State Insurance Regulation and Employers' Decisions to Self-Insure

Article excerpt

Introduction

A major change in the provision of private health insurance for Americans occurred during the 1980s. Employers who self-insure their health plans, rather than purchase insurance from Blue Cross and Blue Shield (BCBS) or commercial insurers, became the leading underwriters of group coverage in the United States, accounting for 58 percent of all group health premiums by the close of the decade. Yet in 1980, self-insured plans accounted for less than a quarter of all group health premiums (Health Insurance Association of America, 1992).

Some observers attribute employers' conversions to self-insurance to the preemption requirements of the federal 1974 Employee Retirement Income Security Act (ERISA). Beginning in 1977 and culminating with a 1985 Supreme Court decision, federal judges have held that states can regulate and tax insurance sold by BCBS and commercial insurers but not employers' self-insured plans (Fox and Schaffer, 1989). In interpreting ERISA's preemption in this way, the courts effectively created incentives for firms to self-insure their health plans. By self-insuring, a firm escapes compliance with state-mandated benefit requirements and can avoid paying state premium taxes. All firms, regardless of their size or industry, are covered by ERISA's exemption from state insurance regulation if they self-insure (Employee Benefit Research Institute, 1993).

Coinciding with the trend toward self-insurance has been a proliferation of state-mandated benefits and increased state insurance taxation. State-mandated benefits are laws that prescribe the terms of coverage for group insurance purchased from BCBS and commercial insurers. Such laws include requirements that plans cover specific services, categories of providers, diseases, or persons who might otherwise have difficulty obtaining coverage. Between 1980 and 1990, the number of state mandates (in place across all the states) almost doubled, from 450 to 854 (Mandated Benefits Manual, 1992). State taxes on the plans sold by these insurers also increased slightly over the decade, largely to help finance new state risk pools for persons otherwise deemed uninsurable (Laudicina, 1988).

The main purpose of this article is to examine the relationship between the growth of state mandates, insurance taxation, and the decisions of firms to convert to self-insurance as a way to provide health benefits. The fact that all three have increased together is consistent with, though not sufficient evidence of, the premise that the growth of state-mandated benefits and insurance taxation fueled the self-insurance movement of the 1980s. Using data from several nationwide employer surveys conducted over the 1980s, we formally test the hypothesis that mandates and state taxation indeed motivated employers to convert from purchased insurance to self-insurance.

Understanding the motives for self-insurance is important for two reasons. First, many mandated health coverages add substantially to employers' costs for insurance benefits (Jensen and Morrisey, 1990; Frank, Salkever, and Sharfstein, 1991; Dyckman and Anderson-Johnson, 1989). If increased foreign competition in the 1990s pressures U.S. firms to increase their efforts to control benefit costs, as suggested by some, then firms that have not already converted to self-insurance may decide to do so. Self-insurance for these mostly medium and small businesses could be risky. While stop-loss coverage is available to most employers, some firms might choose to forego it. Thus, a continuation of the self-insurance trend as a way to avoid costly state regulations could result in increased financial risk for some firms.

Second, self-insurance raises an issue of efficiency in the provision of employer-sponsored insurance. Some research suggests that employers' benefit and administrative costs for self-insured plans are actually higher than those of purchased plans containing the same coverage (Jensen and Gabel, 1988; Jensen and Morrisey, 1990). …