Burden of Health Care Costs: Businesses, Households, and Governments, 1987-2000
Cowan, Cathy A., McDonnell, Patricia A., Levit, Katharine R., Zezza, Mark A., Health Care Financing Review
In this article, we estimate health care spending by sponsor type--businesses, households, governments, and other private funds; track trends in spending over time; and analyze the burden that these expenditures impose on the sponsoring entities. The basis for these estimates is the national health accounts (NHA), the official Federal Government estimates of total U.S. health care spending (Levit et al., 2002).
This presentation differs from the usual NHA arrangement of sources of funding. The NHA structure includes both expenditures for health care services and sources that pay for these services. These sources generally define an entity, usually a third-party insurer, that is responsible for paying the health care bill. These funding sources are broadly classified into private health insurance (PHI), out-of-pocket spending, and specific government programs, such as Medicare and Medicaid. A small portion of expenditures is estimated for other private revenues--philanthropic giving and revenues received by some health care providers from non-health services (e.g., cafeteria and gift shop sales and revenue from educational services). This structure is useful for tracking changes in who (or what public program) is paying for different types of health care services. It is also useful in analyzing the impact of specific public program policy changes on public or private insurance.
For certain financing decisions and policy issues, however, this structure is not optimal. Often the financial burden of paying for coverage resides not with the bill-paying entity, but with the businesses, households, and governments paying insurance premiums or financing health care through dedicated taxes. These entities frequently decide what health care plan is offered to whom, what cost-sharing arrangements (premiums, copayments, and deductibles) will be imposed, and the breadth and depth of coverage. As health care cost burdens change, the decisions made by businesses, households, and governments in these respects are altered, as are policy responses by government to these decisions. Thus, for many purposes, it is helpful to focus not just on who pays the bills for health care services (as tracked in the traditional NHA) but also on the underlying source of financing for health care.
To estimate the burden of health care, the existing NHA estimates for health services and supplies have been disaggregated and rearranged into categories reflecting the sponsors of health care--businesses, households, and governments. This process includes separately estimating PHI premiums paid by private employers, Federal employers, State and local employers, employees, and individuals. In addition, financing sources for Medicare are estimated and counted with their respective sponsors. These sources include private, Federal, State, and local employer and employee contributions through the Federal Insurance Contributions Act (FICA) taxes to the Federal Hospital Insurance (HI) Trust Fund. It also includes Supplementary Medical Insurance (SMI) premiums paid by individuals and Medicaid "buy-ins." (Medicaid buy-ins are payments by State Medicaid programs of Medicare Part A and Part B premiums for eligible individuals.) Finally, workers' compensation spending and temporary disability insurance are reallocated to employers who sponsor these benefits.
Although we categorize sponsors into businesses, households, and governments, individuals ultimately bear the responsibility of paying for health care through taxes, reduced earnings, and higher product costs.
This article is an update of earlier articles (Cowan and Braden, 1997; Cowan et al., 1996; Levit and Cowan, 1991; Levit et al., 1989). Consistent definitions have been used throughout these articles. However, revisions to the NHA, the basis for the estimates presented in this article, have resulted in revisions to these sponsor estimates. In addition, data sources have evolved, and consequently the methodology used to produce these estimates has changed. …