Medicare is one of the largest federal social insurance programs in the United States and the secondary payer for Medicare beneficiaries covered by employer-provided health insurance (EPHI). However, an increasing number of individuals are delaying their Medicare enrollment when they first become eligible at age 65. Using administrative data from the Medicare Current Beneficiary Survey (MCBS), this paper estimates the effects of EPHI, employment, and delays in Medicare enrollment on Medicare costs. Given the administrative nature of the data, we are able to disentangle and estimate the Medicare as secondary payer (MSP) effect and the work effects on Medicare costs, as well as to construct delay enrollment indicators. Using Heckman’s sample selection model, we estimate that MSP and being employed are associated with a lower probability of observing positive Medicare spending and a lower level of Medicare spending. This paper quantifies annual savings of $5.37 billion from MSP and being employed. Delays in Medicare enrollment generate additional annual savings of $10.17 billion. Owing to the links between employment, health insurance coverage, and Medicare costs presented in this research, our findings may be of interest to policy makers who should take into account the consequences of reforms on the Medicare system.

中文翻译：

---

医疗保险成本的实证模型：医疗保险、就业和延迟在医疗保险注册中的作用

Medicare 是美国最大的联邦社会保险计划之一，也是雇主提供的健康保险 (EPHI) 所涵盖的 Medicare 受益人的次要付款人。然而，越来越多的人在 65 岁时首次获得医疗保险资格时推迟了他们的医疗保险注册。 本文使用医疗保险当前受益人调查 (MCBS) 的行政数据，估计了 EPHI、就业和医疗保险注册延迟对医疗费用。鉴于数据的管理性质，我们能够分解和估计医疗保险作为次要支付者 (MSP) 的影响和对医疗保险成本的工作影响，以及构建延迟注册指标。使用 Heckman 的样本选择模型，我们估计 MSP 和受雇与观察积极的医疗保险支出和较低水平的医疗保险支出的概率较低有关。本文量化了 MSP 和就业每年节省的 53.7 亿美元。医疗保险注册延迟每年可额外节省 101.7 亿美元。由于本研究中提出的就业、医疗保险和医疗保险费用之间存在联系，我们的研究结果可能会引起政策制定者的兴趣，他们应该考虑医疗保险制度改革的后果。