Health Affairs, 22, no. 2 (2003): 202-210 doi: 10.1377/hlthaff.22.2.202 © 2003 by Project HOPE	New Online   Pay Cuts For Medicare Docs   Access To Care Woes   Public Coverage More Efficient   Empowering Consumers

This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Gabel, J. R. Articles by Hawkins, S. Search for Related Content PubMed PubMed Citation Articles by Gabel, J. R. Articles by Hawkins, S. Related Collections Insurance Coverage Insurance - Employer-Based System Managed Care

Health Tracking

MARKETWATCH

Self-Insurance In Times Of Growing And Retreating Managed Care

Abstract

 
This paper examines trends in self-insurance and in the content of self-insured plans from 1993 to 2001. The percentage of employees enrolled in self-insured plans fell during these years. Much of the decrease was attributable to the decline of indemnity insurance and the rise of HMO and point-of-service plan enrollment. If the product mix had remained constant throughout these years, self-insured enrollment would have grown between 1993 and 1996 and then declined to its current 50 percent level. As a result of the Health Insurance Portability and Accountability Act (HIPAA), the use of preexisting condition clauses declined dramatically in self-insured plans. Self-insured and purchased plans cost similar amounts and provide similar benefits. Cost sharing is somewhat lower in self-insured PPO plans. During periods of rapid inflation, premiums increase more slowly for self-insured than for fully insured plans.

The second historical accident promoted the growth of self-insured plans. Employers that opt to self-insure are not subject to state financial reserve requirements to minimize the risk of insolvency; state-imposed premium taxes to finance state guaranty funds to pay claims for insolvent plans; state charges to finance high-risk pools that provide coverage for uninsurable people; state-mandated benefit requirements to ensure coverage for essential services; various consumer protection laws; or state insurance reforms intended to minimize harsh medical underwriting.³

Concerns about self-insurance. Critics of self-insurance cite three concerns about the growth of self-insured plans.⁴ First is the issue of equity to employers. ERISA has unintentionally yielded two separate (and unequal) regulatory systems. Second, employees in self-insured plans lack the same consumer protections that employees in purchased plans have with regard to plan insolvency, fraudulent plan behavior, and refusal to pay for treatments and services promised by the employer. Third, because roughly half of U.S. workers are enrolled in a self-insured plan, states find it extraordinarily difficult to assume their role as "the nation’s laboratories" in designing health policy. This limitation applies to a host of critical issues: cost containment, programs for the uninsured, state insurance reform, and consumer protection laws.

Difficulties in observing trends. Self-insurance is not a phenomenon unique to the 1990s. Patricia McDonnell and her colleagues reported from an analysis of a survey of employers in 1984 that a majority of employees covered by job-based insurance were enrolled in a self-insured plan.⁵ Because different surveys use different methods to estimate the extent of self-insurance, it is difficult to observe a trend over the past two decades. The National Medical Expenditure Survey (NMES) estimated that 42 percent of employees were in self-insured plans in 1987.⁶ Susan Marquis and Steve Long estimated that 40 percent of employees were enrolled in such a plan in 1993 and that 33 percent were in 1997.⁷

Impact of legislation. Recent national legislation may have removed some of the advantages of self-insurance and thereby encouraged a shift back to fully insured plans. The Health Insurance Portability and Accountability Act (HIPAA) limited the use of preexisting condition clauses for both fully and self-insured plans.⁸ The Mental Health Parity Act and the Newborns and Mothers Protection Act mandated specific benefits for both types of plans.⁹ Congress enacted all of this legislation in 1996.

State legislation in the 1990s, in contrast, may have encouraged more firms to self-insure. Twenty-one states passed legislation setting minimum length-of-stay standards for inpatient care of mastectomy patients. Thirty-four states passed laws mandating direct access to obstetricians/gynecologists (OB/GYNs), while fourteen states mandated direct access to chiropractors, dermatologists, or eye care providers. Eleven states passed any-willing-provider laws that require insurers to contract with any provider willing to meet the terms of the agreement.¹⁰

Impediments to self-insurance. In tracking trends in self-insurance, analysts need to be aware that a changing insurance product mix can obscure underlying trends in self-insurance. Indemnity plans face comparatively few impediments to self-insurance. The employer simply assumes the financial risk for payment of medical claims. More commonly, the employer purchases stop-loss insurance whereby a reinsurer assumes the financial risk once aggregate medical claims exceed a given threshold or claims for an individual exceed a predetermined level. The employer can rent the services of a utilization management company in an attempt to limit inappropriate hospitalizations and other high-cost services.

There are greater impediments to self-insuring a health maintenance organization (HMO) plan. In addition to assuming financial risk, the employer needs to rent a network of providers, a quality assurance program, and a more complex set of utilization management and disease management programs. In some states employers are prohibited from contracting with providers on a capitated basis.

From 1993 to 1996 employer-based insurance completed the rapid transformation from indemnity to "managed care heavy." In 1993, 49 percent of employees were enrolled in indemnity plans, 22 percent in HMOs, 20 percent in preferred provider organizations (PPOs), and 9 percent in point-of-service (POS) plans. By 1996 a plurality of U.S. workers were in HMOs, and indemnity plans accounted for only 26 percent. From 1996 to 2001 employers and employees retreated from heavily managed care and moved toward hybrid-type managed care plans. PPO enrollment rose from 28 percent to 48 percent of employees, HMO enrollment fell from 31 percent to 23 percent, while enrollment in indemnity plans fell to just 7 percent, and enrollment in POS plans rose from 14 percent to 22 percent.¹¹

Knowledge gaps. Because self-insured plans are preempted from state-mandated benefits and insurance reforms, some observers have questioned whether there are differences in plan design among fully and self-insured plans. Are self-insured plans a safe haven for low-benefit plans and more onerous underwriting activities? A previous analysis of the 1993 Robert Wood Johnson Foundation (RWJF) Employer Health Insurance Survey by Gregory Acs and colleagues reported that self-insured plans have more preexisting condition clauses, more mental health/substance abuse benefits, more maternity and drug benefits, lower deductibles, and similar premiums, compared with purchased plans.¹² More recently, Marquis and Long concluded that the cost of coverage for self-and fully insured plans was similar in 1997.¹³ Others have asked if there has been a differing trend in the content of self-and fully insured plans in recent years. There has been no comprehensive analysis of these issues for 1998–2001.

This paper has three objectives. First, based on surveys with consistent questionnaires and methods, we track trends in self-insurance over the eight-year period 1993–2001. Second, we investigate the role of changing product mix for self-insurance trends over this period. Third, we examine the changes in the content of self-and fully insured plans during the study period. In addressing these issues, the paper updates findings from earlier studies, none of which uses post-1997 data. The paper also provides a preliminary glimpse of the impact of the 1996 federal legislation—HIPAA, the Mental Health Parity Act, and the Newborns and Mothers Protection Act—on the content and prevalence of self-insured plans.

Study Methods

Top Study Methods Study Findings Discussion NOTES

 
Data on employer-based health insurance are from the national surveys conducted by KPMG Peat Marwick in 1993 and 1996 and the Henry J. Kaiser Family Foundation (KFF)/Health Research and Educational Trust (HRET) in 1999 and 2001. The core questions for these surveys are nearly identical throughout the study years. They include (1) the definition of self-insurance; (2) the definition of indemnity, HMO, PPO, and POS plans; (3) the annual increase in the cost of coverage for each type of health plan; (4) estimated enrollment in each of the plan types; and (5) the cost of single and family coverage in each health plan.

The survey’s questionnaire each year asked a series of questions about the firm’s largest indemnity, HMO, PPO, and POS plan. With the exception of 1996, the Washington-based survey firm National Research LLP conducted telephone interviews with employee benefit managers. The unit of analysis on all surveys is the firm rather than the establishment.¹⁴

Dun and Bradstreet provided the sample frame for the list of U.S. public and private employers that employ three or more workers. Firms were randomly selected from a stratified random sample, where firm size and industry defined the sampling cells. Sample sizes ranged from 1,953 in 1993 to 2,001 in 1997. Response rates ranged between 50 percent in 1996 and 60 percent in 1999.

Because firms were randomly selected, it is possible to use statistical weights and thereby calculate national, firm-size, industry, and regional averages. All data reported here are employee weighted (unless specifically noted) to be nationally representative of U.S. workers with employer-sponsored health benefits.

In calculating standard errors, we used SUDAAN to make adjustments for design effects. We use a 95 percent confidence interval and compare whether the individual figure is statistically different from the previous year for the category in each exhibit.

A fundamental concern of our analysis is the role of plan mix over the study period. Hypothetically, the percentage of workers covered by self-insured plans could increase for each type of plan over the study period and yet the total percentage of workers in self-insured plans could decline. This paradox could result from a dramatic shift in enrollment from indemnity plans, which tend to be self-insured, to HMO plans, which tend to be fully insured. Hence, our analysis addresses what trends in health insurance would have been if the mix of plans had remained unchanged over the study years, rather than shifting from conventional to other types of plans.

Study Findings

Top Study Methods Study Findings Discussion NOTES

 
Trends in self-insurance. The percentage of insured workers covered through self-insurance in firms with three or more workers declined slightly over the study years (Exhibit 1). Fifty-five percent of workers were enrolled in self-insured plans in 1993; that number grew to 57 percent in 1996 and then fell to 50 percent in 2001.

Trends by firm size. A few notable trends and patterns stand out in self-insurance from 1993 to 2001 by firm size (Exhibit 2). Throughout the study years, rates of self-insurance have remained relatively stable for firms with 1,000 or more workers. In fact, changes in product mix have counterbalanced trends in self-insurance by firm size and plan type. For example, self-insurance has grown markedly among HMO plans for firms employing 5,000 or more workers. Yet the decline in HMO market share since 1996 has offset this increase, so that the overall percentage of self-insured workers in the largest firms dropped slightly.

Cost of coverage. Premiums increased more rapidly among fully insured than self-insured plans over the study period, with the single exception of 1996 among HMOs (Exhibit 3). In 1996 the average premium change was actually negative for fully insured HMO plans. The health insurance underwriting cycle in 1996 had reached its "soft" phase. Six consecutive years of underwriting profits (profits before investment income) from 1989 to 1995 had induced the entry of insurers into new geographic markets. During this period health insurers viewed HMOs and POS plans as their growth products. They fought for market share by underpricing their competitors, and this fierce price competition drove premiums so low that by 1997, 68 percent of U.S. HMOs were suffering underwriting losses.¹⁵

Cost of single coverage. Yet the cost of single coverage in any given year was statistically equivalent among self-insured and fully insured plans. Presumably, the changing mix of employers purchasing different products and conversions to self-insurance explain this seeming paradox. For some years HMO and POS self-insured plans were more expensive than fully insured plans were. As we show in the subsequent section, these differences may be attributable to more generous coverage within self-insured plans—that is, lower cost sharing or a richer set of covered benefits or both.

Richness of benefits and patient cost sharing. We compared deductibles for each of the four study years by plan type among self-and fully insured plans (Exhibit 4). In 2001 deductibles in PPO plans were 39 percent higher in fully than self-insured plans for use of in-network providers and 18 percent higher for use of out-of-network providers. There were no statistically significant differences among fully and self-insured conventional plans in any year, and there was only one year in which differences were statistically significant for POS plans. Per visit copayments within HMO plans did not vary between purchased and self-insured products over the study years (data not shown).

Discussion

Top Study Methods Study Findings Discussion NOTES

 
Comparisons with previous studies. Several of our findings echo those of Acs and colleagues’ study of self-insurance in 1993 and Marquis and Long’s study of self-insured plans in 1997. However, there are notable differences.

We find that the absolute percentage of employees covered by self-insured plans declined from 56 percent to 50 percent during the study years. This decline is largely attributable to the change in product mix, as employers switched from indemnity plans, which lend themselves to self-insurance, to HMO and POS plans, which face more impediments to self-insurance. Indemnity and PPO plans were less likely, but HMO plans more likely, to be self-insured at the end of the study period than at the beginning. If the product mix throughout the period had held constant as it was in 1996, the high-water mark of heavily managed care, the percentage of workers covered by self-insured plans would have grown from 46 percent in 1993 to 57 percent in 1996 and then declined to its current 50 percent level.

As reported by Acs and colleagues and by Marquis and Long, we find that the average costs of self-and fully insured plans are roughly equivalent. However, self-insured plans have somewhat lower deductibles. Hence, self-insured plans are not a safe haven for employers seeking to offer low-benefit catastrophic coverage to their workforce. One dramatic change has occurred among self-insured firms in recent years: In response to HIPAA, the use of preexisting condition clauses has declined rapidly—more rapidly than among fully insured plans—so that self-insured plans today are no more likely to impose such clauses on their workers than fully insured plans are.

Impact of underwriting cycle. As the health insurance underwriting cycle moves from the "soft phase," characterized by insurer underwriting losses and fierce price competition, to the "hard phase," characterized by rapid inflation and restoration of insurers’ profitability, the comparative attractiveness of self-insurance changes. Relatively small differences in the rate of inflation between fully and self-insured plans become more dramatic. The countervailing force at the moment is a sluggish economy recovering from a recession. Will midsize firms opt to take on greater financial risk during a period when profits are declining?

Likely impact of shift in product mix. Lastly, with health insurers retreating from heavily managed care and seeking products that promote greater employee choice and cost sharing, HMO enrollment is likely to decline, while PPO enrollment is likely to increase. This shift in product mix will raise the number of employees enrolled in self-insured plans. Consequently, the two-decade-old controversy about the merits of ERISA and self-insurance is unlikely to fade away. Stay tuned for more debates at the state and federal levels.

Editor's Notes

 
Jon Gabel is vice-president, Health System Studies, at the Health Research and Educational Trust (HRET) in Washington, D.C. Gail Jensen is a professor of gerontology/economics at Wayne State University in Detroit. Samantha Hawkins is a research analyst at HRET.

The authors thank the Robert Wood Johnson Foundation for financial support through the Health Care Financing and Organization project, under Grant no. 39886.

NOTES

Top Study Methods Study Findings Discussion NOTES

 

1. P. Fronstin, Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 2001 Current Population Survey, Issue Brief no. 240 (Washington: Employee Benefit Research Institute, December 2001).
2. EBRI, Fundamentals of Employee Benefit Programs, 5th ed. (Washington: EBRI, 1997), 5–9.
3. G. Acs et al., "Self-Insured Employer Health Plans: Prevalence, Profile, Provisions, and Premiums," Health Affairs (Summer 1996): 266–278.
4. Readers interested in a more comprehensive discussion of the pros and cons of ERISA should refer to U.S. General Accounting Office, Report to Congressional Requesters, Employer-Based Health Plans: Issues, Trends, and Challenges Posed by ERISA, Pub. no. GA 1.13:HEHS-95-167 (Washington: GAO, 1995).
5. P. McDonnell et al., "Self-Insured Health Plans," Health Care Financing Review 8, no. 2 (1986): 1–16.
6. Reported in Acs et al., "Self-Insured Employer Health Plans."
7. M.S. Marquis and S.H. Long, "Recent Trends in Self-Insured Employer Health Plans," Health Affairs (May/June 1999): 161–166.
8. B. Atchinson and D. Fox, "The Politics of the Health Insurance Portability and Accountability Act," Health Affairs (May/June 1997): 146–150.
9. R. Frank, C. Koyanagi, and T. McGuire, "The Politics and Economics of Mental Health ‘Parity’ Laws," Health Affairs (July/Aug 1997): 108–119.
10. S. Laudicina and K. Pardo, State Legislative Health Care and Insurance Issues: 1999 Survey of Plans (Washington: Blue Cross and Blue Shield Association of America, December 1999).
11. J. Gabel et al., "Job-Based Health Insurance in 2001: Inflation Hits Double Digits, Managed Care Retreats," Health Affairs (Sep/Oct 2001): 180–186.
12. Acs and colleagues’ analysis also includes data from the 1991 Health Insurance Association of America survey of employers and the 1987 National Medical Expenditure Survey. See Acs et al., "Self-Insured Employer Health Plans."
13. Marquis and Long, "Recent Trends."
14. In 1993 KPMG did not ask employers with HMO plans about self-insurance status. Therefore, for HMO plans in 1993 we have used data from the RWJF Employer Health Insurance Survey.
15. InterStudy reported that 68 percent of HMOs suffered underwriting losses in 1997.
16. The exception to this statement is the premium for stop-loss protection paid to reinsurers.

                      What's this?

This article has been cited by other articles:

				M. M. Davis and K. Fant Coverage Of Vaccines In Private Health Plans: What Does The Public Prefer? Health Aff., May 1, 2005; 24(3): 770 - 779. [Abstract] [Full Text] [PDF]

				J. C. Robinson Consolidation And The Transformation Of Competition In Health Insurance Health Aff., November 1, 2004; 23(6): 11 - 24. [Abstract] [Full Text] [PDF]