同类型处方药的更具体的弹性。我们可以利用这些结果来预测D部分（例如，填补甜甜圈洞）下价格的时间路径变化将如何影响处方药和其他类型医疗消费的年度消费模式。

DESCRIPTION (provided by applicant): Dynamics of Plan Choice and Prescription Drug Utilization in Medicare Part D. The Medicare Part D prescription drug benefit pioneered the use of multiple private insurance providers to deliver a public insurance product, and as such raises a host of important policy questions. Are consumers capable of making an informed choice across so many private insurance options? Once enrolled in a plan, do consumers respond appropriately to the non-linear schedule of prices they face throughout the year? The key methodological innovation in our work is to construct models where choices reveal information about consumers' preferences, even when those choices are not necessarily optimal. This builds on our earlier work, which uses an incomplete data set to document evidence of inconsistencies in elders' choice of Part D plans. We will replicate and extend this work using a complete 20% sample of Medicare Part D claims provided by CMS. In particular, we will assess whether consumers tend to improve over time as beneficiaries gain experience in the market or whether consumers tend to remain in sub-optimal plans, either because they attempt to choose a better plan and fail, or because they remain in the same plan through inertia. We will develop a structural model which allows us to decompose the demand and supply side factors which drive changes in plan choices over time. Individuals enrolled in Medicare Part D face a complicated dynamic consumption decision: the prices of drugs today depend on out of pocket payments to date. Forward-looking consumers should consider their entire expected path of spending throughout the year in deciding on drug consumption at each point during the year, while time-inconsistent consumers may not do so. We will build a dynamic model of drug consumption with respect to the nonlinear budget sets facing consumers that incorporates uncertainty as well as potential discounting inconsistencies. Changes in drug consumption around the variance in drug plan pricing schedules allow us to identify both whether choices are consistent over time and how individuals react to price uncertainty. We will also extend our analysis to consider the impacts of prescription drug utilization for downstream physician and hospital spending. Our empirical strategy will allow us to estimate both an overall elasticity of hospital and doctor spending with respect to prescription drug expenditures (and coinsurances), and more particular elasticity's for different types of prescription drugs. We can use the results to provide projections of how changes in the time path of prices under Part D (e.g. filling in the donut hole) will impact annual consumption patterns of both prescription drugs and other types of medical consumption.