Page 1 of 1

Employer-Sponsored Health Insurance

Trends in Cost and Access

Research in Action Issue 17

This article briefly summarizes findings from the Medical Expenditure Panel Survey (MEPS) on employer-sponsored insurance that have resulted from the Agency for Healthcare Research and Quality (AHRQ)-funded research conducted to date.

By Mark W. Stanton, M.A. 


For decades, Federal and State policymakers have sought ways to offer affordable health insurance coverage, but this goal has been elusive. The number of uninsured remains high and costs for health care insurance keep growing, even though legislation has been enacted and regulatory changes have been made that affect both public and private programs. The U.S. employer-based health insurance market provides insurance coverage to nearly two-thirds of the population under 65. In addition, nearly 80 percent of the uninsured live in a family where at least one adult is employed. Therefore, building on these programs might be an attractive component of any solution. But it is essential to have sound, evidence-based information about that system to make informed decisions.

This Research in Action is intended to answer questions that might arise during discussions about options. Based on data obtained by the Federal Agency for Healthcare Research and Quality (AHRQ) in the Medical Expenditure Panel Survey (MEPS), an ongoing series of annual surveys, it answers the questions:

  • What are the recent trends in offer, eligibility, and enrollment rates?
  • What impact does firm size have on whether or not employers offer health insurance?
  • What factors affect whether employees enroll in insurance programs?
  • Who is less likely to have employer-sponsored insurance?

Making a Difference

  • Offer rates have risen while eligibility and enrollment rates have declined.
  • Cost is a major reason for declining enrollment.
  • Larger firms are more likely to offer insurance.
  • Low-wage workers are more sensitive to the size of the employee premium contribution.
  • Hispanics, young adults, near-elderly working women with health problems are more likely to be uninsured.


Having health insurance is valuable for a number of reasons. People who are insured are protected against uncertain and high medical expenses and are more likely to receive needed and appropriate health care. In addition, having health insurance is associated with improved health outcomes and lower mortality, so employees with health insurance are more likely to be productive workers.1

In the first half of 2003, the U.S. employer-based health insurance market provided insurance to over 159 million Americans who constitute nearly two-thirds (63.4 percent) of the population under 65.2 In this voluntary system, employers may choose whether to offer health insurance to their employees and employees may choose to enroll or forgo enrollment. Most employers have chosen to offer health insurance as a fringe benefit to attract employees. It is an attractive option to offer because of the favorable tax treatment to both employer and employee. In addition, employment-based health insurance is likely to be less expensive than individually purchased coverage (for the same set of benefits) and typically provides a broader scope of benefits than is available in individually purchased coverage. The stability of the employer-sponsored insurance system is supported by a tax subsidy that promotes the pooling of risk necessary for the successful functioning of insurance.3,4

When an employer offers health insurance to its employees, usually some conditions are attached, such as a waiting period before benefits take effect for new employees, a requirement that the employee be full time or work a specified number of hours per week, and a requirement that employees pay a portion of the premium. Small employers with predominantly low-wage workers are much less likely to offer health insurance since the greater cost of underwriting and administering coverage for each enrollee in a small workforce adds to the cost of the premium. Since smaller firms face higher rates of employee turnover, it is likely that a smaller percentage of their workers will fulfill the waiting period required for enrollment. Also, since small businesses experience higher rates of business failure than larger firms do, they have greater incentives to keep their costs of doing business low.5 Finally, the lack of coverage at such firms may reflect the weaker preferences of small-firm employees for compensation in the form of health insurance benefits compared to wage income.6

The system of employer-sponsored health insurance has long provided coverage to the vast majority of America's workers and their dependents. Many of the rest are insured through public programs such as Medicaid and the State Children's Health Insurance Program (SCHIP) or through privately purchased insurance in the individual market. Yet there remains a sizable group of workers and their dependents who are uninsured. Persons who are uninsured over the long term experience poorer health and earlier death than the insured population, and it is estimated that such problems are associated with societal costs ranging from $65 billion to $130 billion per year.7

The Medical Expenditure Panel Survey

AHRQ's Medical Expenditure Panel Survey (MEPS) is an ongoing series of annual surveys designed to continually provide policymakers, health care administrators, businesses, and others with timely, comprehensive information about health care use and costs in the United States and to improve the accuracy of their economic projections. As a health care survey, MEPS emphasizes and provides a broad range of estimates on insurance status. Its utility as a natural benchmark has been reinforced by the adoption of MEPS premium cost estimates in the Gross Domestic Product. MEPS is unique in its ability to link data on individuals and households (including demographics, health status, employment, and income) to information on their use of health care. No other survey contains such a wide range of data essential for analyzing health spending and insurance coverage.

For purposes of this synthesis, the household and insurance components of MEPS are especially important since they offer detailed statistical information on many health-insurance-related questions. In 2003, the MEPS-HC (Household Component) surveyed 15,000 families including 37,000 individuals; it has had a response rate of over 70 percent. The MEPS-IC (Insurance Component) surveys 42,000 business establishments and governments and has a response rate of 78 percent.

For more information on MEPS, go to

Factors Affecting Insurance Trends

What Are the Recent Trends in Offer, Eligibility, and Enrollment rates?

In recent years (1996-2002), countervailing trends among offer, eligibility, and enrollment rates have characterized the employer-sponsored health insurance market. Offer rates have risen while eligibility and enrollment rates have gone down, whether measured in relation to all private-sector employees, private-sector full-time employees, or private- sector part-time employees (Table 1).

For example, from 1996 to 2002, among all private-sector employees, the offer rate increased from 86.5 percent to 88.3 percent, the eligibility rate decreased from 81.3 percent to 77.1 percent, and the enrollment rate dropped from 69.6 percent to 62.4 percent. In addition, the enrollment-when-eligible rate dropped 4.5 percentage points, from 85.5 percent in 1996 to 81.0 percent in 2001.8 Some of the decrease in the enrollment rate may result from an increase in workers who enroll in their spouse's employer-sponsored plans instead of their own. One study reporting on a drop in enrollment in the period 1987-96 found that it was partly offset by a rising rate of enrollment in a spouse's plan.6

According to recent research, the principal reason for decreasing enrollment during the period 1987-2000 was that the cost to employees had risen substantially.9 An earlier analysis of enrollment trends in the period 1987-96 suggested that decreasing enrollments have resulted from a variety of factors, including increasing costs of insurance, rising employee share of health insurance premium cost, expansions in Medicaid, declining real incomes, and increased price-consciousness among workers.6


  • The offer rate is the percent of employees who work where insurance is offered.
  • The eligibility rate is the percent of employees eligible to enroll where insurance is offered. In most cases, to be eligible for employer-based coverage, employees must have full-time status and also pass through a waiting period after commencement of employment.
  • The enrollment rate is the percent of all who work where insurance is offered who enroll.
  • The enrollment-when-eligible rate is the percent of eligible workers who enroll in offered coverage.

Note: Definitions are taken from the MEPS-IC (Insurance Component).

What Is the Impact of Firm Size on Offer Rates?

Firms with large numbers of employees are more likely to offer their employees health insurance:

  • Among medium and large private-sector employers (firms with 50 or more employees), 97.8 percent of all employees worked where health insurance was offered in 2002.10
  • Among small private-sector employers (firms with fewer than 50 employees), 63.5 percent of all employees worked where health insurance was offered.11

Small employers not offering health insurance typically cite the cost of insurance and the belief that employees have other sources of coverage (e.g., through a spouse) as their principal reasons (Figure 1). The firms least likely to offer coverage are those with low-wage workers, high turnover, no unions, and many part-time employees.12 The smaller the firm size, the more likely are its employees to be uninsured. However, it is interesting to note that the proportion of uninsured workers employed even by quite large firms (500 or more workers) has grown in recent years. Between 1987 and 2001, the proportion of uninsured workers in these firms expanded from 25 to 32 percent. Declines in manufacturing jobs and unionization rates are the most important reasons for the rise in uninsured workers at these large firms.12

What Factors Affect Whether Employees Enroll in Insurance?

Many factors affect whether or not employees enroll in a health insurance plan. Employees who enroll are more likely to be male, work full time, belong to a union, and work in public administration or the public sector. Married workers with or without children are less likely to enroll than unmarried workers since, in many cases, married people have the option of being covered under a spouse's policy.

Two important factors affecting enrollment are the size of the employee premium contribution and the wage level of the workers.

Premium Contributions

In general, the higher the employee contribution is, the less likely employees are to enroll:

  • In 1999, employers offering a plan without employee contributions had an enrollment rate of 87.3 percent; employers requiring an employee premium contribution had an enrollment rate of 77.0 percent.13
  • Between 1996 and 2002, the average employee premium contribution per enrolled private-sector employee for single coverage rose 65.2 percent, from $342 to $565.8
  • In the same period, the average employee premium contribution per enrolled private-sector employee for family coverage rose 55.8 percent, from $1,275 to $1,987.8
  • In 2002, the average employee premium contribution per enrolled employee for single coverage was 19.0 percent lower in small firms than in large firms.14
  • However, the average employee premium contribution per enrolled employee for family or employee-plus-one coverage was over 10 percent higher in small firms than in large firms.14

In an AHRQ study, a model was developed to examine the influence of out-of-pocket contributions, total premiums, and employer and workforce characteristics on workers' enrollment decisions. The researchers estimated that eliminating employee premium contributions had the potential to reduce the rate of employees declining coverage by almost a fifth. If the cost to employees of single coverage were to fall to zero, 18.1 percent, or 2.5 million of the 13.8 million private-sector employees who currently decline insurance coverage from their employers, might enroll in such insurance. Using a statistical model, the researchers also found that the take-up rate with coverage at no cost to employees in predominantly low-wage establishments would increase by almost 10 percentage points, compared to an increase of 3.9 percentage points in predominantly high-wage establishments.13

Wage Levels

Although people insured through their employment are relatively less sensitive to premium size than those in the individual market since they pay only a fraction of the full premium, low-wage workers are more sensitive to the size of the contribution than are higher income workers.15

  • During 2002, 63.5 percent of employees in predominantly low-wage establishments (50 percent or more of employees earn less than $9.50 per hour) were enrolled in a health insurance plan.
  • In establishments in which less than 50 percent of employees earned less than $9.50 per hour, 83 percent of workers were enrolled in a health insurance plan.8

Who is less likely to have employer-sponsored insurance?

In the first half of 2002, 60 percent of uninsured adults and 75 percent of uninsured children lived in a family with at least one adult full-time worker.16 Groups less likely to have employer-sponsored insurance include:

  • Workers in small establishments.
  • Minorities (especially Hispanic males).
  • Young adults (19-24).
  • Near-elderly working women with health problems.
  • Retirees.a

In fact, as the establishment size increases, the percentage of uninsured workers drops (Figure 2). Other groups less likely to be insured are the self-employed, low-wage workers, nonunionized workers, and part-time workers.


Declining enrollment rates for Hispanic workers are increasing the disparity in Hispanic-white coverage rates, according to an AHRQ study.17 Between 1987 and 1996, no change was observed in the large differential in health insurance coverage between most minority groups and white Americans, except for Hispanic men, where this differential widened. The most important factors contributing to this widening gap between minorities and whites were a declining union membership rate and a greater likelihood among non-elderly American workers of being in a poor or low-income household. Among non-elderly American workers:

  • Employment-related insurance coverage among Hispanic men declined 13.0 percentage points between 1987 and 1996.
  • The decline among Hispanic and black women was 7 percentage points, while the decline was less than 3 percentage points for white men.18
  • By 2003, 38.5 percent of Hispanic males under 65 were uninsured, compared to 16.2 percent of white males.19

Young Adults, Near-elderly Working women, Retirees

Young adults and near-elderly working women with health problems were other groups less likely to have employer-sponsored insurance and more likely to be uninsured. Although most retirees are eligible for Medicare, they are becoming less likely to receive employer-sponsored retiree health benefits.

Young adults ages 19-24 were the age group most likely to be uninsured, with 36.3 percent uninsured for all or part of 2003, according to MEPS data.20 At age 19, most young adults who do not go to college are dropped from their parents' insurance coverage. In making the transition to work, many young adults are in low-wage or temporary jobs where health insurance is not offered. If they are from low-income families (below 200 percent of Federal poverty level), they are much more likely to be uninsured.20

In 1996, near-elderly working women (ages 55-64) with health problems, a group comprising 1.8 million persons, were at much greater risk of being uninsured (22.7 percent) than their counterparts in very good or excellent health (9.5 percent). Although near-elderly workers were more likely than prime-age workers (30-54) to purchase coverage outside the workplace, there has been a decline in private coverage purchased outside the workplace by women in this age group. Factors associated with the greater likelihood of near-elderly women with health problems being uninsured appear to be lower income and less access to jobs offering insurance.21

The availability of employer-sponsored health insurance for retirees is shrinking.b The availability of employer-sponsored health insurance is much broader for current full-time employees than for retirees and varies by economic sector and size of the employer. With the exception of large governments, the percent of full-time employees who worked where health insurance was offered to retirees fell from 1998 to 2000.22

The decrease in the offer rate to retirees varied by size of firm. For retirees age 65 and over, from 1998 to 2002:

  • Among firms with less than 50 employees, the availability of retiree coverage declined from 5.6 percent to 2.1 percent.
  • Among firms with 100 to 999 employees, the decrease was from 18.7 percent to 9.2 percent.
  • Among firms with 1,000 or more employees, the decrease (from 41.2 percent to 41.1 percent) was not significant.8

 aAn establishment refers to a particular workplace or location. A firm (the entire company) consists of one or more establishments under common ownership or control. A large firm may have many small establishments.

 bRetirees typically are insured by Medicare but about half of people 65 and over also have supplemental insurance through their former employers or through the private individual market.

Effects of Small Group Insurance Reforms

Beginning in the early 1990s, a number of States passed legislation to enhance the availability and affordability of health insurance to employees of firms with fewer than 50 employees. These reforms included provisions to guarantee the issuance and renewal of health plans, to diminish premium variability and rates of growth across small employer groups, to limit the use of pre-existing condition exclusions, and to enhance portability of coverage.

A study of the impact of these reforms using data from the 1987 National Medical Expenditure Survey (NMES) and 1996-97 MEPS found that the reforms had little impact on offer rates and enrollment rates. However, the study found that in States with the most stringent reform, employment-related coverage increased for high-risk workers relative to low-risk workers.23

MEPS Data and Research Publications

MEPS disseminates its health care data and publications in a variety of formats.

  • Public use files can be:
    • Downloaded from the MEPS Web site:
    • Ordered from AHRQ as diskettes (floppy disks) or CD-ROMs.
  • MEPS publications:
    • Are available on the MEPS Web site to be viewed online or printed as PDF files.
    • Can be searched through the MEPS Publications Database (Web site:, which allows searching for specific publications and sorting of results by various options.
  • MEPS research done by outside researchers:
    • May be available as reprints through the AHRQ Web site:
    • May be viewable through MEDLINE® or other publication databases of the National Library of Medicine®:


America's employer-sponsored health insurance system continues to provide health insurance to the vast majority of workers and their families. However, significant numbers of Americans lack access to coverage through the employment-based system. Ongoing increases in health care costs, the attendant increases in health insurance premiums, and the rise in employee contributions for coverage pose challenges for employers who are deciding whether to provide such coverage and for employees who are considering whether to enroll in coverage offered by their employers. The information presented here highlights some of these challenges and is intended to provide employers and policymakers with objective information to use as they address these difficult problems. Work on these issues is ongoing, however, and additional information will continue to be released as it becomes available.


  1.  *Eisenberg J, Power EJ. Transforming insurance coverage in quality health care. Voltage drops from potential to delivered quality. JAMA 2000; 284:2100-7.
  2.  *Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality. MEPS HC-064: 2003 P7R3/P8R1 Population Characteristics. Web site:
  3.  *Bernard D, Selden TM. Employer offers, private coverage, and the tax subsidy for health insurance:1987-1996. Int J Health Care Finance Econ 2003; 2:297-318.
  4. *Monheit AC, Nichols LM, Selden TM. How are net health insurance benefits distributed in the employment-related insurance market? Inquiry Winter 1995-96; 32(4):379-91.
  5.  Lee J. Are health insurance premiums higher for small firms? The Synthesis Project. The Robert Wood Johnson Foundation; Sept 2002. Research Synthesis Report No. 2.
  6.  *Cooper PF, Schone BS. More offers, fewer takers for employment-based health insurance: 1987 and 1996. Health Aff 1997; 16(6):142-9.
  7.  Institute of Medicine, Board on Health Care Services, Committee on the Consequences of Uninsurance. Hidden costs, values lost: uninsurance in America. Washington (DC): The National Academies Press; 2003. Web site: Accessed Aug 4, 2004.
  8.  *Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality. MEPSnet/IC [online database]. Web site: Accessed Jan 21, 2004.
  9.  Cutler D. Employee costs and the decline in health insurance coverage. In D Cutler and A Garber, eds. Frontiers in health policy research, volume 6. Cambridge (MA): MIT Press; 2003.
  10.  *Sommers J. Employer-sponsored health insurance for large employers in the private sector, by industry classification: 2000. Rockville (MD): Agency for Healthcare Research and Quality; 2003. Statistical Brief No. 47. Web site: Accessed Aug 4, 2004.
  11.  *Sommers J. Employer-sponsored health insurance for small employers in the private sector, by industry classification: 2002. Rockville (MD): Agency for Healthcare Research and Quality; 2003. Statistical Brief No. 46. Accessed Aug 4, 2004.
  12.  Glied S, Lambrew JM, Little S. The growing share of uninsured workers employed by large firms. Commonwealth Fund; Oct 2003.
  13.  *Cooper PF, Vistnes J. Workers' decisions to take-up offered health insurance coverage: assessing the importance of out-of-pocket premium costs. Med Care 2003; 41(7 Suppl):III35-43.
  14.  *Branscome JM. Employer-sponsored single, employee-plus-one, and family health insurance coverage: selection and cost, 2001. Rockville (MD): Agency for Healthcare Research and Quality; 2004. Statistical Brief No. 50. Web site: Accessed Aug 4, 2004.
  15.  *Blumberg LJ, Nichols LM, Banthin JS. Worker decisions to purchase health insurance. Int J Health Care Finance Econ 2001; 1:305-25.
  16.  *Hudson J. The uninsured in America, first half of 2002: estimates of the uninsured living in working families for the civilian noninstitutionalized population under age 65. Rockville (MD): Agency for Healthcare Research and Quality; 2004. Statistical Brief No. 39. Web site: Accessed Aug 4, 2004.
  17.  *Monheit AC, Vistnes JP. Race/ethnicity and health insurance status: 1987 and 1996. Med Care Research Rev 2000; 57(1 Suppl):11-35.
  18.  *Rhoades J, Chu M. Health insurance status of the civilian noninstitutionalized population: 1999. Rockville (MD): Agency for Healthcare Research and Quality; 2000. MEPS Research Findings No. 14. AHRQ Pub. No. 01-0011. Web site:
  19.  *Center for Financing, Access, and Cost Trends, Agency for Healthcare Research and Quality. MEPS HC-064: 2003 P7R3/P8R1 Population characteristics. Web site: Accessed Aug 24, 2004.
  20.  Collins SR, Schoen C, Tenney K, et al. Rite of passage? Why young adults become uninsured and how new policies can help. Commonwealth Fund; May 2003. Issue Brief. Web site: Accessed Oct 8, 2003.
  21.  *Monheit AC, Vistnes JP, Eisenberg JM. Moving to Medicare: trends in the health insurance status of near-elderly workers, 1987-1996. Health Aff 2001; 20(2):204-13.
  22.  *Sommers J. Employer sponsored health insurance for full time employees and retirees: 1998 and 2000. Rockville (MD): Agency for Healthcare Research and Quality; 2003. Statistical Brief No. 17. Accessed Aug 4, 2004.
  23.  *Monheit AC, Schone BS. How has small group market reform affected employee health insurance coverage? J Public Econ 2004; 88:237-54.

 *AHRQ-funded/sponsored research.

Page last reviewed September 2004
Internet Citation: Employer-Sponsored Health Insurance: Trends in Cost and Access. September 2004. Agency for Healthcare Research and Quality, Rockville, MD.