Skip to content Skip to navigation

Public versus Private Provision of Health Insurance

Aug 2015 to Jul 2018




Alfred P. Sloan Foundation

Project Goal:

The goal is to study the differences between health insurance managed by private firms and the public sector, by comparing the experience of Medicare beneficiaries in Medicare Advantage plans with those in public Medicare. Dimensions to be compared include services provided, costs and health outcomes.


The US has the most expensive health care system in the world, and relative to most other developed countries relies far more on private insurance companies. The project will take advantage of newly available data to compare the relative cost of private and publicly managed insurance, the mix of services they provide, and differences in health outcomes. It will focus on the US Medicare population, of whom around 30 percent now enroll in private insurance plans. Broadly, the project will contribute to a long-standing debate in economics over the relative efficiency of the public and private sector in providing essential services, and the regulation of sectors where the government turns to private providers.

Proposed Activities:
The project will involve a statistical analysis of newly available data on private insurance claims obtained from the Health Care Cost Institute, along with government data on all public Medicare claims. The project is highly collaborative due to the scale of the data, and will engage research assistants at both Stanford and MIT as parallel analyses will be run on public and private data to comply with confidentiality restrictions.

Expected Products:
The expected output is a series of academic papers that would be submitted to leading journals, presented at conferences and research seminars, and disseminated widely on the internet.

Expected Outcomes:
The main outcome of the research will be evidence on the comparative costs, services and outcomes of private insurers relative to the federal government. The project may also generate insights into the role of privatization in the healthcare sector and more broadly, and it is possible there could be methodological contributions related to the analysis of large-scale insurance claims data.