PART II. POTENTIAL EFFECTS OF HIPAA: A REVIEW OF THE LITERATURE
Stephen H. Long and M. Susan Marquis
The Health Insurance Portability and Accountability Act of 1996 (HIPAA) establishes a federal role for regulating the employer group and individual insurance markets (Atchinson and Fox, 1997). The goals of Title I of the legislation are to provide coverage security for those currently insured. Title I guarantees the availability of insurance to all small employers (with 2 or more employees) and assures that individuals who leave employment are able to maintain health insurance coverage. Thus, HIPAA ensures access to insurance for some employer groups and individuals who previously were unable to purchase health insurance or unable to purchase adequate coverage. What effect this will have on the number of uninsured or the price people pay for insurance is, however, a matter of some uncertainty. Moreover, variability among states in existing insurance legislation, and the flexibility that states are given to implement the individual market reforms suggests that the answer to these questions will vary from place to place.
This paper examines the extant literature on issues pertaining to HIPAA and the general insurance access problems it was designed to address. The goal of the review is to generate hypotheses about the likely effects of HIPAA, the magnitude of those effects, how the effects are likely to vary between population groups, and how background, policy, or implementation characteristics might influence the magnitude of the effects. These hypotheses and the identification of mediating factors will inform the development of a plan for evaluating HIPAA.
In the next section of this paper, we provide a brief overview of the provisions of HIPAA. This is followed by a discussion of the potential effects of these provisions in the group market and the individual market. To estimate these effects, we searched the literature in research journals, the employee-benefits trade press, and reports of private and public organizations for information about HIPAA and the provisions it contains, including: guaranteed issue, guaranteed renewal, portability, pre-existing condition exclusion limits, and non-discrimination provisions. We also reviewed literature on the performance of state high-risk pools, one of the alternative mechanisms that states are allowed to adopt to meet the HIPAA provisions for the individual market reforms. We conclude with some summary observations about the implications for an evaluation design.
The specific provisions of HIPAA related to the reform of the group market and the individual market are summarized below.
Group market reforms. The group market reform provisions that apply to all group plans, including self-insured plans, limit pre-existing condition exclusions and prohibit exclusion of individuals from a group health plan based on health status. In addition, health insurance issuers are required to guarantee renewability of coverage for all groups and guarantee issue all products for small groups.
Preexisting conditions exclusions for all group plans (including insured and self-insured plans) are limited to 12 months (18 months for late entrants) for conditions treated or diagnosed in the prior 6 months. Moreover, group health plans must credit prior public or private insurance coverage toward preexisting condition periods, provided the coverage has not lapsed for more than 63 days (group-to-group portability).
HIPAA prohibits all group health plans (insured or self-insured) from denying coverage or charging higher prices based on health status. It requires that health insurance companies guarantee renewal of all group plans, to both large and small groups. It also requires insurance companies that offer health coverage in the small group market (2-50 employees) to make all products available to all applicants (guaranteed issue). All small groups must be accepted and all eligible members of a group must be accepted.
Individual Market Reforms. Unless the state implements an "acceptable alternative mechanism," insurers in the individual market are required to guarantee issue and apply no preexisting condition exclusions to individuals who had 18 months of continuous coverage, the most recent under a group plan, within 63 days of obtaining new coverage (group-to-individual portability). The individual also must have exhausted all other sources of coverage (including COBRA coverage, other group coverage, Medicare or Medicaid). Individual health insurers must guarantee issue at least two different individual insurance products. These products may be either a) the two highest volume products, or b) a low and high policy option that represent individual policies offered by the insurer in the state that are subsidized, risk-adjusted, or covered by a risk-adjustment mechanism. The guaranteed renewal provisions of the law for plans issued to groups also apply to plans in the individual market.
State Implementation. HIPAA provides for substantial state flexibility in implementing the intent of the individual market reforms to ensure that those who leave a group health insurance plan are able to maintain coverage and are not denied individual insurance if that is the only coverage option available. States may adopt an "acceptable alternative mechanism" to the federal provisions outlined above. To meet the requirements of an "acceptable alternative mechanism", a state program must provide all eligible individuals with a choice of coverage, including one providing comprehensive benefits, and not impose pre-existing condition exclusions on them. For example, 22 states will use high risk pools to meet the requirements, in most cases expanding on an existing risk pool (IHPS, 1998). Other state alternative programs include mandatory group conversion policies, or guaranteed issue of designated individual policies.
EFFECTS OF HIPAAS GROUP MARKET PROVISIONS
HIPPAs group market provisions were designed to eliminate insurer practices that discriminate against employer groups and their members on the basis of health status, industry, or other characteristics. Redlining (that is, excluding specific types of businesses from coverage), denying coverage to employees with poor health or to their entire group, and excluding coverage for pre-existing conditions are documented practices that can pose barriers to employers (especially small employers) that wish to offer health insurance as a benefit (McLaughlin and Zellers, 1994; Zellers, McLaughlin, and Frick, 1992). HIPPA eliminates the first two practices in the small group market, and limits the exclusion on pre-existing conditions for all group plans. Proponents of these reforms believe that they will expand coverage; critics argue that they will lead to increases in premiums that might in turn lower access to coverage (Atchinson and Fox, 1997).
The research evidence on the effect of guaranteed issue, guaranteed renewal, guaranteed coverage for individuals, and pre-existing condition limits in group plans is scarce. But it suggests that these provisions have only small effects on premiums and coverage. Moreover, many states had already taken steps to eliminate discriminatory carrier practices and have standards that meet or surpass HIPAA standards. Thus, HIPAA imposes few new requirements in the group market. We review this evidence below.
Effects on Access
Many states (38) had some form of guaranteed issue legislation for small groups at the time HIPAA was enacted (Litow and McClelland, 1997). Of these, 16 require guaranteed issue of all insurance products and 22 have statutory plans that are guaranteed issue. In about 12 of these states, however, the state law does not encompass all small groups covered by HIPAA.
In the 24 states that do not guarantee issue insurance products to small groups or have guaranteed issue legislation that does not encompass all groups from 2 to 50, HIPAA may result in changes in the number of firms that offer health insurance. The research evidence on the effect of this type of reform is very limited because most of the states small group reform legislation is relatively recent. One empirical study using state data from 1989-1993 suggests that guaranteed issue legislation may result in an increase in the number of firms offering insurance (Jensen et al., 1995). This studys estimate implies about a 10 percentage point increase in the number of small firms offering insurance in states in which availability is guaranteed, holding constant other regulations and market characteristics. Another recent empirical study of the effect of insurance market reforms on states uninsured rates found that guaranteed issue, when part of a package of small group reforms including portability and limits on pre-existing conditions, reduces the uninsured rate (Marsteller et al., 1998). However, this conclusion is likely to vary from state to state depending on other state regulations concerning the health insurance market. HIPAA for example, guarantees access to insurance for all small groups, but it does not guarantee affordability. Whether guaranteed access leads to additional group insurance purchases is likely to depend on whether and what type of rating restrictions the state insurance legislation places on premiums that insurers can charge small groups. For example, Marsteller and her colleagues (1998) suggest that there are no changes in the uninsured rate when rating restrictions are combined with the reforms including guaranteed issue.
Most states (30) have legislation for small group plans that meets or surpasses the HIPAA limits on pre-existing conditions periods that can be imposed (Ladenheim, 1996). Many of these state regulations also credit prior coverage in determining these periods (GAO, 1995). However, the HIPAA limits on pre-existing conditions and group-to-group portability would also apply to large businesses, which are typically not subject to state regulations, and to self-funded plans, which are exempt from state regulation. The GAO (1995) estimates that about 12 million employees and almost 7 million dependents of employees may benefitin terms of reduced or eliminated waiting periodfrom the HIPAA group-to-group portability provisions (GAO, 1995).
While potentially benefiting many current plan enrollees, the provisions are unlikely to lead to an increase in the number of firms offering insurance. Jensen and her colleagues (1995) found no evidence that restricting pre-existing condition waiting periods increased offer rates among small employers. Similarly, the results from the Urban Institute study (Marsteller et al., 1998) found a small, and only marginally significant, effect of limits on pre-existing conditions and portability if guaranteed issue is not part of the package. In larger firms, the HIPAA provisions would benefit only a fraction of employees and so have limited influence on the group decision, assuming the preferences of the median worker dominate (Goldstein and Pauly, 1976). Moreover, offer rates in large firms (those with 50 or more employees) currently exceed 90 percent (Cantor, Long, and Marquis, 1995).
While we believe it is unlikely that the pre-existing condition limits and portability will increase the number of firms offering insurance, the absence of waiting periods may induce some persons to enroll in group plans that they otherwise would turn down. Research indicates that individuals are more likely to purchase insurance when the expected benefit of the plan is greater (see for example, Marquis and Holmer, 1996). This increase, however, is also expected to be small, because about 90 percent of workers offered group insurance enroll in the plan (Long and Marquis, 1993; Cooper and Schone, 1997).
Portability may also benefit individuals who wish to change jobs but who currently stay in their job because of fear of losing insurance. We examine this benefit of the HIPAA reforms in more detail below.
The non-discrimination provisions also are unlikely to substantially affect firm offer rates or employee enrollments. Health insurance offer rates by small employers are not significantly different in states that have limits on the exclusions allowed under pre-existing condition clauses nor are they significantly different in those that have laws prohibiting occupational exclusions from coverage (Jensen, Morrisey, and Morlock, 1995). Medical underwriting, the practice of excluding workers from coverage if they have specific preexisting conditions, occurred in both large and small employer group plans, though it was more common in the latter. However, it was rarely practiced; fewer than 5 percent of firms report that specific individuals are excluded from coverage (Cantor, Long, and Marquis, 1995).
Finally, guaranteed renewal is not an effective guarantee of access to coverage without limits on premium increases (Blumberg and Nichols, 1996). HIPAA does not provide for affordable coverage; it only ensures the continued right to purchase a plan.
Several state-specific experiences after adopting a package of regulations similar to the HIPAA provisions also suggest that there will be at most modest access effects, and that other market factors and regulatory conditions may influence outcomes. California enacted a package of reforms, effective in 1993, for firms with 3-50 employees that included guaranteed issue, guaranteed renewal, limits on pre-existing conditions, and restrictions on premium variability. In the post-reform period, about 10 percent more firms with 3-24 employees offered insurance than in the pre-reform period (Buchmueller and Jensen, 1997). No change in offer rates were found among firms with 25-50 employees, however. Minnesotas experience is similar to Californias. Minnesota also enacted a package of reforms for the small group market that included guaranteed issue and renewal, and limits on pre-existing condition exclusions. After reform, the number of enrollees in the small group market increased by more than 8-12 percent (Nichols et al., 1997).
However, Buchmueller and Jensen (1997) did not find increases in offer rates among small firms in a group of states that adopted reforms similar to Californias during the same period. They concluded that intense competition in the California health market that exerted downward pressure on prices may have provided an environment in which the access reforms were successful in providing a small expansion in coverage. Observers of Minnesotas market draw similar conclusions (Nichols et al., 1997). Moreover, Minnesotas reforms included a provision prohibiting individuals from purchasing individual policies on their own if they were eligible for group coverage. It is believed that some of the increase in the group market is a move from the individual market (IHPS, 1995).
Effects on Premiums
The HIPAA group provisions ensure access to coverage for groups and individuals in those groups, but also place some limits on insurers ability to segment risks. Insurers cannot charge higher prices to high risk individuals in a group; they cannot exclude entire groups from the risk pool. Risk segmentation results in lower premiums for the healthy than the sick; greater pooling of risks may result in adverse selection and increasing premiums that will cause the healthy to drop coverage. However, HIPAA does not place any restrictions on the manner in which insurers can set premiums, and so insurers retain substantial ability to segment the market (Blumberg and Nichols, 1996). Thus, the HIPAA provisions are unlikely to result in substantial premium increases for those individuals currently purchasing insurance. The Health Insurance Association of America estimates that guaranteed issue provisions have only a small impact on premiums2 to 4 percent (Thompson, 1992). The empirical evidence supports this. Jensen, Morrisey, and Morlock (1995) found no evidence that guaranteed issue, pre-existing condition limits, or laws limiting exclusions on the basis of condition or occupation resulted in premium increases. Moreover, an analysis of claims experience from a large insurer specializing in the small group market found no difference in average claims for groups that were guaranteed issue and those that were medically screened (Glazner et al., 1995)
However, all states that have enacted guaranteed issue provisions have also adopted some form of rating restrictions for small groups, and observers believe that other states are likely to adopt rating reforms for small groups as they modify their laws to meet the guaranteed issue provision of HIPAA (Litow and McClelland, 1997). Rating reforms are intended to enlarge the risk pool on which premiums for small employers are based, thus making insurance more affordable for high risk groups and encouraging them to purchase insurance. On the other hand, small employers that have a good risk profile (for example, predominantly young workers) may pay higher rates after reforms are enacted because insurers will be limited in the extent to which they can take into account this favorable profile. The effect of rating reforms on premiums will therefore vary from company to company depending on the characteristics of the group.
On average, premiums may rise as well, as higher risk groups are attracted into the market and the lower risk groups discouraged from purchase. Thus, rating reforms in combination with other market reforms may increase the premiums for small businesses that purchase insurance, and lead some of them to drop coverage. There are other forces, however, which might lead premiums to fall with a change in market regulations and rate restrictions. Reforms may encourage greater price competition because insurers are limited in the extent to which they can compete on the basis of risk selection (Buchmueller and Jensen, 1997). Reforms may also improve information and lead to greater shopping.
The overall effect on premiums and access is likely to depend on how stringent the rating reforms are. Under full community rating, the insurer can vary premiums only by family type, geographic location, and benefits design. Only New York currently requires full community rating. Modified community rating reforms (or community rating by class), which allow insurers to vary premiums by a limited number of characteristics of the enrollees (such as age), are more common but do not permit health status or prior claims experience to be factored into the pricing. Some states limit the variation in premiums attributable to these factors. Other states have enacted rating reforms that permit use of health status in setting premiums, but limit the differentials in premiums allowed due to health status differences.
The American Academy of Actuaries (1993) conducted simulation analyses that suggest that average premiums for small businesses would rise by only about 2 percent with modified community rating and by about 5 percent under full community rating. There is little empirical evidence on the effects of rating reforms from states that have enacted them. What little evidence exists is consistent with the analysis of the Academy suggesting that rating reforms result in only modest increases in average premiums. For example, premiums in New York in the small group market rose about 5 percent during the first year that community rating was in effect (Chollet and Paul, 1994). Minnesota, which adopted restrictions on premium rate variations, also experienced premium rate increases of less than 5 percent in the year after it enacted these rating reforms in combination with a number of other small group reforms (Blumberg and Nichols, 1996). In California, premiums fell slightly after the small group reform legislation (Buchmueller and Jensen, 1997). Again analysts attribute the California experience to the highly competitive nature of the market.
Average premium changes of this magnitude are unlikely to lead to significant changes in the number of employers offering insurance. The published research literature suggests that a 5 percent increase in premiums would lead to a reduction of between 10 to 15 percent in the number of employers offering health insurance (Leibowitz and Chernew, 1992; Jensen and Gabel, 1994; Morrisey, Jensen and Morlock, 1994). However, this literature uses proxy measures of the price faced by those who do not currently offer insurance to estimate the response. Experimental evidence suggests that the response is much lower. A number of demonstration studies of the effect of subsidies to small groups on insurance purchase decisions indicate that premium changes up to 50 percent lead to changes of less than 10 percent in the number of firms offering insurance (Thorpe, et al., 1992; Helms et al., 1992). Based on these experimental results, we would expect only about a 1 percent decrease in the number of firms offering insurance in states that adopt a comprehensive new package of small group reforms to meet the HIPAA requirements.
Even if the average price does not change, states that adopt rating reforms along with other group regulatory changes may see substantial changes in the prices faced by individual firms. For example, the American Academy of Actuaries (1993) estimates suggest that about 20 percent of firms would realize premium increases of 20 percent or greater and 20 percent of firms would realize premium decreases of 20 percent or greater under modified community rating. This, in turn, might lead to changes in the firms that do and do not offer insurance, even if the overall numbers remain steady. Over time, however, rating reforms lead to greater stability in premiums faced by individual firms and so on offer decisions (Buchanan and Marquis, 1997).
HIPAA imposes new requirements that may increase the administrative costs of insurance. HIPAA requires employers to provide certification of creditable coverage to employees and their dependents as they leave a group plan; in subsequent contracts, employers are likely to place this responsibility on insurers (Finlay, 1996). Insurers view this as costly and unnecessary, favoring certification on demand. They note that state portability rules have been successfully implemented without global certification requirements (GAO, 1998). The necessity of certifying periods of coverage for spouses and dependents may pose special problems and require costly new record keeping systems. Employers and insurers currently do not have records to track this coverage (Humo, 1997). In addition, there may be other new costs to employers or insurers as their agents in oversight and adherence to the regulations concerning waivers of coverage and timely enrollment of HIPAA eligibles (Deru, 1997).
Some writers also suggest that carriers will adopt some new practices in response to the guaranteed acceptance provisions of HIPPA that may raise administrative costs. They observe that insurers are more closely assessing the risk of an entire group using medical questionnaires for all employees, even if not immediately eligible for coverage, to identify the potential risk of the group should these employees later seek coverage (Deru, 1997, Grosjean, 1997). This practice potentially increases the insurers risk assessment costs. Carriers are also more strictly enforcing contract guidelines concerning employee participation and employer contributions (Deru, 1997; Grosjean, 1997), which also raises administrative costs. With guaranteed issue, it is especially important to spread the risk as much as possible, which requires enrolling as many people in the group as possible. Although these new administrative costs are important to the groups affected by them, they are unlikely to lead to significant changes in the overall price of insurance.
Finally, HIPPAs non-discrimination provisions prohibit carriers from raising rates for particular individuals in a group because of health. This may lead to increases in the price charged for all members of the group. However, guaranteed renewal provisions adopted in 43 states place some limits on these price increases (Landenheim, 1996).
Some believe the main benefit of the HIPAA group market reforms is that it relieves individuals from worry about gaps in coverage as they change jobs (Clark, 1996). And many believe that such worry now impedes people from leaving current jobsreferred to as "job-lock"and leads to an inefficient distribution of labor. The empirical evidence on the importance of job-lock is mixed, however. A number of studies have found evidence of lower job-turnover as the value of ones own employer-provided insurance increases (Madrian, 1994; Cooper and Monheit, 1993; Buchmueller and Valetta, 1996). These studies suggest that insurance reduces turnover by about 25 percent. Gruber and Madrian examine turnover in states that do and do not have continuation of coverage mandates, and conclude that one year of continued benefits increases turnover by about 10 percent among those with employer health insurance. In contrast, several other investigators have found little evidence to support job-lock (Holtz-Eakin, 1994; Kapur, 1998).
The effects of HIPAA on job mobility and the distribution of labor are likely to be minor, however, even if job-lock exists. The Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA) requires that firms with 20 or more employees permit employees and their dependents to continue to purchase the companys health insurance for up to 18 months after their job is terminated. Although employees must pay the premium for COBRA coverage, this coverage should have substantially eliminated job-lock for most workers and their dependents in firms of 20 or more, especially if job-lock stems from short term medical considerations such as pregnancy rather than longer horizon issues about being medically underwritten. But in addition, 40 states already have adopted group-to-group portability regulations for small groups requiring that periods of coverage under a previous employer health insurance plan be credited toward waiting periods for coverage under a new employers plan (GAO, 1995).
HIPAA forbids discrimination based on health status. There is some uncertainty about the implications of this rule for plan benefit design. Some observers believe that the rule does not prohibit exclusions or caps on coverage of particular treatments, nor different copayments or deductibles for specific conditions, nor mandatory case management for specific health problems (Barker and OBrien, 1997; Schwappach and Sipes-Early, 1996). However, certain wellness programs may need to be redesigned, especially those that appear to impose a premium surcharge on those in poorer health (Barker and OBrien, 1997). But, ambiguities in the law, litigation, and subsequent case law may lead to further constraints on benefit design (Barker and OBrien, 1997).
There are also some early indications that some issuers may be using benefit design to circumvent HIPAAs restrictions on pre-existing condition limits for those leaving a job or losing group coverage. There is anecdotal evidence that some insurers are establishing waiting periods for coverage for certain costly conditions from all enrollees coverage, which effectively precludes pre-existing conditions from coverage (GAO, 1998).
EFFECTS OF GROUP-TO-INDIVIDUAL PORTABILITY PROVISION
The largest controversy surrounding HIPAA relates to the probable effects of the individual insurance market reforms. Critics of these provisions argue that they would lead to steep price increases in the individual market and an eventual erosion of the individual market. Proponents believe they will guarantee continued access to insurance at little cost. Below we discuss estimates of the effects of the group-to-individual portability provisions and lessons that can be learned from state experiences with individual market reforms. We organize the discussion around two implementation mechanisms: guaranteed issue of some or all products and high risk pools.
Effects of Guaranteed Issue
Access and Premiums
Widely discrepant estimates of the effects of the group-to-individual portability provisions of HIPAA on access and premiums in the individual market have been offered. The Health Insurance Association of America, for example, estimated that the HIPAA individual market provisions would lead about 1.6 million new individuals to purchase individual insurance each year (HIAA, 1995). In the long run, this would lead to an increase of 28 percent in the number of covered lives in the individual market. These new entrants are assumed to have higher spending that those currently insured, otherwise they would likely participate in the current market. Assuming spending levels that are twice that of the currently insured, HIAA estimates that prices in the individual market would rise by about 25 percent (HIAA, 1995). This price increase in turn might discourage some who currently purchase in the individual market from doing so. As healthier individuals leave the market, premiums would rise further potentially leading to a collapse in the market.
However, others have concluded the effects would be much smaller. For example, analysis at RAND (Klerman, 1996) indicates that the HIAA estimates substantially overestimate the number of individuals who exhaust their COBRA benefits and thus become eligible under the HIPAA provisions. The RAND analysis also suggests that the HIAA estimates overstate the number of individuals leaving insured jobs at smaller employers who are eligible for HIPAA because they assume that all have 18 months of creditable coverage. The RAND analysis suggests that the increase in covered lives in the individual market would be only about 6 percent, and premiums would rise by only about 6 percent.
The estimates are also sensitive to assumptions about the health risk of the newly eligible population. Evidence from states that have continuation of coverage requirements indicates that claims cost for those who enroll in such coverage is about 150 percent of the average claims cost of members of the group (Gruber, 1994). Combining this estimate of the cost of the new entrants in the individual market with the RAND estimates of the number of entrants, premiums are estimated to rise by only 2 percent (Klerman, 1996).
Moreover, the implications of these estimates of premium increases depend upon states actions regarding rate regulation in the individual market. Insurance companies worry that states will impose restrictive rate regulation on the individual market as they pass new legislation to conform to HIPAA (Findlay, 1997). With restrictive rate regulation, those currently insured in the individual market would see their premiums increased as the new, and less healthy, HIPAA eligible individuals enroll. In response, some of those currently insured might leave the market. Because the healthy would be more likely to insure, this would lead to further increases in premiums and flight from the market by the healthier. Some fear that this would then lead carriers to eliminate individual coverage from their portfolios (Deru, 1997: Litow and McClelland, 1997).
However, to date, few states have adopted new rate regulation since HIPAA was enacted. Absent rate regulation, carriers are likely to segment the conversion policies and set separate, and higher, rates for these policies. This may lead to prohibitively high premiums for these policies and discourage the new HIPAA eligibles from purchasing coverage . Early results from the states that adopted the Federal fallback suggest that this, in fact, may be happening in some states (GAO, 1998). However, under this scenario, premiums for the currently insured would not be expected to increase.
The experience in states that had adopted guaranteed issue and rating reforms prior to HIPAA provides some confirmation of fears that imposing very strict pooling in this market too quickly might harm the individual market. The Urban Institute analysis of the effect of insurance market reforms on uninsured rates using the Current Population Survey (CPS), found some evidence that guaranteed issue and rating restriction in the individual market results in an increase in the uninsured rate (Marteller et al., 1998). However, this result is based on observations in a limited number of states. The Galen Institute (Schriver and Arnett, 1998) also used CPS data to compare uninsured rates in 16 states they classified as having stringent individual and group market reforms with other states. Uninsured rates in the former increased by 2 percentage points more than in other states between 1990 and 1996, leading the authors to conclude that the reforms had unintended effects. However, there was considerable variability among the 16 states in the magnitude of the increase, and overall uninsured rates in the tightly regulated states remained below those in the other states. Turning to some state specific experiences, premiums for individual coverage in Kentucky rose more than 60 percent following the enactment of guaranteed issue and community rating for the individual market, driving healthier insured individuals out of the market (Page, 1997a). Carriers faced losses when left with the sicker individuals, and 40 companies left the market. Similar problems were encountered in New Hampshire, leading the largest individual insurer in the state to withdraw (Page, 1997b). Washington also saw a number of insurers leave the state following individual insurance market reforms, and remaining insurers are reported to be incurring large losses (Chollet and Kirk, 1998). New York also saw a reduction in the number of individual policies following strict rating reforms and anecdotal evidence suggests that it was younger, healthier individuals who left the market (Chollet and Paul, 1994; IHPS, 1995). However, few commercial insurers left the New York market, despite threats to do so (IHPS, 1995). Moreover, it is difficult to attribute the changes in the individual market in New York to community rating because Blue Cross and Blue Shield, the dominant carrier prior to the reforms, practiced community rating prior to the New York legislation (Chollet and Paul, 1994).
Other states have implemented guaranteed issue along with rating bands in the individual market; there appear to be less disruptive effects when insurers are not required to adopt complete pooling. While there is little hard data, insurers in Vermont report growth in their business and little adverse selection (Chollet and Paul, 1994). Few insurers left the state following the reforms. Minnesota, which also adopted rating bands rather than strict community rating, reports some fall in participation in the individual market (IHPS, 1995). But as noted earlier, some of this is believed to be a movement to the group market. Some carriers have left the market, but the impact was limited because they held a small market share (IHPS, 1995).
In sum, the effect on both access and premiums in the individual market will depend on how states implement the portability provisions of HIPAA and on how they regulate the marketespecially, whether and how they restrict premiums for individual insurance coverage. Ten states already had guaranteed issue and portability in the individual market, and all of these had some rate regulation (GAO, 1996). HIPAA would not be expected to have additional effects in these markets. Twenty-two states and the District of Columbia have introduced risk pools or modified existing risk pools to meet the HIPAA requirements, and so HIPAA will not directly effect premiums or coverage in the individual market. Furthermore, in the remaining states, the effect of the legislation on premiums for current policies will depend on whether the state has or introduces rate regulation, since insurers would be expected to price policies to the new entrants separately if free to do so.
Employment. The availability of post-retirement health insurance is a significant factor in the decision to retire. Employer-provided retirement health insurance reduces the age at retirement by 6 months to 2 years (Madrian, 1994; Blau and Gilleskie, 1997). In addition, the probability of retirement is greater among older workers who have access to mandated continuation coverageeither COBRA or state mandated continuation benefits (Gruber and Madrian, 1995). Thus the group-to-individual portability provisions of HIPAA may affect the supply of labor. The magnitude of this effect is not known, but will likely vary depending on the type of insurance products that are available and state regulations to control their price.
Some have also suggested that portability will enhance entrepreneurial activity, and that the employment-based health insurance system discourages individuals from leaving firms to start their own business. While there is limited empirical work, the evidence does not support this hypothesis (Holtz-Eakin, Penrod, and Rosen 1996).
Benefit Design. The new legislation may also affect the characteristics of products available in the market if they affect carriers decisions to participate in the market or alter product lines. Plan design can be a technique to segment the market; for example, plans with high cost sharing are likely to be more attractive to healthier individuals than to sicker individuals. There is some evidence that insurers have turned to plan design as a competitive response in states that have adopted community rating (IHPS, 1995). Plan design changes may also be used to hold down premium increases (IHPS, 1995). These responses will vary depending on whether the state regulation imposes strict pooling through rating reforms.
Administrative Costs. Several states failed to pass conforming legislation and the federal government will be responsible for oversight of HIPAA implementation. The state, however, will continue to enforce existing law. This may lead to administrative duplication and added administrative costs (Aston, 1997a).
Effects of Risk Pools
Twenty-two states and the District of Columbia have adopted state risk pools as an alternative mechanism. Many of these risk pools were in operation prior to HIPAA, though some characteristics of the pools have to be modified to meet the group-to-individual portability provisions. For example, eligibility needs to be extended to all HIPAA eligibles and waiting periods for pre-existing conditions need to be eliminated for those with 18 months of creditable prior coverage. Some pools also have to expand the choices available to serve as an alternative mechanism (Aston, 1997b).
State high risk pools for individuals who are unable to obtain insurance have been in existence since Connecticut and Minnesota introduced the first pools in 1976 (Bovberg and Koller, 1986). The experience of these pools suggests that they are unlikely to attract large numbers of new HIPAA individuals. We summarize the evidence on the effects of these pools on access below, and discuss some other potential effects for states that introduce new pools.
Access and Premiums
Existing high risk pools are a form of guaranteed issue to individuals (Blumberg and Nichols, 1996). They provide a source of coverage for individuals who wish to purchase individual insurance and are willing to pay for it, but are denied coverage or offered coverage that is prohibitively expensive or restrictive. However, the evidence suggests that most of these pools cover only a small share of the target population. In 25 of the states with pools, fewer than 5 percent of persons with individual insurance are enrolled in the pool (Gao, 1996). Assuming the uninsurable population is about 1 percent of the under 65 population in a statea standard metric (Laudicina, 1988)1994 enrollment data for the 25 state pools that were operational indicate that most of the pools reach only about 5 to 25 percent of the target population (enrollment data from Stearns et al., 1997). In contrast, about 40 to 50 percent of workers who do not have access to group coverage purchase individual insurance (Marquis and Long, 1995; Marquis and Buchanan, 1992). So, assuming preferences among workers lacking insurance are similar to preferences of persons unable to purchase individual insurance, the low rate of participation in risk pools suggests that there remain barriers to enrollment in risk pools.
The premiums in risk pools appear to be at least one such barrier. Table 1 shows premiums and enrollments in 8 risk pools studied by Stearns and Mroz (1995). Overall, enrollments appear to be inversely related to premium levels.
Table 1. Premiums and Enrollments in Eight State High Risk Pools
Source: Stearns et al., 1997; Stearns and Mroz, 1995.
a. Quarterly premium for 50 year old male.
b Target set at 1 percent of the uninsured population under age 65
c 1992 premium
d 1990 premium
e 1989 premium
By their very nature, risk pools start out with adverse selection; persons who are able to purchase individual insurance at a lower rate than in the pool would choose to do so. The premiums in risk pools are subsidizedthey are typically capped at a fixed percentage (usually about 150) of rates charged by private insurers for standard risks. Nonetheless, Table 1 suggests that high premiums in many plans are a barrier to participation and more rigorous analysis of disenrollments from risk plans also shows that price is a factor in participation (Stearns and Mroz, 1995). Thus, unless states subsidize risk pools more heavily than in the past or attract a broader risk pool to lower premiums, it is unlikely that many new HIPAA eligibles will enroll in such plans. Most risk pools allowed limited pre-existing condition exclusions, which will have to be waived for HIPAA eligibles if the risk pool is to serve as an alternative mechanism. This may increase the pools attractiveness to some individuals. However, in states with an existing risk pool, most HIPAA eligibles were previously eligible to participate in the pool, or were able to buy private insurance at a lower price than offered by the pool but were unwilling to pay the price. The experience of existing risk pools suggests that new pools will attract only a small share of the target population unless prices are lower.
States with existing risk pools will have to modify eligibility criteria for the pool to serve as an alternative mechanism. An unanswered question is whether broader eligibility rules will serve to attract a broader risk pool, thereby keeping premiums down and so increasing enrollments among those who might otherwise go uninsured. Without premium restrictions, private insurers will offer low prices to better risks, and so risk pools are likely to continue to attract the less healthy. However, risk pools have lower administrative costs than typical private insurance plans (Stearns et al., 1997), which may serve as a partial offset to private insurers ability to attract lower risks through lower premiums. The state risk pool in Connecticut provides an illustration of the experience of an open risk pool. Historically, the Connecticut risk pool was open to any resident. Yet it appeared to attract a less healthy population. The premiums in the pool were as high as or higher than premiums in other states, though it did not have lower loss ratios (Stearns et al., 1997 and Stearns and Mroz, 1995). In addition, the Connecticut pool attracted a disproportionate number of persons with severe mental health problems (Stearns and Slifkin, 1995). Thus, even with broad eligibility requirements, state risk pools appear to experience adverse selection. Premiums are thus likely to remain high, and so enrollments are likely to remain low.
Risk pool premiums are not designed to fully cover premium and administrative costs and so the pools realize losses. The most common way to finance these is by an assessment on insurers (Stearns et al., 1997; Sneider, 1991; Bell, 1997). Most of this cost is then passed on to businesses and individuals in the form of higher prices for insurance (Bell, 1997; Sneider, 1991). New assessments on commercial carriers to finance risk pools may also lead some of them to withdraw from the market (Blumberg and Nichols, 1996).
Some view pools as a major loophole in HIPAA (Lieberman, 1997). Critics point to their insolvency. Some also argue that they constrain choice, which was one of the intentions of the act (Atchinson and Fox, 1997). Some also fear that many self-employed will find their premiums raised because they will be forced to join risk pools of less healthy individuals (Atchinson and Fox, 1997).
Most states had adopted small group market reforms prior to HIPAA that met or exceeded the provisions of the HIPAA legislation. Thus, we expect limited effects nationally on access, coverage, and premiums in the group market. There may be greater effects in the few states that had not adopted market reforms. But even in these states, the literature suggests that the effects of guaranteed coverage are likely to be small without price reforms. The implication for evaluation design is to focus on selected states and populations within them that are most likely to benefit from the reform, such as persons in high risk industries or occupations and employers with a larger share of older workers. The key outcomes to monitor include the availability of group coverage, enrollment in group coverage, the cost of group coverage, and job mobility.
Evaluation requires a comparison group as well to control for secular change. This is especially difficult because we would like to compare what would have happened over time without reforms to the changes that we observe in states in which the HIPPA provisions lead to new group reforms. We cannot observe the former, but must rely on change over time in states that had implemented reforms prior to the HIPAA legislation as control groups. However, reforms are likely to affect premium growth and the stability of the market, and thus the change we observe in reform states is only an imperfect proxy for the control measure that we would like to have.
While the individual market reform provisions of HIPAA were much more controversial than the group market reforms, the evidence also suggests that these reforms are likely to have limited effect on the number of uninsured and the premiums they would have to pay for care. Many states already have high risk pools that will, with minor changes, meet the group-to-individual portability provisions of the law. A number of other states already had individual market reforms that provide portability. Again, careful selection of states and subpopulations is indicated for the evaluation. For example, one would want to include states that modified existing risk pools, states that have adopted new risk pools, states with new guaranteed issue provisions, and states that had prior individual market reforms. Focusing on access, coverage, and premiums for high risk individuals leaving the group market (e.g., early retirees, those in poor or fair health) would increase the likelihood of finding possible effects.
It is not surprising that we expect limited quantifiable effects. HIPAA was targeted to address the most important insurance abuses and not to ensure affordable coverage (Lieberman, 1997). But, there are many implementation issues and problems facing employers, insurers, and regulators that also need to be addressed in the evaluation. One observer notes that HIPAA "may appear modest in scope to public policy makers, but it is anything but simple for the private sector" (Findlay, 1997). Monitoring employer and insurer efforts to document creditable coverage, insurer benefit design practices and the characteristics of products available to HIPAA eligibles, and state practices to insure compliance with HIPAA requirements, are all important issues that will need to be addressed as a part of a qualitative assessment of HIPAA.
American Academy of Actuaries, "An Analysis of Mandated Community Rating", Monograph 3, March 1993.
Aston, G., "States Revamp Laws to Meet New Federal Rules", American Medical News, v.40, n.33, 1997b, pp.29-30.
Aston, G., "Insurance Reform Moves To the States", American Medical News, v.40, n.5, 3 February 1997a, pp.3-4.
Atchinson, B. and D. Fox, "The Politics of the Health Insurance Portability and Accountability Act", Health Affairs, v.15 n.3, May/June 1997, pp.146-150.
Barker, R. and K. O'Brien, "Nondiscrimination Rules Under HIPAA: Limits or License for Health Plan Design?", Benefits Law Journal, v.10 n.1, Spring 1997, pp. 5-29.
Bell, A. "Texas Prepares to Launch Health Risk Pool", National Underwriter Life & Health, v.101, n.48, 1 December 1997, p34+.
Blau, D.M. and D. B. Gilleskie, "Retiree Health Insurance and the Labor Force Behavior of Older Men in the 1990s", NBER Working Paper No. 5948, National Bureau of Economic Research, Cambridge, MA, 1997.
Blumberg, L. and L. Nichols, "First, Do No Harm: Developing Health Insurance Market Reform Packages", Health Affairs, v.15 n.3, Fall 1996, pp.35-53.
Bovbjerg, R. and C. Koller, "State Health Insurance Pools: Current Performance, Future Prospects", Inquiry, v.23, n.2, Summer 1986, pp.111-121.
Buchanan, J. L. and M. S. Marquis, "Who Gains and Who Loses with Community Rating for Small Business", paper presented at the RAND Conference on Employer Health Benefits, July 1997.
Buchmueller, T.C. and G. A. Jensen, "Small Group Reform in a Competitive Managed Care Market: The Case of California, 1993 to 1995", Inquiry, v.34, 1997, pp.249-263.
Buchmueller, T.C. and R.G. Valetta, "The Effects of Employer-Provided Health Insurance on Worker Mobility", Industrial and Labor Relations Review, v.49, 1996, pp.439-455.
Cantor, J., S.H. Long, and M.S. Marquis, "Private Employment-Based Health Insurance in Ten States", Health Affairs, v.14, n.2, 1995, pp.199-211.
Chollet, D. and R. Paul, Community Rating: Issues and Experience, Alpha Center: Washington, D.C., December 1994.
Chollet, D. J. and A. M. Kirk, Understanding Individual Health Insurance Markets, Kaiser Family Foundation, March 1998.
Clark, J., "What Health Care Reform Can Give You", Kiplinger's Personal Finance Magazine, v.50, December 1996, p.116.
Cooper, P.F. and A.C. Monheit , "Does Employment-Related Health Insurance Inhibit Job Mobility", Inquiry, v.30, 1993, pp.400-416.
Cooper, P.F. and C. S. Schone, "More Offers Fewer Takers For Employment-Based Health Insurance: 1987 and 1996", Health Affairs, v.16, n.6, 1997, pp.142-149.
Deru, S.E., "Side Effects of Kassebaum-Kennedy Bill", National Underwriter Life and Health--Financial Services Edition, v.101, n.18, 5 May 1997, p.20.
Findlay, S., "Figuring It Out", Business and Health, v.14, n.11, 1 November 1996, pp.21-22.
Findlay, S., "The Heavy Load of Implementing Portability", Business & Health, v.15, n.3, March 1997, p.18.
Glazner, J., W. Brainthwaite, S. Hull, and D. Lezotte, "The Questionable Value of Medical Screening in the Small-Group Health Insurance Market", Health Affairs, v.14 n.2, Summer 1995, pp.224-234.
Goldstein, G. S. and M. V. Pauly, "Group Health Insurance as a Local Public Good", The Role of Health Insurance in the Health Services Sector, R.H. Rosett, ed. New York, NY: NBER, pp.73-107.
General Accounting Office, Health Insurance Portability, GAO/HEHS-95-275, September 1995.
General Accounting Office, Health Insurance Standards: New Federal Law Creates Challenges for Consumer, Insurers, Regulators, GAO/HEHS-98-67, February 1998.
General Accounting Office, Private Health Insurance: Millions Relying on Individual Market Face Cost and Coverage Trade-Offs, GAO/HEHS-97-8, November 1996
Grosjean, H.C., "Portability Act Harms Most Small Employers", Business Journal - Serving Phoenix and the Valley of the Sun, v.17, n.51, 17 October 1997, p.19.
Gruber J., "State-Mandated Benefits and Employer-Provided Health Insurance", Journal of Public Economics, v.55, 1994, pp.433-464.
Gruber, J. and B. C. Madrian, "Health Insurance Availability and the Retirement Decision", American Economic Review, v.85, 1995, pp.938-948.
Gruber, J. and B. Madrian, "Health Insurance and Job Mobility: The Effects of Public Policy on Job-Lock", Industrial and Labor Relations Review, v.48, n.1, October 1994, pp.86-102.
Health Insurance Association of America, "The Cost of Ending Job Lock or How Much Would Health Insurance Costs Go Up if Portability of Health Insurance Were Guaranteed? Preliminary Estimates", July 26, 1995.
Helms, W. D., A. K. Gauthier, and D. M. Campion, "Mending the Flaws in the Small-Group Market", Health Affairs, v.11, n.2, Summer 1992.
Holtz-Eakin, D., "Health Insurance Provision and Labor Market Efficiency in the United States and Germany", in Social Protection Versus Economic Flexibility: Is There a Tradeoff, R.M. Blank, ed. University of Chicago Press, Chicago, pp.157-187.
Holtz-Eakin, D., J.R. Penrod, and H.S. Rosen, "Health Insurance and the Supply of Entrepreneurs", Journal of Public Economics, v.62, October 1996, pp. 209-235.
Humo, T., "Implementing HIPAA: Portability of Coverage and the Future of Health Care Reform", Managing Employee Health Benefits, v.5, n.4, Summer 1997, pp.25-29.
Institute for Health Policy Solutions, Small Group and Individual Insurance Market Regulatory Database, prepared for the Health Care Financing Administration, 1998.
Institute for Health Policy Solutions, State Experiences with Community Rating and Related Reforms, Institute for Health Policy Solutions, September 1995.
Jensen, G., M. Morrisey, and R. Morlock, "The Effects of State Initiatives in the Small Group Insurance Market", presented at a conference entitled "The Rapidly Changing Insurance Market: Policy and Market Forces", sponsored by the Robert Wood Johnson Foundation, March 1995.
Jensen, G.A. and J. R. Gabel, "State Mandated Benefits and the Small Firms Decision to Offer Insurance," Journal of Regulatory Economics, v.4, n.4, December 1992, pp.379-404.
Kapur, K., "The Impact of Health on Job Mobility: A Measure of Job Lock", Industrial and Labor Relations Review, v.51, 1998, pp.282-297.
Klerman, J., New Estimates of the Effect of Kassebaum-Kennedys Group-to-Individual Conversion Provision on Premiums for Individual Health Insurance, RAND, Santa Monica, CA: 1996.
Landenheim, "A Drop in Your State's Bucket", Business and Health, v.14, n.10, 1 October 1996, pp.37-41.
Laudicina, S., "State Health Risk Pools: Insuring the Uninsurable", Health Affairs, v.7, n.4, Fall 1988, pp.97-104.
Leibowitz, A. and M. Chernew " The Firms Demand for Health Insurance," in Health Benefits and the Workforce, U.S. Department of Labor, 1992.
Lieberman, T., "You Can't Take It With You", Columbia Journalism Review, v.35, n.2, 1 July 1997, pp.50-52.
Litow, M. and P. McClelland, "The Impact of the New Health Care Reform Law", Managing Employee Health Benefits, v.5, n.2, Winter 1997, pp.7-12.
Long, S.H. and M.S. Marquis, "Gaps in Employer Coverage: Lack of Supply or Lack of Demand?", Health Affairs, Supplement, 1993, pp.282-293.
Madrian, B.C., "Employment-Based Health Insurance and Job Mobility: Is There Evidence of Job Lock?", Quarterly Journal of Economics, n.109, 1994, pp.27-54.
Madrian, B.C., "The Effect of Health Insurance on Retirement", Brookings Papers on Economic Activity, n.1, 1984, pp.181-232.
Marsteller, J.A., L.M. Nichols, A. Badawi, S. Rajan, and S. Zuckerman, Variation in the Uninsured: State and County Level Analyses, Urban Institute, June, 1998.
Marquis, M. S. and J. L. Buchanan, "Subsidies and National Health Care Reform: The Effect on Workers Demand for Health Insurance Coverage", Health Benefits and the Workforce, Department of Labor, 1992.
Marquis, M. S. and S. H. Long, "Worker Demand for Health Insurance in the Non-Group Market", Journal of Health Economics, v.14, n.1, 1995, pp.47-64.
Marquis, M.S., and M.R. Holmer "Alternative Models of Choice Under Uncertainty and Demand for Health Insurance", The Review of Economics and Statistics, v.78, n.3, 1996, pp.421-427.
McLaughlin, C.G. and W.K. Zellers, Small Business and Health Care Reform, University of Michigan, 1994.
Morrisey, M.A., G.A. Jensen, and R. J. Morlock, "Small Employers and the Health Insurance Market," Health Affairs, v.13, n.5, Winter 1994, pp.149-161.
Nichols, L.M., L.J. Blumberg, G.P. Acs, C.E. Uccello, and J. A. Marsteller, Small Employers: Their Diversity and Health Insurance, Urban Institute, 1997.
Page, L., "Kentucky Poised To Scrap Health Insurance Reforms", American Medical News, v.40, n.36, 22 September 1997a, pp.3-4.
Page, L., "NH Faces Turmoil in Individual Insurance Market", American Medical News, v.40, n.42, 10 November 1997b, p.8.
Schriver, M.L. and G. Arnett, "Uninsured Rates Rates Rise Dramatically in States with Strictest Health Insurance Regulations", The Heritage Foundation Backgrounder, n.1211, August 11, 1998.
Scwappach, K., and K. Sipes-Early, "Health Reform 1996: Implications for Plan Sponsors", Benefits Law Journal, v.9, n.4, Winter 1996, pp. 99-123.
Sneider, J., "State Health Plan in World of Trouble", Business Journal-Milwaukee, v.8, n.13, 7 January, 1991, pp.1-2.
Stearns S. C. and R. T . Slifkin, "State Risk Pools and Mental Health Care Use", Health Affairs, v.14, n.3, 1995, pp.220-231.
Stearns S.C. and T. A Mroz, "Premium Increases and Disenrollment from State Risk Pools", Inquiry, v.32, 1995, pp.392-406.
Stearns, S.C., R.T. Slifkin, K.E. Thorpe, and T.A. Mroz, "The Structure and Experience of State Risk Pools: 1988-1994", Medical Care Research and Review, 1997.
Thompson, R., "States Take Lead in Health Reform", Nations Business, v.80, n.4, April 1992, pp.18-26.
Thorpe, K. E., "Reducing the Number of Uninsured by Subsidizing Employment-Based Health Insurance, Results From a Pilot Study", JAMA, v.267, n.7, 1992, pp.945-948.
Zellers, W., C. McLaughlin, and K. Frick, "Small-Business Health Insurance: Only the Healthy Need Apply", Health Affairs, v.11, n.1, Spring 1992, pp.174-180.
Go to next section of this document.
Return to table of contents.
Exit to Health Policy Home Page