Medigap Reform Legislation of 1990: Have the Objectives Been Met?

Abstract

The 1990 medigap reform legislation had multiple objectives: To simplify the insurance market in order to facilitate policy comparison, provide consumer choice, provide market stability, promote competition, and avoid adverse selection. Based on case study interviews with a cross-section of individuals and organizations, we report that most of these objectives have been achieved. Consumers of medigap plans are able to make more informed choices, largely because they can adequately compare policies based on standard benefits. Marketing abuses have apparently declined, as evidenced by a decrease in the number of consumer complaints. Finally, no major detrimental impact on the insurance industry was detected. Beneficiaries still face some confusion in this market, however, such as understanding the rating methodologies used to set premiums and how this may affect their choices. Confusion could increase with the growth of managed care options.

Introduction

The 1990s ushered in a new era of health care competition with the inclusion of a little-known provision in Congress' Omnibus Budget Reconciliation Act (OBRA) of 1990. Calling for the establishment of mandatory standardized benefit packages, this legislation revolutionized the marketing and sale of Medicare supplemental (“medigap”) insurance to the elderly. Future insurance reform efforts based on managed competition would be remiss in not considering the lessons learned from standardizing the medigap market.

All medigap policies sold after July 31, 1992¹ must conform to 1 of 10 uniform benefit packages, labeled A through J, as stipulated by OBRA 1990 and subsequent State regulations. Previously, hundreds of different benefit packages were available to consumers. Standardizing benefits, effectively limiting the number of options available, was designed to simplify consumer purchases, thereby making it easier to compare benefits and premiums.

Standardized benefit packages fulfill a key tenet of managed competition by facilitating comparisons between alternative health insurance choices (Enthoven, 1988). The Clinton Administration's Health Security Act and several rival bills proposed that health benefits be standardized. Although these reform efforts failed, it is still possible to learn about standardization by a careful examination of the medigap experience. As set forth in the House Conference Report, the OBRA 1990 medigap legislation had five objectives:

• Simplifying the market to facilitate policy comparisons.

• Providing consumer choice.

• Providing market stability.

• Promoting competition.

• Avoiding adverse selection.

Based on interviews and data collected from nine States over a 4-year period, this article is organized around these five objectives and the extent to which they have been met. In addition, we have included a section on Federal and State administration of the medigap program since regulatory oversight cuts across several of the objectives.

The next part of the article provides background on the problems in the medigap market and regulatory responses. Then, the data sources and methods employed for the study are discussed. This is followed by the results, in which we examine the success or failure of the legislation in meeting its five objectives. We conclude with a summation of the legislation's overall effect and a discussion of remaining problems in the medigap market.

Background

Problems in the medigap market date back to the inception of Medicare. The program was not designed to pay all of the health care costs of the elderly and disabled. Therefore, gaps in coverage remain, in the form of various copayments and services that are not covered at all, such as long-term care and prescription drugs. To fill some of these gaps, private insurers responded by selling supplemental health insurance, known as “medigap” policies.

In 1991, almost 90 percent of the elderly population had some kind of health coverage in addition to Medicare.² Of these, 35 percent were covered through a current or former employer, 37 percent owned individual coverage, 7 percent had both individual and employer coverage, and 12 percent were eligible for Medicaid (Chulis, Eppig, and Poisal, 1995; Chulis et al, 1993). This article pertains to the 42 percent of Medicare beneficiaries (approximately 13 million people) who own individual supplemental coverage, because only these policies are subject to the OBRA 1990 legislation.³ Retiree benefits were specifically excluded from the scope of the legislation.

The sale of individual medigap policies has been subject to a number of problems, albeit some of them isolated: marketing abuses by companies and agents; low rates of return on premiums; duplicate coverage; and poor consumer knowledge. Hearings on these problems spurred Congress to enact the “Baucus Amendments” of 1980 (Public Law 96-265).⁴ This legislation established voluntary certification standards designed to encourage States to enact legislation that incorporated certain basic requirements established by the National Association of Insurance Commissioners (NAIC), the association of chief State insurance commissioners. Specifically, medigap policies had (among other things) to meet minimum benefit package requirements, meet minimum loss ratio standards (60 percent for individual policies and 75 percent for group policies),⁵ and comply with various disclosure provisions, including providing a consumer guide and outline of benefits to prospective policyholders.

Although the Baucus legislation achieved many of its goals (U.S. General Accounting Office, 1986), one major problem remained. No limits existed on how many and what type of benefits could be offered in excess of the minimum standards. Consumers, faced with hundreds of policy configurations, found it difficult to comparison shop effectively.

OBRA 1990 was designed to address this shortcoming of the Baucus legislation. It stipulates that after July 1992 (with the exception described in footnote 1) the only policies that could be sold as Medicare supplements were the 10 specified benefit packages (labeled A, B, C, … J), which are shown in Table 1. All carriers selling medigap policies are required to offer Plan A but can also choose to sell any or all of the remaining packages. The comprehensiveness of the packages generally increases from A through J. All packages cover a core set of services, which includes the Part A hospital coinsurance for stays longer than 60 days plus coverage for 365 additional lifetime reserve days after Medicare benefits end, the Part B coinsurance, and the Parts A and B blood deductible. In 1996, the Part A hospital coinsurance was $184 for the 61st through 90th day of hospitalization per benefit period and $368 per day for days 91 through 150.⁶ Coverage for the Part B coinsurance generally includes 20 percent of Medicare's approved amount or 50 percent of outpatient mental health services after meeting the annual $100 Part B deductible.

Table 1. Benefits Covered in the Medigap Policy Prototypes Under OBRA 1990.

Benefits	Plan Type
	A	B	C	D	E	F	G	H	I	J
Core Benefits1	X	X	X	X	X	X	X	X	X	X
SNF Coinsurance	—	—	X	X	X	X	X	X	X	X
Part A Deductible	—	X	X	X	X	X	X	X	X	X
Part B Deductible	—	—	X	—	—	X	—	—	—	X
Part B Excess Charges	—	—	—	High2	Low2	—	—	High2	High2
Foreign Travel	—	—	X	X	X	X	X	X	X	X
At-Home Recovery	—	—	—	X	—	—	X	—	X	X
Prescription Drugs	—	—	—	—	—	—	—	Low3	Low3	High3
Preventive Medical Care	—	—	—	—	X	—	—	—	—	X

¹Core benefits include coverage of all Part A (hospital) coinsurance for stays longer than 60 days, the 20-percent Part B coinsurance, Parts A and B blood deductible, and the 365 lifetime reserve days of hospital care.

²Low excess charge coverage pays 80 percent of the difference between the physician's charge and the Medicare-allowable rate; high coverage pays 100 percent of the difference.

³Low prescription drug coverage has a $250 annual deductible, 50 percent coinsurance, and a maximum annual benefit of $1,250; high coverage is similar but it has a $3,000 maximum annual benefit.

NOTES: OBRA is Omnibus Budget Reconciliation Act. SNF is skilled nursing facility. Plan types A-J represent the 10 uniform benefit packages mandated by OBRA 1990. Plan A represents the least comprehensive package; Plan J represents the most comprehensive.

SOURCE: National Association of Insurance Commissioners: Medicare Supplement Insurance Minimum Standards Model Act, July 30, 1991.

With regard to the non-core benefits, more than one-half of the packages cover the Part A deductible, skilled nursing facility (SNF) coinsurance, and foreign travel services. In 1996, the Part A deductible was $736. Coverage for SNF coinsurance amounts to $92 per day for days 21 through 100 per benefit period.⁷ Packages with the foreign travel benefit pay 80 percent of medically necessary emergency care in a foreign country following a $250 deductible (with a lifetime maximum of $50,000).

Only a few of the plans cover preventive medical care and at-home recovery. The preventive screening benefit pays a maximum of $120 per year for physician-ordered health care screenings. The at-home recovery benefit pays up to $40 per visit for up to 7 visits a week and can be used up to 8 weeks after Medicare-covered home health care benefits cease. The ceiling for this benefit is $1,600. Three packages (H, I, and J) cover prescription drugs; each plan is subject to a $250 deductible and 50 percent coinsurance for expenses above the deductible. Plans H and I have a maximum benefit amount of $1,250, while Plan J has a maximum benefit amount of $3,000.

The process undertaken to establish the benefit packages was unique in that it gave a private body, NAIC, the first opportunity to formulate their specifications with the Federal Government playing only a default role if NAIC failed to act in a timely fashion. NAIC had 9 months to complete this process with little guidance from Congress as to content or process.

Although policy standardization represents the most far-reaching provision of the legislation, OBRA 1990 included several other provisions, including the following:

• Higher loss ratio requirements for individual policies. Loss ratios were increased from 60 percent to 65 percent and refunds to consumers are required when policies fail to meet the new standards. Minimum loss ratios for group policies remain at 75 percent but are now subject to the refund requirements.

• Severe penalties on agents and insurers who knowingly sell duplicate coverage.

• Limits on agent commissions during the initial year of coverage under a new policy. Commissions are restricted to no more than twice the amounts paid for policy renewals in subsequent years.

• Requirement that insurers hold a 6-month open enrollment period when beneficiaries 65 years of age or over first enroll in Part B of Medicare.⁸ Thus, during that 6-month period, a person can purchase any policy offered regardless of his or her health status and receive the company's most favorable rate.

• The requirement that pre-existing condition exclusions could not exceed 6 months in duration. Also, a second preexisting condition period may not be imposed if a person switches plans or carriers. (However, after the 6-month open enrollment period, the carrier can refuse coverage or charge a higher premium for the policy based on past health care experience.)

Data and Methods

Data for this study were gathered through extensive case study interviews with key participants in the medigap supplemental market. Three waves of interviews were conducted; the first wave was completed in spring 1992, the second roughly 1 year later, and the final wave in spring 1995. This timeframe allowed us to capture both pre- and post-OBRA 1990 experiences. Since most States did not implement until mid-1992, the first round of interviews reflects the regulatory environment prior to standardization.

Interviews were conducted with multiple officials in the department of insurance (DOI) or other State agency that regulates the medigap market in each of the nine States. The States–California, Florida, Minnesota, Missouri, New York, South Carolina, Texas, Washington, and Wisconsin-were selected with assistance from NAIC with the intent of including States with diverse regulatory environments. Other criteria used for choosing case study States were having a large elderly population, geographic diversity, and waiver status.⁹ Representatives from the State Medicare consumer counseling program and/or consumer division of the DOI were also interviewed. Counseling programs were developed or enhanced in all States through funding provided by OBRA 1990. The programs have different names in most States but serve the same function: to provide free, unbiased information, counseling, and assistance (ICA) to Medicare beneficiaries facing various types of health insurance decisions.

Interviewees also included representatives from insurance carriers selling medigap policies, public interest groups, executive branch officials, trade associations, and congressional staff members.¹⁰ Including State representatives, nearly 100 interviews were conducted during each wave. As a result of the diversity of participants, we were able to obtain many perspectives on medigap insurance. With few exceptions, the same individuals were interviewed during each wave, which enhanced the consistency of the findings over the duration of the study. Interviews with carriers were conducted by telephone, while all State interviews and most of the remaining interviews were conducted in person. An interview guide was prepared in advance of each wave to help structure the discussion.

Discussion topics for interviews with carriers included the impact of the legislation on competition, marketing, beneficiaries, and administration of their medigap product. State interviews addressed issues regarding administration of carriers, enforcement of the legislation, and the impact of standardization on beneficiaries and carriers. Some quantitative data, such as the cost of insurance premiums and the number of consumer complaints, were also retrieved from State agencies when available. All data were examined to identify reported patterns and trends, with particular attention dedicated to assessing whether the objectives of legislation were achieved.

Findings

Findings from the study are presented in this section, which is structured in accordance with the objectives of the legislation (as cited in the Background section). We begin with the goal of simplifying the medigap market. Next, we discuss the impact of the legislation on consumer choice and the extent to which market stability resulted and competition was promoted. The issue of adverse selection is addressed in the subsequent section. We close with a section on the Federal and State oversight of the medigap market.

Simplifying the Market to Facilitate Policy Comparison

By most accounts, OBRA 1990 has simplified the medigap market. Establishing 10 standardized packages of benefits was an underlying factor leading to a market that respondents described as more understandable and beneficial to consumers. According to the vast majority of study participants, the 10-plan structure successfully facilitates direct policy comparison, making shopping for a policy easier and more straightforward. This was the position most frequently held by State officials and representatives of the State consumer ICA programs. Most believed that consumer confusion has diminished as a result of the standardization of benefits, and beneficiaries are now able to make more educated purchasing decisions.

As one measure of lessened confusion, the number of consumer complaints filed with the State DOIs declined following implementation of OBRA 1990. A standard complaint form, developed by the NAIC and adopted by many States in the early 1990s, allows States to track medigap-specific consumer complaints, which not all had done previously. Before using the standard form, several States simply collapsed all complaints under a broad “Medicare-related” category.

Reliable longitudinal data on consumer complaints are not available nationally. We collected these data directly from the non-waivered States for the years immediately preceding and following implementation of OBRA 1990 (1990-94).¹¹ As shown in Figure 1, the data from most States show a noticeable downward trend in the number of complaints during that time period. For example, in Missouri the number of consumer complaints logged declined from 226 in 1990 to 56 in 1994, a 75-percent reduction. The most dramatic decrease occurred in Florida, where the number of complaints fell from 812 in 1990 to 178 in 1994, a 78-percent reduction.

Figure 1. Consumer Complaints Reported to State Departments of Insurance.

New York was the one exception among the eight States. In fact, complaints increased in the State from 21 in 1990 to 209 in 1994.¹² However, other significant changes in the insurance market were occurring in New York beginning in early 1993 when the State adopted individual and small group market reform for most types of insurance, including medigap. The market reform mandated continuous open enrollment, development of a State risk pool, and the use of community rating for groups of 3 to 50 and in the sale of individual coverage. Since these reforms included the medigap market, they may have influenced New York's complaint data. During that time period, Empire Blue Cross/Blue Shield (BC/BS) also raised its medigap premium rates, which may have led to consumer complaints.

Marketing and agent practices have historically been among the most common reasons for consumer complaints. Since benefits are standardized, plans are easier to compare, leaving sales agents with less of a role in explaining products. As such, fewer opportunities exist to engage in unscrupulous marketing practices. This may have contributed to the reduction in consumer complaints. In addition, insurance agents now have less monetary incentive to sell medigap policies to replace existing policies, i.e., “churning,” because of the restrictions placed on commission levels by OBRA 1990. Specifically, first-year commissions cannot exceed 200 percent of policy renewal commissions. Some States have instituted even stricter requirements. For example, Washington State mandates level commissions. The penalties against agents and insurers who knowingly sell duplicate coverage also provide a disincentive against the sale of unnecessary coverage.

One variable for which we were unable to account when examining the trend in consumer complaints was the potential change in the repository of complaints. As previously mentioned, State ICA programs for Medicare beneficiaries were established or enhanced during this time period. It is possible that consumers began to voice their complaints to ICA counselors instead of filing them with the DOIs. Approximately 200,000–not quite 1 percent of elderly Medicare beneficiaries nationwide—received assistance from an ICA program in 1993¹³ (McCormack et al., 1996). Representatives from the ICA programs and State agencies also reported an apparent overall shift away from medigap issues to other insurance matters, which is further evidence of lessened confusion among seniors.

The potential for consumer confusion increases with the growth of Medicare health maintenance organizations (HMOs) and related managed care plan arrangements. Until recently, the Medicare risk program, which makes capitation payments to HMOs, was little more than a slowly evolving pilot program.¹⁴ However, that has changed. Between April 1994 and January 1996, a 21-month period, the number of plans with Medicare risk contracts increased 60 percent–from 118 to 189–and the number of enrollees rose 68 percent–from 1.9 million to 3.2 million (Health Care Financing Administration, 1994 and 1996). Furthermore, the number of HMO risk applications under review by HCFA is at an all-time high, and various proposals introduced in Congress would enhance the already explosive growth.

HMO benefits can be as confusing as indemnity (medigap) benefits. They differ, for example, in the prescription drug coverage, which is much more prevalent for HMO than medigap enrollees, e.g., in the maximum benefit amounts, whether maximums are stated on a quarterly or annual basis, and the copayment amount and structure, including whether there are financial incentives to purchase generic rather than brand name drugs. HMOs commonly offer dental, vision, hearing, and prevention benefits which differ. Also, the copays associated with emergency room, urgent care, and physician use vary based on whether the physician is a specialist or primary care doctor and whether care is sought during or after hours.

The growth in new types of managed care plans can be expected to exacerbate confusion. HCFA recently clarified that Medicare risk HMOs could offer point-of-service (POS) options, which can differ across health plans in their non-network benefits. The expansion of the Medicare SELECT¹⁵ program to all States (from only the 15 demonstration States) (Garfinkel et al., 1996) could also compound the situation. Finally, Congress is considering loosening the requirements on plans and provider groups to become a Medicare risk contractor, which will result in further growth in options available to the consumer.

Providing Consumer Choice

Carriers selling medigap policies differ in their perspectives regarding the impact of the legislation on consumers. Some believe that OBRA 1990 had a positive effect on consumers, whereas others hold that consumer choice along with creativity was restricted by having only 10 standardized benefit packages available. On the other hand, consumer representatives generally indicated that 10 plans were too many, particularly given the growing number of other options available to Medicare beneficiaries. Individuals who were involved in the development of the benefit packages undoubtedly discussed concerns about providing consumers with sufficient choice. In the end, they decided to develop 10 plans, which was the maximum number allowed under the new law. More recently, many of those who originally favored fewer plans now believe that the number should not be changed in order to avoid further confusion because the elderly have become accustomed to the 10 plans. In short, their plea is for stability over perfection.

Table 2 shows the average distribution of enrollment in plans A through J, using data collected from Florida, Missouri, New York, South Carolina, Texas, and Washington for 1994.¹⁶ A high proportion of beneficiaries are enrolled in a small number of plans; two-thirds of beneficiaries have purchased Flan B, C, or E Most enrollees want coverage more extensive than the minimum (Plan A), which accounted for only 5 percent of sales, although it is the only plan that carriers are required to sell. Plan F enrolls the highest proportion of beneficiaries (29.7 percent) followed by Plan C (21.2 percent). Plans F and C are similar in structure except that Plan F pays for Part B excess charges. Many consumers may not understand the physician balance-billing limitations in the Medicare Act (title XVIII of the Social Security Act) and are unaware of the increase in the assignment rate in recent years, which reduces the need for such coverage in medigap policies.¹⁷ The plan with the third highest enrollment is Plan B (17.1 percent), which only covers the Part A deductible in addition to the core benefits.

Table 2. Distribution of Sales of Standardized Plans in 1994.

Benefits	Plan Type
	A	B	C	D	E	F	G	H	I	J
Core Benefits	X	X	X	X	X	X	X	X	X	X
SNF Coinsurance	—	—	X	X	X	X	X	X	X	X
Part A Deductible	—	X	X	X	X	X	X	X	X	X
Part B Deductible	—	—	X	—	—	X	—	—	—	X
Part B Excess Charges	—	—	—	—	High	Low	—	—	High	High
Foreign Travel	—	—	X	X	X	X	X	X	X	X
At-Home Recovery	—	—	—	X	—	—	X	—	X	X
Prescription Drugs	—	—	—	—	—	—	—	Low	Low	High
Preventive Medical Care	—	—	—	—	X	—	—	—	—	X
Percentage Distribution of Sales	5.1	17.1	21.2	8.4	0.8	29.7	2.2	2.7	5.9	6.9

NOTES: SNF is skilled nursing facility. Plan types A-J represent the 10 uniform benefit packages mandated by OBRA 1990. Plan A represents the least comprehensive package; Plan J represents the most comprehensive.

SOURCE: Data provided by the State departments of insurance in Florida, Missouri, New York, South Carolina, Texas, and Washington, 1994-95.

During the interviews, some carriers noted that certain benefits found in the 10 standardized plans are of limited interest or value, namely, the preventive care benefit and Part B deductible, which were referred to as “dollar trading.” For example, the preventive care benefit reimburses actual charges up to $120 and the Part B deductible covers the first $100 of physician charges. Electing plans with these benefits essentially results in prepayment. Some policyholders find this attractive since they do not have to concern themselves with paying bills and are able to plan their annual health care expenses. Still, the limited interest in preventive benefits is evidenced by the low demand (0.8 percent of the total sales) for Plan E which covers preventive care, and the higher demand for Plan C which covers nearly the same benefits as Plan E except preventive care. Another benefit that carriers cited as being of limited interest to consumers was at-home recovery. The data support this claim-plans D, G, I, and J cover this service; however, enrollment in these plans ranges from only 2.2 to 8.4 percent. Consumers have the opportunity to purchase all 10 plans from national carriers like AARP/Prudential. Thus, the low level of sales reflects a demand as opposed to a supply issue.

Some carriers also remarked that consumers were without access to certain kinds of coverage altogether because they are no longer able to purchase selected comprehensive policies that were sold prior to 1992, especially those that had significant drug coverage or catastrophic benefits. Only three of the standardized plans (H, I, and J) cover prescription drugs, each with sizable cost–sharing amounts. In 1993, only about 1 in 10 medigap purchasers had one of these policies, suggesting little interest on the part of consumers, at least at the prevailing prices. Yet drug policies were not popular before standardization; 4 to 17 percent of purchasers had a plan that covered prescription drugs in 1990 (Rice and Thomas, 1992). Thus, the market reform does not appear to have affected consumers' demand for drug coverage.

Some consumer groups have noted the absence of coverage for individuals with high outpatient prescription drug expenses such as those taken by transplant, acquired immunodeficiency syndrome (AIDS), cancer, and end stage renal disease patients. However, raising the maximum payment limits for the three plans that cover prescription drugs (Plans H, I, and J) would have resulted in higher premiums and exacerbated the pattern of biased selection that was observed for these plans (see the Avoiding Adverse Selection section later).

One approach would have been to include catastrophic drug protection as a basic benefit in all of the 10 standardized plans. The State of Wisconsin did just that. Wisconsin is a waivered State (along with Minnesota and Massachusetts) because it had adopted its own standardization program prior to the passage of OBRA 1990. In 1994, it modified its core program to cover prescription drug expenses above an annual deductible of $6,250. That deductible integrates with the optional drug rider that is offered under the Wisconsin program and is equivalent to the drug benefits in Plan J of the national program, which in turn ceases to pay once $6,250 in expenses have been reached.¹⁸

Earlier research has also demonstrated that some of the previously sold benefits, such as private duty nursing, were of little value. Although one-half of medigap policy holders had a private duty nursing benefit prior to standardization, carriers and consumer groups found it difficult to argue in its favor since it was more of a luxury than a necessity (Rice and Thomas, 1992). In addition, the structure of the benefit was very confusing. Thus, consumers may in fact be better off without the option of purchasing this benefit.

Based on the distribution of enrollment, it appears that consumers find coverage of Medicare's coinsurance and the Part A deductible valuable and coverage for less costly services, e.g., preventive screening and at-home recovery, to be less attractive (Fox, Rice, and Alecxih, 1995). With 10 plans available from which to choose, consumers gravitate toward only three plans (F, C, and B) suggesting that elimination of the hundreds of different medigap packages that existed prior to OBRA 1990 has not been detrimental to consumer choice.

So as not to impede market driven innovation, OBRA 1990 allows “innovative benefits” to be sold if they are approved by the State. Most States felt that innovative benefits constituted an 11th benefit package and were reluctant to approve them. Furthermore, requiring that each State decide separately on innovative benefits poses a challenge to carriers that market in multiple States. These carriers must obtain approval from every State in which they wanted to sell the benefit. As a result, innovative benefits are more likely to be available from carriers that sell policies in only one State. Only one of our case study States and fewer than a half-dozen States nationally have approved innovative benefits as of early 1995. For example, BC/BS of Florida is selling a dental benefit and an Arkansas carrier is offering a vision care benefit (Health Care Financing Admini-stration, 1995). It appears that the use of innovative benefits is generally limited to BC/BS plans.

Providing Market Stability

As a measure of market stability, we examined changes in the number of carriers selling medigap policies before and after the market reform. The number of carriers in New York remained fairly steady between 1990 and 1995. However, all of the other non-waivered States in our sample reported a decline in the number of carriers selling medigap policies immediately after standardization. The number of carriers was at a peak immediately prior to implementation, bottomed out in 1992, and began to slowly rise in the years following implementation. The actual pattern is State-specific. In California, there were approximately 100 carriers prior to standardization but only one-half as many by 1995. At a low point in 1992, only 10 carriers were approved to sell policies in the State. However, at no point in time did the number of medigap carriers fall below 10 in any case study State.

This pattern is even more evident when it is compared with data from two waivered States, Minnesota and Wisconsin–the number of carriers in these two States remained steady throughout the early to mid-1990s. The decline and slow rise in the number of carriers can be attributed to an initial hesitation about entering a market where loss ratios are dictated by law and enforced through refunds. For the most part, it was the small- and medium-sized companies that exited the market following standardization, and the same groups that later rejoined. It is not surprising that the smaller carriers did not transition as quickly as larger carriers, whose greater resources permitted more flexibility.

Despite the reduction in the number of carriers, consumers still had ample choices. We found a general consensus among regulators and carriers that consumers were not disadvantaged as a result of the reduction in the number of carriers, since a sufficient number of companies remained. Although the market is led by a few carriers-namely the AARP/Prudential and State BC/BS plans–which together constitute nearly two-thirds of sales nationally, no significant barriers to entry were detected.

Promoting Competition

Standardization is intended to improve the ability to compare benefit packages and, hence, enhance price competition. If it achieves this objective, one would expect to see a higher proportion of the premium dollar paid out in benefits rather than retained for administration and profits, i.e., higher loss ratios.

Loss ratios were, in fact, higher during the study period. For individual policies in particular, the loss ratios were at high levels by historic standards in 1994, the most recent year for which data are available–75.2 percent (Table 3). In contrast, they ranged between 70.3 and 71.3 percent in 1990-93, and in particular were unchanged in the first year in which standardization occurred. We hypothesize, based in part on carrier interviews, that companies in the initial year or two of standardization priced conservatively because they were having to sell new products for which they lacked direct claims experience. Over time they priced more aggressively, reflecting competition and solid profit margins.

Table 3. Medigap Loss Ratios for Individual and Group Policies: 1990-94.

Year	Individual	Group
1990	70.3	81.0
1991	71.6	80.5
1992	70.5	76.4
1993	71.3	70.4
1994	75.2	82.3

NOTE: The data refer to polices issued in the year in question and the 2 prior years.

SOURCE: Data provide by the National Association of Insurance Commissioners, 1995.

However, OBRA 1990 also raised the minimum loss ratio from 60 to 65 percent, effective for individual policies sold or issued after November 5, 1991. (The loss ratio requirement for group policies remained at 75 percent.) In theory, either standardization or the higher minimum loss ratio requirements could have generated the increase in loss ratios nationally. We suspect that the impact of standardization leading to enhanced competition was the dominant reason. Most carriers have loss ratios that are considerably above the minimum and, thus, are unaffected by the requirements, whereas enhanced competition affects all carriers in the market.

Companies failing to meet the standard are required to provide refunds or credits to policyholders. No such refunds or credits had been issued as of our last wave of interviews (spring 1995); however, the legislation called for 3 years of data to be used when assessing loss ratios. Most States implemented standardization in mid-1992, thus, 3 years of loss ratio data were barely available at the end of our study. Consumers should have received their first distribution of refunds and credits in late 1995.

The process required to evaluate companies' loss ratios is likely to be administratively complex and time consuming, one that most States we spoke with had not made or were not planning to make a priority. Some States expect that they will enforce the loss ratio provision prospectively by denying future rate increases or requiring a rate reduction. Other States had no specific agenda for addressing this potential problem.

One possible response to enhanced competition is for carriers to set premiums to reflect more accurately how costs vary as a function of enrollee age. Carriers can rate medigap policies in three ways: (1) without any age rating; (2) with issue age rating; and (3) attained age rating.¹⁹

Issue age premiums are based on the age of the beneficiary when the policy is issued, increasing only for inflation but not because the beneficiary ages. Under attained age rating, carriers increase rates as policyholders become older. One problem with attained age rating is that it results in higher policy premiums as beneficiaries grow older, when their incomes in real terms are likely to decline.

Based on interviews with carriers and States, it appears that there has been an increase in the number of carriers that use attained age rating to price their products. However, this shift is not universal, as evidenced by the data we collected from large carriers on rating methods. In fact, some of the larger carriers are not changing to attained age pricing as a business strategy. The movement to attained age rating may be occurring among the smaller carriers in an attempt to attract business by having lower initial premiums.

Table 4 shows the average annual premiums in 1993 and 1994 for medigap policies using data collected from six case study States. Premiums ranged from $524 to $1,553 in 1992 and $528 to $1,811 in 1994. The plan with highest enrollment (Plan F) cost more than $1,000 for several years. In 1994, annual premiums for the drug plans (H, I, and J) averaged $1,453-nearly 3 times higher than premiums for the most basic policy-pricing it out of reach of many elderly consumers.

Table 4. Average Annual Premiums Paid for Medigap Polices: 1992 and 1994.

Plan Type	1992 Premium	1994 Premium	Percent Change
A	$524	$528	1
B	813	838	3
C	888	908	4
D	828	872	5
E	830	870	5
F	1,088	1,117	3
G	1,013	1,050	4
H	1,888	1,203	3
I	1,270	1,344	6
J	1,553	1,811	17

SOURCE: Data provided by the State departments of insurance in Missouri, New York, South Carolina and Texas, 1993-95.

Premium increases between 1992 and 1994 ranged from 1 to 17 percent, suggesting that the utilization experiences of the plans were quite different and that carriers were still learning how to price these new products. Plan A's premium had the smallest increase; however, only 5 percent of consumers have elected this option. The most popular plan (Plan F) experienced a 3-percent increase in its premium ($1,088 to $1,117). Plan J experienced the most dramatic premium increase (17 percent). On average, some of these increases are small compared with some of those that were projected for 1996. AARP/Prudential indicated that it planned to raise premiums on their medigap policies an average of 30 percent following a significant increase in claims volume (Washington Post, 1995). Selected carriers have filed for large premium increases for their closed prestandarized blocks of business as a result of their shrinking and aging risk pools.

Avoiding Adverse Selection

Adverse selection is an issue primarily for the three packages that include prescription drugs (Plans H, I, and J). The extent of biased selection can be inferred from a comparison of premium levels of these three plans with those that do not cover prescription drugs, assuming a reasonable relationship between premium and benefit payment levels. Unfortunately, an analysis of relative premiums is complicated by the fact that each of the seven non-drug plans incorporates other changes, thus precluding direct comparisons. However, we were able to account for other changes. To do this, the firm of Bob Gold and Associates was engaged to provide actuarial estimates.

Three close matches to Plans H, I, and J were identified from other standard plans. These matches and the benefit differences other than drugs are as follows:

• Plans H and C differ only in the coverage of part B deductible.

• Plans I and D differ only in the coverage of excess charges.

• Plans J and E differ only in the coverage of the Part B deductible and at-home recovery.

For each pair, the impact of the benefit differences other than prescription drugs was removed, leaving a dollar amount that can be attributed to drug coverage, as shown in Table 5.²⁰

Table 5. Premium Adjustments to Remove Impact of Non-Drug Benefits.

Plan	Adjustment
Plan H to Plan C
Plan H Premium	$1,182
Part B Deductible	+110
Adjusted Premium	1,292
Plan C Premium	-852
Premium Attributable to Drug Coverage	440
Plan I to Plan D
Plan I Premium	$1,310
Physician Excess Charges - High	-103
Adjusted Premium	1,207
Plan D Premium	-817
Premium Attributable to Drug Coverage	390
Plan J to Plan E
Plan J Premium	$1,828
Part B Deductible	-110
At-Home Recovery	-90
Adjusted Premium	1,628
Plan E Premium	-833
Premium Attributable to Drug Coverage	795

NOTE: Calculations assume 20 prescriptions per person.

SOURCE: Premiums are based on average 1994 annual premiums in Missouri, New York, South Carolina, and Texas, 1993-95.

A reasonable estimate for the cost of a single prescription in an indemnity environment is $40 per fill. In 1987, the last year in which utilization data were collected through a national probability sample survey, the average number of prescriptions per capita for the elderly was 14.7 (U.S. Department of Health and Human Services, 1989). Using a figure of 20 prescriptions per person, which is selected to overestimate increases since 1987, and multiplying by $40, we obtain an estimated annual expenditure for the average elderly person of $800.

The prescription drug benefit for Plans H and I has a $250 annual deductible, 50 percent coinsurance, and a limit on benefit payments of $1,250. Thus, a high proportion of fills do not generate any benefit payments (because they are below the deductible or above the expenditure limit), and those that do are subject to 50 percent coinsurance. Actuarial estimates show that 23 percent of total drug expenditures for a typical elderly population would be below the $250 deductible, and another 21 percent would be above the cap, leaving 56 percent of prescriptions for which benefits are paid. Dividing 56 percent in half to reflect the 50 percent coinsurance results in 28 percent of total drug expenses-equal to $224-being reimbursed.²¹ However, the $224 does not include the costs of administering the benefit. Assuming a loss-ratio (ratio of benefits to premiums) of 80 percent, the total cost becomes $280 (equal to $224 divided by 0.8). This is well below the differences in drug-only portion of the premiums of $440 and $390 calculated in Table 5 for Plans H and I, respectively.

For Plan J, which has an annual benefit limit of $3,000 rather than $1,250, some 35 percent of expenditures can be expected to be reimbursed. Since the portion of the premium attributable to drug coverage for Plan J is estimated at $795, the impact of biased selection is even greater.

This analysis demonstrates that carriers do set premiums to account for significant adverse selection when pricing coverage that includes prescription drugs. Three points are important in interpreting these results. First, we are interested only in significant differences that are not sensitive to minor changes in estimating assumptions, e.g., related to the average retail price per prescription or the dollar amount attributable to non-drug benefits. Second, the analysis presumes that the carriers price to achieve roughly comparable loss ratios across the benefit packages that they sell. Third, if the prescription drug coverage does attract less healthy enrollees than average, as hypothesized, this effect is likely to manifest itself across all benefits, not just in the prescription drug use.

Program Administration

Medigap is the only type of private health insurance for which responsibility for regulation falls to the Federal Government, mostly HCFA, in the U.S. Department of Health and Human Services (DHHS). However, HCFA's role is one of policy setting and oversight, with ongoing administration conducted by the States, which also have the right to establish regulations that are more stringent than those provided in OBRA 1990.

Two principal organizations within HCFA address medigap issues–the Bureau of Policy Development, which establishes policy guidelines, and the Bureau of Program Operations, which oversees the implementation. Typically, after legislation is passed, HCFA issues implementing regulations, as prescribed by the Federal Administrative Procedures Act. These elaborate on the legislation and must be consistent with it. Most importantly, they have the force of law, which is generally not true of other forms of policy issuances. Some legislation is self-implementing, i.e., does not require regulations, or some may require regulations only to address relatively narrow matters.

At the time of this writing, HCFA had yet to issue regulations relating to the OBRA 1990 medigap provisions. The exact impact of its not having done so is debatable. Some argue that since there are no material problems at present, the absence of regulations, which are by necessity detailed and prescriptive, may be advantageous because it allows State flexibility. However, regulations are required, for example, to implement the Federal civil and criminal penalty provisions, such as those associated with misrepresentation, fraud, and the sale of duplicate coverage. The enforcement of these provisions at the Federal level is through the Office of the Inspector General (OIG) within DHHS, but independent of HCFA. Enforcement of the penalty provisions is, apparently, of low priority to the OIG, given its desire to focus resources on matters that, unlike medigap, adversely impact the Federal budget.

Other areas that would presumably be addressed through the regulations include:

• How HCFA will perform oversight of the States (e.g., whether to perform onsite surveys)–some HCFA staff believe that the absence of regulations gives them administrative flexibility.

• The definition of “innovative benefits”–which may be inherently difficult to define.

• How Part B psychiatric benefits are defined (whether the medigap plans are liable for the full 50-percent coinsurance facing the beneficiary or just the 20 percent that applies to other services). HCFA's position is that the plans are liable for the full 50 percent.

HCFA periodically issues bulletins on policy and operational matters to State regulators that are only advisory to the State and, thus, are not a substitute for regulations.

To date, Federal enforcement has principally resulted from complaints from beneficiaries or from reviewing changes to State regulations.²² There have been instances in which HCFA has intervened with the States to prevent sale of policies that were out of compliance with the standardization requirements, but this has been rare.

State reaction to how HCFA has fulfilled its role during the implementation process has been mixed. HCFA was required to approve all of the State programs to ensure that they met the minimum requirements of the legislation. Some report the process as being smooth, others as drawn-out and painful, with answers to questions entailing long delays. The problems that did arise reflect to a significant degree the newness of the program for HCFA, which has not traditionally played a role in regulating private indemnity insurance. Another source of the States' criticism is the involvement of multiple governmental offices, including the General Counsel, which issues legal interpretations. The medigap provisions of OBRA 1990 apparently do not represent a high priority for HCFA, which regards ongoing enforcement as fundamentally a State responsibility.

Some States experienced confusion in administering the legislation. Unclear legislative language addressing the sale of duplicate coverage created ambiguities for States and carriers alike. In particular, exactly what types of Medicare supplemental policies (e.g., dread disease, cash indemnity policies [typically paying a fixed amount per day for every day in the hospital], and long- term care policies) were subject to the duplicate policy provisions was contentious. States did not enforce the provision prior to enactment of amendments in 1994, the most important component of which requires disclosure of duplication. Even now, ambiguity remains regarding whether long-term care policies that coordinate with Medicare can be sold (in particular by paying nursing home benefits only after the beneficiary is no longer covered, typically because they do not meet the skilled nursing requirement).

Changes imposed by OBRA 1990 affected how the insurance carriers could define and sell insurance, and how much they could profit from their products. Despite the regulatory metamorphosis that occurred in a roughly short period of time, the transition process was fairly smooth once the initial burden of implementation was overcome. Unequivocally, the greatest challenge in implementation was the demanding time schedule. Overall, administration of the medigap market has been free from major barriers. Several States reported that the current structure is more straightforward and efficient, but this did not occur without an increase in paperwork that is required to meet the new prior approval regulations. As a result of the general ease in administration, none of the States we cite visited added staff as a result of OBRA 1990.

Carriers, for their part, had mixed to positive reactions to the performance of States. Some carriers complained of long delays in the approval processes in some States. Also, some carriers that sell in multiple States felt that, as long as the benefits were standardized, the policy forms should have been standardized. This would obviate having to meet myriad State requirements that differed mostly with regard to how the policies are presented physically, e.g., required type size and font, rather than their substance. With regard to the legislation itself, the poorly written anti-duplication provisions previously discussed created confusion, and many carriers have found the at-home recovery benefit in Plans D, G, I, and J difficult to administer because the benefit is somewhat complex.²³

Conclusion

Overall, we believe that the medigap reforms in OBRA 1990 have met their goals. The most dramatic change is the standardization requirements. Although a definitive assessment will require the passage of more time, especially with regard to loss ratios, several findings have emerged that are favorable. Consumer complaints have decreased in most States. There are ample numbers of carriers selling the product, despite the more competitive environment.²⁴ State official and consumer representatives are, with few exceptions, supportive of the standardization requirements. Loss ratios are higher than they had been, at least in the last year (1994) for which data are available at the time of this writing. Finally, although the implementation process has had some administrative problems, these have proven solvable, and after the initial transition period, standardization has reduced the review burden on the States.

On the more negative side, OBRA 1990 has apparently not prevented adverse selection, as evidenced by carrier pricing of policies that include the prescription drug benefit. Arguably, preventing adverse selection was an unrealistic expectation. Indeed, the enhanced ability of consumers to make price comparisons, one of the key objectives of OBRA 1990, may increase adverse selection as a result of consumers' better understanding what they are purchasing and tailoring their choice of plans to their individual needs.

As is true with any significant legislative change, issues remain, ranging from those with broad philosophical import to minor and technical ones. One broad issue relates to the resulting limitation in consumer choice. Standardization reduces confusion, thereby promoting competition, but at the expense of diversity. To be sure, not all diversity is good, and OBRA 1990 reduced the availability of benefits that are arguably of little value, e.g., private duty nursing, and eliminated minor benefit package variants that add to confusion, but did not constitute meaningful differences. Competition and consumer confusion are inherently antithetical.

On the other hand, the consumer is also denied access to certain policies that many would judge legitimate. Of particular note is the lack of access to catastrophic-only coverage that would reimburse expenses above a significant deductible, thus offering adequate financial protection to many while being less expensive than any of the 10 standardized plans. In addition to increasing consumer choice, a catastrophic-only option would have a favorable effect on the Medicare budget by retaining patient cost-sharing, which in turn restrains health services utilization.

Another issue relates to the inconsistencies in regulation between medigap and Medicare HMO coverage, a matter that was not addressed as part of the deliberations leading to the OBRA 1990 medigap reforms. Medicare HMOs compete directly with medigap plans. The confusion that now exists will be exacerbated by both the growth in HMOs with Medicare risk contracts and the recently issued HCFA regulations allowing HMOs to market Medicare “point-of-service” products. These products pay for services rendered by providers that are not part of the HMO's provider network, with the enrollee facing higher cost-sharing than if they obtain services outside of the network. Each HMO can decide the circumstances under which it will allow access to non-network providers as well as the associated levels of cost sharing.

Other disparities also exist. For example, medigap carriers are not required to have open enrollment beyond the first 6 months of Medicare eligibility (and, for the disabled under 65 years of age on Medicare, for the 6 months after they turn 65). In contrast, HMOs are precluded from rejecting anyone based on health status and, also, are precluded from excluding pre-existing conditions. As another example, medigap plans are allowed to age rate, i.e., vary premiums based on the age of the enrollee, whereas HMOs must charge a flat premium.

Finally, one of the topics that arose frequently in discussions with State officials and beneficiary representatives was the need for stability in medigap regulation and in the Medicare program itself. The objective of stability competes with that of continuously improving Federal programs and responding to changing circumstances. However, the enactment of Medicare catastrophic coverage in 1988, its repeal 1 year later, and the further enactment of the OBRA 1990 reform legislation caused considerable confusion among beneficiaries. The desirability of any future changes in the structure of the 10 standardized plans needs to be weighed against the goal of minimizing beneficiary confusion.