Po l i c y m a k e r s a t both the federal

Transcription

1 Recent Trends In Self-Insured Employer Health Plans A look at the interplay among market forces, regulation, and employers decisions to self-insure. b y M. S u s an M ar q u i s a n d S t e ph e n H. L o n g Po l i c y m a k e r s a t both the federal and state levels are interested in knowing how many private employer health plans are self-insured and in identifying forces that affect this number. These trends are important because fully insured plans and selfinsured plans are subject to different regulatory regimes. 2 State insurance laws, which are traditionally far more comprehensive, cover only fully insured products purchased from insurers. In contrast, federal laws cover both types of plans, but with comparatively limited provisions. Recently, however, federal regulation of health coverage has been extended to cover new dimensions. The ultimate reach of regulatory provisions in affecting enrolled workers and their dependents, therefore, depends on the extent of self-insurance among employers. Numerous provisions of state insurance regulation may cause an increase in the propensity of employers to self-insure. State regulations frequently mandate that certain benefits be covered in all policies issued by health insurers, potentially adding to the cost of fully insured plans. Other regulations that may add to the relative cost of fully insured plans are continuation of coverage and rate regulation in the small-group market and limits on preexisting conditions. 5 States levy taxes on the premium revenues of insurance companies, which also raise the cost of fully insured plans. 6 Employers whose workplaces extend across many states may object to the added complexity or inequity among their workers as a result of having to offer different benefits in different states. 7 Although the trend over this decade toward increasing state regulation of the small-group market was thought to add to the incentive to self-insure, the Health Insurance Portability and Accountability Act (HIPAA) of 996 included some provisions that apply to both fully and self-insured plans, thereby contributing to a more level playing field. Recent dynamics in the health insurance market also may be leading to changes in selfinsurance. The large cost increases of the late 90s and early 990s created strong pressure for cost savings. 9 There were rapid changes in the types of products available to employers especially in the availability of health maintenance organizations (HMOs) to employers of all sizes. 0 Small employers had improved access to HMOs. For large employers looking to control costs, this trend first provided new, fully insured alternatives to selfinsuring, but very recently there has been talk of self-insured employers being able to purchase access to HMOs while remaining selfinsured a phenomenon often termed selfinsured HMOs. In this paper we analyze data from two large surveys of employers about their health insurance benefit offerings. The data provide evidence on recent trends in employers propensity to self-insure and employees propensity to enroll in self-insured plans. We examine some possible causes of these trends. 6 Susan Marquis and Steve Long are senior economists at RAND in Washington, D.C. H E A L T H A F F A I R S ~ M a y / J u n e ThePeople-to-People Downloaded from Health HealthAffairs.org Foundation, on Inc. February 2, 20.

2 62 METHODS We compared estimates from the and Robert Wood Johnson Foundation (RWJF) Employer Health Insurance Surveys to measure recent trends in self-insured plans. The survey interviewed 22,7 private employers in ten states. The survey interviewed 2,55 private employers nationwide. The sample was concentrated in the sixty communities followed by the RWJF Community Tracking Study and in twelve states having significant small-group rating reforms for research on this state intervention. These cases are supplemented by a sample from the remainder of the continental United States to better represent the nation s business establishments. 2 For both surveys the sampling frame was the Duns Market Identifiers national census of employment establishments. Within the geographic units described above, the samples were further allocated to strata defined by the number of workers at the establishment. In the response rate was 7 percent, whereas in it was 60 percent. The sample establishments have been weighted for these analyses to account for different probabilities of selection and nonresponse. The weighted sample represents all private employment establishments with at least one employee. We used sampling weights to make estimates for three different units of observation: establishments, employees, and insurance plan enrollees. We examined trends in employers decisions to self-insure by comparing rates in the seven states that are represented in both surveys: Colorado, Florida, Minnesota, New York, Oregon, Vermont, and Washington. The survey included 5,0 employers in these seven states, and the survey included,7 employers. We also examined several hypotheses about correlates of the self-insurance decision using the full national sample from the survey. The surveys, based on similar questionnaires and definitions, used computerassisted telephone interviews averaging thirty minutes each to collect the data. The interviews were conducted with the person or persons in each establishment most knowledgeable about health benefits and firm and worker characteristics. A number of specialized concepts and definitions were used in this work. We based our definition of self-insured plans on the respondent s self-assessment. In the question asked about self-insured status was, Are you self-insured, or are you fully insured? In the question was, Is this a plan that is purchased from an insurance company or HMO, or is it a self-insured plan? Further explanation of this distinction, involving whether the firm or the insurer bears financial responsibility for workers medical claims, was provided in both surveys if needed. A self-insured employer was defined as one that offered one or more self-insured plans. Classification of a plan as an HMO also was based on the respondent s self-assessment, aided by a complete definition of plan types if needed. Because self-insurance status and plan type were collected independently, we were able to identify self-insured HMOs. Although the sample and analysis unit is the establishment (a physical location of business), for many analyses we categorized establishments according to the size of the firm, which includes employees at all locations nationwide. This is because insurance decisions, including the decision to self-insure, typically are made at the regional or national level in firms with several establishments. As a measure of the extent of a firm s multistate business operation, we used the number of the firm s employees working in the same state as the sampled establishment, divided by the number of the firm s employees nationwide. 5 We classified establishments according to regulatory and market characteristics of their location. The degree of small-group reform in the state was measured as the interaction between whether all insurance products must be guarantee-issued and how tightly premium rates are regulated, if at all. We defined three classes of reform: high (all products guaranteeissued; rates cannot be based on health status, H E A L T H A F F A I R S ~ V o l u m e, N u m b e r Downloaded from HealthAffairs.org on February 2, 20.

3 with limited or no age variation allowed), medium (all products guarantee-issued; rates cannot be based on health status, with no or limited restrictions on age variation), or low (fewer or no restrictions). 6 HMO penetration was measured for the sixty Tracking study sites and is based on tabulations from the employer survey data of the share of employees enrolled in their employer-sponsored plan who were in an HMO. It is defined as low if the community is in the bottom quartile and high if in the top quartile. Finally, we measured both the current health insurance premium level and the change in the premium from the previous year. The current premium level is based on plan-specific data, reported separately for individual and family coverage. For establishments offering multiple plans, we computed an average premium for the establishment by weighting the premiums for the separate plans by the number of enrollees. The latter measure is based on an establishmentlevel question about change in total health insurance cost per enrollee compared with the previous plan year. RESULTS Between and self-insurance declined in the seven states studied. The percentage of self-insured establishments fell in all size groups (Exhibit ), and the decline ranged from 67 percent among small employers to percent among large employers. Overall, in the seven states the number of selfinsured employers fell from 9 percent to percent. Among employees enrolled in their employer s group plan, the number enrolled in a self-insured plan fell percent. The likelihood of being in a self-insured plan rises strongly as firm size increases. The decline in self-insurance appears to be related to rapid change in market forces during this period. Foremost was a very strong shift toward employers offering managed care plans. The percentage of employers offering only fully insured HMOs soared from 2 percent to 5 percent in the four-year period in the seven states (Exhibit 2). 7 Counting employees facing this type of offer, rather than employers, yields a similar conclusion. The shift to managed care came at the expense of nearly all other combinations of plan offerings and occurred among employers of all sizes but was especially marked among workers in firms with fewer than 00 workers, which are shown separately in Exhibit 2. Self-insured HMOs represented a small part of the picture (2 percent of establishments and 6 9 percent of employees). Hence, if there is to be a resurgence of selfinsurance through the vehicle of self-insured HMOs, it has yet to emerge. The strong move to HMOs appears to have been driven, at least in part, by cost considerations. In U.S. establishments offering an HMO had average premiums across all of their plans that were 7 percent lower for single coverage ($6 versus $76 $77) and 7 percent lower for family coverage than were 6 EXHIBIT Self-Insurance In Seven States, By Firm Size, And Percent self-insured establishments Percent of enrollees in self-insured plans Firm size Fewer than 00 employees 9% employees % or more employees 6% 56 SOURCE: and Robert Wood Johnson Foundation Employer Health Insurance Surveys. NOTE: The seven states are Colorado, Florida, Minnesota, New York, Oregon, Vermont, and Washington All 9% 0 H E A L T H A F F A I R S ~ M a y / J u n e Downloaded from HealthAffairs.org on February 2, 20.

4 EXHIBIT 2 Types of Health Plans Offered In Seven States, And Percent of establishments offering insurance All establishments Establishments with fewer than 00 employees Any self-insured plan Any Other Fully insured plans only self-insured self-insured HMO Non-HMO HMO and HMO plan only only only other plan 2% 2 7% 2 2% % % Percent of employees offered insurance All establishments Establishments with fewer than 00 employees 6 9 SOURCE: and Robert Wood Johnson Foundation Employer Health Insurance Surveys. NOTES: The seven states are Colorado, Florida, Minnesota, New York, Oregon, Vermont, and Washington. HMO is health maintenance organization premiums in establishments offering only other types of plans, whether self- or fully insured (Exhibit ). Similarly, total premium costs rose less between 996 and for establishments offering an HMO than for others, whether self-insured or fully insured. Opportunity to take advantage of HMOs also affects the likelihood of self-insuring. In markets with high HMO penetration only percent of employers self-insure, compared with 9 percent in markets with low HMO penetration (Exhibit ). This differential prevails among all firm sizes except the smallest employers. Tight small-group regulations do not appear to cause an increase in self-insurance among small employers. The share of small establishments that self-insure is fairly constant, regardless of whether they are located in states with low, medium, or high degrees of small-group reform (Exhibit ). Moreover, the absence of an upward trend in the number of small employers that self-insured over the study period, during which these reforms were legislated and implemented, further supports the conclusion that small-group reforms have not led to an upsurge in self-insuring by small employers (see Exhibit ). States with strict rating reforms provide small employers some longer-run protection against price in- EXHIBIT Premiums And One-Year Change In Premiums, By Type Of Plan Offered, Nationwide, Plan offered Any HMO Non-HMO Any self-insured Fully insured only Monthly premium Single $ Family $ 2 One-year change.6% SOURCE: Robert Wood Johnson Foundation Employer Health Insurance Survey. NOTES: Results are establishment-weighted for employers in all states. HMO is health maintenance organization H E A L T H A F F A I R S ~ V o l u m e, N u m b e r Downloaded from HealthAffairs.org on February 2, 20.

5 EXHIBIT Percentage Of Establishments That Self-Insure, By Firm Size And Characteristics Of Establishment s Location, Nationwide, HMO penetration a Low High Small-group insurance reforms b Low Medium High Share of business operation in state Less than one-third One-third to two-thirds Two-thirds to almost all All Firm size (number of workers) Fewer than or more All % All employers d SOURCE: Robert Wood Johnson Foundation Employer Health Insurance Survey. NOTES: Establishment-weighted results for employers nationwide. HMO is health maintenance organization. a Low is lowest quartile of market areas. High is highest quartile of market areas. Results are from the sixty Community Tracking Study sites only and so do not match the national estimates presented in this and other exhibits. b Low reform if few or no restrictions. Medium reform if there is guaranteed issue of all plans and moderate rate regulation (Florida, Kentucky, New Hampshire, Oregon, Vermont, Washington); moderate rate regulation if health status is not allowed, but there is no or limited restriction on age variation. High reform if guaranteed issue of all plans and tight rate regulation (Maine, Maryland, Massachusetts, New Jersey, New York); tight rate regulation if health status is not allowed and there is limited or no age variation. c Not applicable; the small-group insurance reforms apply only to plans purchased by small employers. d The corresponding percentages of self-insured plan enrollees are four, eight, twenty-nine, and sixty for the firm-size categories and thirty-nine for all establishments. 5% 6 7 % % % creases in purchased plans when a group member becomes ill. The advent of these provisions over the study period may have been a factor in offsetting shifts to self-insurance that otherwise might have occurred. In contrast, the long-standing observation that multistate employers are more likely to self-insure is borne out in our data. Clearly, most multistate firms are large employers. Controlling for the effect of firm size on selfinsuring by examining firms with more than 500 employees, 5 percent of establishments in firms with all of their employees in a single state self-insure, compared with 6 percent of establishments in firms with fewer than onethird of their employees in the same state. The r ed uc tio n in self-insured establishments that occurred in seven states between and suggests that self-insurance behavior is indeed dynamic. Our results suggest that market change may have been more important than regulatory change as a determinant of self-insurance decisions. This topic deserves to be monitored closely, however, in the face of continuing rapid market and regulatory change. This research was supported by Grants no and 0565 from the Robert Wood Johnson Foundation (RWJF) and by Contract no. J-9-P from the Pension and Welfare Benefits Administration, U.S. Department of Labor (DOL). Any views expressed herein are solely those of the authors, and no endorsement by the RWJF, the DOL, or RAND is intended or should be inferred. The authors thank Linda Andrews and Roald Euller for their efforts in preparing the survey data on which this paper is based. H E A L T H A F F A I R S ~ M a y / J u n e Downloaded from HealthAffairs.org on February 2, 20.

6 66 NOTES. U.S. General Accounting Office, Employer-Based Health Plans: Issues, Trends, and Challenges Posed by ERISA ( Washington: GAO, 995); and C. Copeland and B. Pierron, Implications of ERISA for Health Benefits and the Number of Self-Funded ERISA Plans (Washington: Employee Benefit Research Institute, 99). 2. Copeland and Pierron, Implications of ERISA; P. Butler and K. Polzer, Private-Sector Health Coverage: Variation in Consumer Protections under ERISA and State Law (Washington: George Washington University, 996); and G. Acs et al., Self-Insured Employer Health Plans: Prevalence, Profile, Provisions, and Premiums, Health Affairs (Summer 996): B. Atchinson and D. Fox, The Politics of the Health Insurance Portability and Accountability Act, Health Affairs (May/June ): 6 50; and L.M. Nichols and L.J. Blumberg, A Different Kind of New Federalism? The Health Insurance Portability and Accountability Act of 996, Health Affairs (May/June 99): J. Gabel and G. Jensen, The Price of State Mandated Benefits, Inquiry (Winter 99): 9 ; S. Garfinkel, Self-Insuring Employee Health Benefits, Medical Care Research and Review (December 995): 75 9; J. Gruber, State- Mandated Benefits and Employer-Provided Health Insurance, Journal of Public Economics (June 99): 6; and G. Jensen, K. Cotter, and M. Morrisey, State Insurance Regulations and Employers Decisions to Self-Insure, Journal of Risk and Insurance (June 995): An Analysis of Mandated Community Rating (Washington: American Academy of Actuaries, ); GAO, Employer-Based Health Plans and Health Insurance Regulation, Varying State Requirements Affect Cost of Insurance (Washington: GAO, 996); and M.A. Morrisey and G.A. Jensen, State Small-Group Insurance Reform, in Health Policy, Federalism, and the American States, ed. R.F. Rich and W.D. White (Washington: Urban Institute Press, 996). 6. Gabel and Jensen, The Price of State Mandated Benefits; GAO, Health Insurance Regulation; and Jensen et al., State Insurance Regulations. 7. M. Schachner, Growth in Health Care Self- Funding Slows: Few Additional Employers Likely to Self-Insure, Business Insurance, 25 January,.. L.J. Blumberg and L.M. Nichols, First, Do No Harm: Developing Health Insurance Market Reform Packages, Health Affairs (Fall 996): K. Levit et al., National Health Expenditures in : More Slow Growth, Health Affairs (November/December 99): J.R. Gabel, P.B. Ginsburg, and K.A. Hunt, Small Employers and Their Health Benefits, 9 996: An Awkward Adolescence, Health Affairs (September/October ): The ten states are Colorado, Florida, Minnesota, New Mexico, New York, North Dakota, Oklahoma, Oregon, Vermont, and Washington. For a description of the survey, see J. Cantor, S.H. Long, and M.S. Marquis, Private Employment- Based Health Insurance in Ten States, Health Affairs (Summer 995): The survey methods are described in Robert Wood Johnson Foundation Employer Survey: Data Collection Methodology Report (Rockville, Md.: Westat, Inc., 995). 2. See Employer Health InsuranceSurvey: FinalMethodology Report (Research Triangle Park, N.C.: Research Triangle Institute, 99); and P. Kemper et al., The Design of the Community Tracking Study: A Longitudinal Study of Health System Change and Its Effects on People, Inquiry (Summer 996): The twelve states are California, Colorado, Connecticut, Florida, Maryland, Massachusetts, Minnesota, New Jersey, New York, Oregon, Vermont, and Washington.. The exceptions to this are two states in for which the sampling frame was all employers in the state unemployment insurance files.. We did not identify products for which employers self-insure some of the expenses (such as outof-network use) that is, the employer made the judgment whether to classify the plan either as self-insured or fully insured. 5. Ideally, we would prefer a measure reflecting the number of different states in which the firm operates, but the measure we used was the best available to us, and we expect that it is correlated with the former. 6. Based on a review of legislation in all states for the period 990 by the Institute for Health Policy Solutions. See R. Curtis et al., Health Insurance Reform in the Small-Group Market, Health Affairs (May/June 999): See notes to Exhibit for a list of the states in the high and medium groups. 7. About 20 percent of employers classified as selfinsured in Exhibit 2 also offer fully insured products; about 0 percent of employees in selfinsured firms are offered both self-insured and fully insured products. See Exhibit for the numbers of employees enrolling in self-insured versus fully insured plans.. Because we computed an average establishment premium, the cost comparison between establishments that offer HMOs and those that do not is not biased by enrollee selection effects among types of plans within the establishment s offerings. However, the cost differences may reflect differences in the risk characteristics of establishments that do and do not offer HMOs. H E A L T H A F F A I R S ~ V o l u m e, N u m b e r Downloaded from HealthAffairs.org on February 2, 20.