Skip Navigation Archive: U.S. Department of Health and Human Services
Archive: Agency for Healthcare Research Quality
Archive print banner
Performance Budget Submission for Congressional Justification

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Medical Expenditure Panel Survey (MEPS)

Purpose and Method of Operation

The objectives of AHCPR's Medical Expenditure Panel Survey are to provide public and private sector decision makers with the ability to:

  • Obtain timely national estimates of health care use and expenditures, private and public health insurance coverage, and the availability, costs, and scope of private health insurance benefits among the U.S. population.
  • Analyze changes in behavior as a result of market forces or policy changes (and the interaction of both) on health care use, expenditures, and insurance coverage.
  • Develop cost/savings estimates of proposed changes in policy.
  • Identify the impact of changes in policy for key subgroups of the population (i.e., who benefits and who pays more).

These objectives are accomplished through the fielding of the Medical Expenditure Panel Survey (MEPS). MEPS is an interrelated series of surveys that replaces the National Medical Expenditure Survey (NMES). MEPS not only updates information that was last collected a decade ago in FY 1987 but it will also provide more timely data, at a lower cost per year of data, through the move to an ongoing data collection effort.


AHCPR's Medical Expenditure Panel Survey collects detailed information regarding the use and payment for health care services from a nationally representative sample of Americans. No other surveys supported by the Federal Government or the private sector provide this level of detail regarding: the health care services used by Americans at the household level and their associated expenditures (for families and individuals); the cost, scope, and breadth of private health insurance coverage held by and available to the U.S. population; and the specific services that are purchased through out-of-pocket and/or third-party payments.

This level of detail enables public and private-sector economic models to develop national and regional estimates of the impact of changes in financing, coverage, and reimbursement policy and estimates of who benefits and who bears the cost of a change in policy. No other survey provides the foundation for estimating the impact of changes on different economic groups or special populations of interest, such as the poor, elderly, veterans, the uninsured, or racial/ethnic groups. That is why these surveys have always been seen as a critical national resource by the private sector (e.g., RAND, Heritage Foundation, Lewin-VHI, and the Urban Institute) as well as the public sector (e.g., Office of Management and Budget (OMB), Congressional Budget Office (CBO), Physician Payment Review Commission (PPRC), Prospective Payment Assessment Commission (ProPAC), and Treasury Department).

Within the past few years, the Department has initiated a Survey Integration Plan. The major goal of the Plan is to improve the effectiveness and efficiency of the major HHS surveys. MEPS is the result of that effort. Unlike NMES, which developed its own large sampling frame of families to interview, MEPS relies upon an existing nationally representative sampling frame that was developed by the National Center for Health Statistics (NCHS). The linkage between the MEPS household survey and NCHS' National Health Interview Survey (NHIS) has achieved significant savings in sample frame development and enhancements in analytical capacity.

MEPS builds upon the strengths of the 1977 and 1987 NMES and streamlines the Department's data collection efforts. This integration will assure a better return on the Federal Government's investment by moving from a periodic survey once every ten years to a continuing longitudinal data collection effort. Over time, MEPS will provide more comprehensive data for public and private sector decision makers.

The first MEPS data (from 1996) became available in April 1997 and key findings are summarized in the table below. This rich data source has become not only more comprehensive and timely, but MEPS' new design has enhanced analytic capacities, allowed for longitudinal analyses, and developed greater statistical power and efficiency.

MEPS Data—Key Findings

Date Public Use Data ReleaseKey Findings
April 1997 Round 1 1996

Medical Expenditure Panel Survey Component person-level demographics, employment, health status, and health insurance.
17% of the U.S. populationwas uninsured throughout the first half of 1996. 19.6% of privately insured children in single-parent families get health insurance coverage from a policyholder not residing in their household. Almost 53% of children covered by Medicaid have at least one parent that works.

A greater percentage of workers are being offered health insurance by their employers in 1996 than in 1987; however, a smaller proportion are accepting insurance.
April 1997 Round 1 1996

MEPS—Nursing Home Survey Component—person-level demographic and health status data and Round 1 facility characteristics data.
Almost 66% of U.S. nursing homes in 1996 were operated for-profit.

Over 80% of nursing home residents needed help with 3 or more activities of daily living. Almost half of all nursing home residents have some form of dementia.
September 1997 Release 2 1996

Household Survey Component—Round 1 parent identifiers and managed care data, and Round 2 health status and access to care data.
Nearly 19% of the population had no usual source of health care in 1996.

About 12% of all American families experienced barriers to receiving needed health care services.
Summer 1998 MEPS—Household Survey Component 1996—population characteristics and utilization data file. Full-year demographics, health insurance, employment, and utilization.
Summer 1998 MEPS—Nursing Home Survey Component 1996—population characteristics and residence history data file. Nursing home resident person level demographics, residence history, and insurance variables.
Winter 1998 MEPS—Nursing Home Survey Component—use and expenditure data file. Nursing home resident person level health status; demographics; variables indicating living spouse, siblings, children, or parents; prior long-term care use; veteran status; use of health care services; use of prescribed medicines; insurance; income and assets; and health care expenditures.
Winter 1998 MEPS—Household Survey Component 1996—use and expenditure data file. Person level use, source of payment, expenditures, health status, health insurance, and income.
1999 1996-1997 Panel Household Component—conditions, events, and job data files. Detailed condition, event, and job information.
1999 Linked Household Component Survey—Insurance Component Survey data file. 1996 full-year use and expenditure file linked with 1996 Insurance Component file.

Funding History

Funding for The MEPS program prior to FY 1999 has been as follows:

Year AmountFTEs
1994 $10,000,000 ---
1995 $15,000,000 ---
1996 $15,100,000 ---
1997 $38,886,000 ---
1998 $36,300,000 ---

Sources of MEPS funding follow:

Year Budget Authority 1% EvaluationTotal
1994 Actual $10,000,000 --- $10,000,000
1995 Actual $9,918,000 $5,082,000 $15,000,000
1996 Actual $10,000,000 $5,000,000 $15,000,000
1997 Actual $224,000 $38,662,000 $38,886,000
1998 Appropriation $0 $36,300,000 $36,300,000

Rationale for the Budget Estimate

The FY 1999 request for MEPS totals $27,800,000 in 1% evaluation funds. The total reflects a decrease of $8,500,000 from the FY 1998 level of $36,300,000.

The decrease of $8,500,000 from the FY 1998 level is the net effect of three factors: the conclusion of the Nursing Home Component of MEPS, the end of the development and testing phase of the survey, and the move to the lower sample size of a non-peak year in the sample (the Department's Survey Integration Plan calls for five-year peaks in sample size for MEPS.

In FY 1999, data collection will be ongoing for the MEPS Household Survey, the MEPS Medical Provider Survey, and the MEPS Insurance Component (MEPS-IC), which consists of the MEPS Health Insurance Plans Survey and the former National Employer Health Insurance Survey (NEHIS). More specifically, in-person interviews will be conducted with 9,000 families to obtain calendar year 1998 health care data, and with 5,600 new families sampled for the 1999 MEPS household survey. The Medical Provider Survey conducted in 1999 will consist of interviews with approximately 3,000 facilities, 12,000 office-based providers, 7,000 hospital-identified physicians, and over 500 home health providers. In addition, the Insurance Component in 1999 will consist of interviews with more than 40,000 employers and 1,000 insurance carriers.

All of the MEPS components will be heavily engaged in survey-related activities directed to the following tasks: data editing, imputation, data preparation and data processing, development of estimation weights and variance estimation capabilities for the component surveys, preparation of public use tapes, and development of analytical and methodological reports.

The design of the surveys under the ongoing HHS Survey Integration Plan calls for a smaller panel in each of the four years between the quinquennial peak sample size years, yielding over time considerably more data for reduced fixed costs. The implementation of the HHS Survey Integration Plan will provide the Congress, OMB, the Department, the private sector, and the research and policy community with continuous, up-to-date data on health care, use, expenditures, and the employer health insurance market. This plan greatly enhances the analytic capabilities over the separate NHIS, NMES and NEHIS surveys. The Department, in cooperation with OMB, is continuing the survey integration initiative in order to achieve more efficiencies and savings.

Data collection for the Insurance Component (MEPS-IC)is now being conducted by the Census Bureau rather than by the original contractor. This move is cost efficient as a consequence of the sample design capacities that the Census Bureau has made available to the survey at marginal cost, and the opportunity to use and make marginal changes to existing survey control systems rather than having to redevelop them. AHCPR is continuing its efforts to negotiate further efficiencies in all areas of the MEPS.

A Critical Resource

Since 1977, AHCPR's expenditure surveys have been an important and unique resource for public and private sector decision makers. No other surveys effectively and efficiently link data about persons, their families, employers, and providers with data on health care use, expenditures, insurance coverage, and health status.

The data from earlier surveys (1977 and 1987) have quickly become a linchpin for the nation's economic models and their projections of health care expenditures and utilization. In the public sector, OMB, CBO, PPRC, ProPAC, and the Treasury Department among others, rely upon these data to evaluate health reform policies, the effect of tax code changes on health expenditures and tax revenue, and proposed changes in government health programs such as Medicare. In the private sector, these data are used by many private businesses, foundations and academic institutions to develop economic projections. These data represent a major resource for the health services research community at large.

MEPS dramatically improves the value and cost-effectiveness of AHCPR's data collection effort. Survey integration means that MEPS will yield a broader array of data than NMES. The move from a large survey every ten years to following a smaller cohort of families on an ongoing basis has four primary benefits: it will decrease the cost per year of data collected; it will provide more timely data on a continuous basis; it will create for the first time the ability to assess changes over time; and it will permit the correlation of these data with the National Health Accounts.

The timeliness of the data provided by MEPS will enhance the accuracy of the economic models upon which the Administration, and the private sector rely. MEPS will provide concrete and current insights regarding the dynamics of insurance coverage, job lock, the adequacy of insurance coverage, spousal coverage, health insurance coverage from multiple sources, and the sites and providers of specific health care services.

Major Components of MEPS

MEPS consists of a series of interrelated surveys. The individual components of MEPS and the information each will provide is outlined as follows.

  • Household Survey—Six interviews will be conducted with each new sample of households for MEPS over a 3-year period to obtain health care use, and expenditure and insurance coverage data for two consecutive calendar years. The 1996 MEPS sample included 10,800 families, the 1997 MEPS sample included 6,300 families, the 1998 MEPS sample will include 5,200 families and the 5,600 new families. This is the only survey that collects health care expenditures of American families.
  • Medical Provider Survey—Interviews will be conducted with 3,000 hospitals, 12,000 office-based physicians, 500 home health providers, and 7,000 hospital-identified physicians in 1999 to obtain health care information on MEPS Household Survey participants. A separate interview will be conducted to obtain information for each of the calendar years associated with the MEPS Household Survey Panel. This survey allows AHCPR to verify information collected at the household and to get information from providers when it is not known by the household (e.g., households receiving Medicaid or that are in capitated plans will not know their expenditures).
  • MEPS Insurance Component—Health Insurance Plans Survey—Interviews will be conducted with 10,000 employers, 500 unions, and 500 insurers in 1999 to obtain detailed information on the health insurance benefits and premiums associated with health insurance coverage held by the MEPS Household Survey participants. A separate interview will be conducted to obtain information for each of the calendar years associated with the MEPS Household Survey Panel. This survey allows evaluation of insurance purchasing choices and the impact of those choices on use and access to care.
  • MEPS Insurance Component Establishment Level (formerly NEHIS)—Interviews will be conducted with 30,000 establishments to obtain national and state-specific (40 states) estimates of the availability of health insurance at the workplace, the type of coverage provided by employers, and the associated costs of coverage. For each establishment surveyed, the MEPS-IC Establishment Level Survey will obtain information on the number and characteristics of plans offered, the scope and breadth of benefits included in each plan and the corresponding copayment provisions, the number of current workers and retirees enrolled in each plan, and whether each plan is fully or self-insured. The MEPS-IC Establishment Level Survey data will also include characteristics of each establishment including its size, the type of workforce employed, aggregate data on payroll and available fringe benefits, industrial classification, and corporate status. National Nursing Home Expenditure Study—This survey will provide calendar year expenditure estimates for nursing home care for persons residing in a nursing home anytime during 1996. In addition to making expenditure estimates by sources of payment, the survey estimates changes in sources of payment (e.g., Medicaid) over the year. The design of the survey permits expenditure estimates by important sub-populations including persons admitted or discharged during the year; by functional health status, insurance coverage, and income distributions; and by characteristics of the nursing home residents' care-givers who reside in the community. Also, the survey can provide estimated physician use, prescribed medicine use, and hospital use by nursing home residents. No other national nursing home survey has the capability of uniformly measuring health status using data collection materials based on HCFA's Resident Assessment Instrument or collecting data from the nursing home residents' next-of-kin who reside in the community.

The information derived from these surveys will enable the Congress, the Administration, and other public and private sector policy makers to evaluate the impact of:

  • Growing enrollment in managed care.
  • Enrollment in different types of managed care.
  • Changes in how chronic care and disability are managed and financed.
  • Alternative approaches to provision of long-term care.
  • Changes in employer-supported health insurance.
  • Changes in Federal and State policy.

In addition to assessing broad trends such as these, MEPS will address a host of specific, policy-relevant questions. Examples of these questions are outlined in the section on significant questions addressed by MEPS.

MEPS will make a significant contribution to improving the accuracy of the nation's economic models in at least two ways. First, with each passing year, attempts to extrapolate the 1987 data to the current year are increasingly subject to error; current data are essential and MEPS will provide them. Second, the fact that this survey will provide a third data series for these sophisticated models will significantly improve their ability to identify and assess trends that may not be obvious with data from only two points in time (1977 and 1987). Thus, the critical importance of collecting this third series of data cannot be overemphasized.

The Future

Integrating many complex surveys while decreasing costs and enhancing the Department's analytic capabilities is a continuing effort. During 1997, AHCPR explored the possibility of using the Census Bureau for collection of Household Survey data. This proved to be unfeasible and AHCPR will publish a solicitation in February 1998 for this purpose, with decisions expected by Summer 1998.

In FY 1999, the MEPS will support research on the health effects of welfare reform. Particular attention will be given to research that investigates the health consequences of welfare to work programs. Specific areas of emphasis will include health insurance coverage among former welfare recipients who became employed, and health care access and utilization by former welfare recipients who exhaust their traditional Medicaid benefits.

Return to Budget Justification Statement

Current as of February 1998


The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care