by Brigitte C. Madrian
THE RESPONSE OF THE CLINTON ADMINISTRATION and Congress to the failed attempt at fundamental health care reform in 1993 has been a series of incremental measures that expand the government's role in the health care system. One recent initiative, the Health Insurance Portability and Accountability Act of 1996 (HIPAA), restricts the ability of insurers to exclude preexisting medical conditions from insurance coverage. It also requires insurance companies to issue or renew coverage for individuals who have been insured but who no longer have access to group health insurance coverage. Those provisions are intended to address the issue of health insurance portability-the ability of individuals to maintain health insurance coverage regardless of changes in their employment or personal situation.
Although it is too early to assess directly the effects of the HIPAA, they can be anticipated in part from the experiences of individuals and employers with an earlier health insurance portability law, COBRA, named for the budget bill-the Consolidated Omnibus Budget Reconciliation Act of 1985-to which it was unceremoniously attached. That law mandates that employers who provide a group health insurance plan must offer most employees who lose or leave jobs the option to continue coverage under the employer's plan.
Under COBRA many individuals who might have dropped health insurance when changing jobs have been able to maintain their coverage. COBRA has also allowed some individuals to take advantage of better job opportunities and enabled older workers to retire early. However, those benefits have come at a cost. Companies have been forced to bear the excess costs of both insuring COBRA recipients and of administering the program. Further, while early retirement may benefit older individuals, it results in both reduced economic output and lower tax revenue. COBRA is a good example of how the federal government's idiosyncratic approach to health insurance reform creates new problems even as it attempts to solve others.
| Charles D. Spencer & Associates estimates that in 1997, the annual per beneficiary medical expenditures for COBRA recipients exceeded those for active employees by $2,260. If that figure is multiplied by the number of former employees covered by COBRA, estimated at 5 million, the total direct excess costs to employers is about $11.5 billion per year. |
WHAT IS COBRA?
COBRA essentially confers some degree of health insurance portability on individuals with employer-provided health insurance. The law, which took effect on 1 July 1986, mandates that employers who provide a group health insurance plan must offer employees who would lose coverage because of termination of employment or a reduction in work hours the option to continue their health insurance coverage for up to eighteen months. It allows up to thirty-six months of coverage for spouses and dependent children who lose coverage because of an employee's death, divorce, or eligibility for Medicare; and it also allows for thirty-six months of coverage for children who otherwise would lose coverage because they are too old to maintain their classification as dependents. Those changes in circumstance (job change, divorce, etc.) that would result in a loss of health insurance coverage are termed "qualifying events."
COBRA coverage ceases if the period of eligibility expires, if an individual ceases to pay the required premiums, or if the individual becomes eligible for group health insurance from another source, such as from a subsequent employer; that does not exclude preexisting conditions. Individuals who become eligible for Medicare also lose their COBRA eligibility, although their dependents would be able to maintain their coverage.
Under COBRA, employers have the responsibility of notifying their covered employees, spouses, and dependents when they become eligible to continue coverage. COBRA coverage for beneficiaries must also be identical to that provided to similarly situated active employees and their dependents. The exception is that firms may charge COBRA beneficiaries up to 102 percent of the average employer cost of providing health insurance under its group plan rather than the rate that is nominally charged to active employees which is often highly subsidized. (The average COBRA premium is about $400 per month.)
Thus, for example, assume that a company with a group plan can insure each employee for an expected cost of $400 per month, less than the nongroup rate of, say, $600 per month, that an employee would pay if purchasing insurance as an individual in the market. A typical employer might cover that cost by charging each employee a monthly contribution of $100 and paying the remaining $300 directly out of revenues. Under COBRA, if the employee leaves that job, he or she can continue this coverage by paying 102 percent of the $400 per employee monthly total, that is, $408. In that case, direct individual expenditures on health insurance would increase by $308, from $100 per month to $408 per month. But the former worker would still be better of than if he or she had to pay the nongroup rate of $600.
WHO USES COBRA?
In any given year about 10 percent of the workforce experiences a qualifying event that could potentially trigger COBRA eligibility. Some 20 percent of that group, or about 2 percent of the workforce, will actually elect COBRA coverage. While that might seem a small number, the Employee Benefit Research Institute places the number of former employees covered by COBRA at some point during the year at about 5 million with an additional 2.5 million to 3 million covered dependents. To put those numbers in perspective, about 4.5 million individuals under the age of sixty-five are covered by Medicare, and another seven million individuals under the age of sixty-five are covered by the military's CHAMPUS/VA programs. So COBRA is of roughly the same magnitude as those two sources of public health insurance for the nonelderly population. The average duration of COBRA coverage for individuals who are eligible for eighteen months of coverage is 10.3 months, while for individuals eligible for thirty-six months of coverage it is 22.9 months.
One reason why many individuals do not elect to continue their insurance under COBRA is that they have alternative sources of health insurance and either do not qualify or do not really need COBRA. For example, an employee who leaves one job with health insurance for another job with health insurance probably does not need COBRA and, in fact, would be ineligible for COBRA coverage unless the new employer's insurance excludes preexisting conditions. Similarly, an individual who retires or who is laid off likely would not need COBRA if health insurance coverage through a spouse is available, and would also be ineligible if he or she did in fact subsequently become covered under a spouse's health insurance policy. The law is open to some interpretation about whether individuals who were jointly covered through their own and a spouse's health insurance at the time of a qualifying event are eligible for COBRA and there is a case currently pending before the Supreme Court on just that issue. In Geissel v. Moore, the plaintiff was initially covered by health insurance through both his own and his wife's employer. After being laid-off, his former employer denied the plaintiff COBRA coverage because he was already insured through his wife's plan. The plaintiff, however, desired to maintain both sources of coverage because he had cancer. The plaintiff maintains that the provision of the law which ends the COBRA eligibility for those with another source of group health insurance applies only to health insurance obtained subsequent to a COBRA qualifying event, not to health insurance held at the time of a qualifying event.
Among those who do not have reasonable alternative sources of health insurance coverage, the take up rate for COBRA coverage is about 65 percent. Thus, COBRA is an important source of health insurance coverage for that segment of the population that finds itself temporarily unable to obtain employer-provided health insurance. My research with Jonathan Gruber of MIT shows that the availability of continuation coverage has increased the probability of insurance coverage for potential early retirees, those aged fifty-five to sixty-four, by at least 6 percent. It has also decreased the likelihood that the long-term unemployed, those whose job loss lasts at least twelve months, will be uninsured by 20 percent.
COBRA AND THE LABOR MARKET
In addition to its effects on insurance coverage, COBRA has significant effects on individual labor force behavior. For the majority of full-time workers, employment and health insurance go hand-in-hand. Any change in employment moving from one job to another, becoming self-employed, retiring, or being laid off will involve relinquishing one's current employer-provided health insurance. Further, not all health insurance plans are created equal. Some have higher deductibles than others, some require copayments, some give individuals more say in their choice of physicians, some exclude preexisting conditions, some cover more services than others. That variation in plan characteristics coupled with the substantial variability in individual needs imply that some individuals will place a higher value on maintaining their current employer-provided health insurance than will others. The former individuals will be reticent to make job changes that will also result in a change in or loss of health insurance coverage a so-called "job lock." COBRA, however, allows individuals to maintain the health insurance that they receive from their current employer if their employment situation changes. Consequently, individuals who would be wary of leaving their current job because the costs of relinquishing the health insurance that goes with it are too high may decide that the health insurance portability afforded by COBRA now allows them to do so. It is true that an employee on a new job that does not cover preexisting conditions can maintain coverage under the older employer's policy through COBRA for only eighteen months. But in many cases that is significant to reduce job lock. For example, a pregnancy lasts no more than nine months. Many cancer therapies, if successful, do not last more than several months. Thus, for many individuals, eighteen months provides sufficient time to either find or fully qualify for alternative health insurance while still keeping coverage through COBRA.
EFFECTS ON YOUNG AND OLD
Perhaps COBRA's most significant effect on employment is its promotion of early retirement. Individuals are currently not eligible for Medicare until they reach age 65. (President Clinton has recently proposed giving individuals aged sixty-two to sixty-five the option to buy into Medicare). While some employers allow retirees to maintain their health insurance after they retire, many do not.
COBRA enables those without access to employer-provided retiree health insurance either to fully or partially bridge the "insurance gap" until they are eligible for Medicare. My research with Jonathan Gruber suggests that COBRA has increased the probability that individuals aged fifty-five to sixty-four will retire in any given year by almost one-third over what it would be without COBRA.
That increase in the retirement rate is derived by comparing the probability of retirement before COBRA was passed with that after COBRA passed. We can refine the comparison to account for the overall secular decline in retirement probabilities over the past several decades by comparing the effect of COBRA on retirement in states thathad COBRA-like mandates already in place when COBRA was passed with the effect in states without any similar mandates (between 1974 and 1986 when COBRA took effect, twenty-two states had enacted similar health insurance continuation laws.) The table above illustrates the differences in those retirement probabilities.
In the years preceding the adoption of COBRA, 1983 to 1985,6.8 percent of men aged fifty-five to sixty-four retired each year. In the years following the implementation of COBRA, the fraction of men retiring in states that had similar continuation laws already in place increased only slightly, to 7.2 percent, a 0.4 percentage point or 6 percent increase. One would have expected only a small change for that group because COBRA did not dramatically increase the availability of health insurance for individuals in those states.
In contrast, in states that had not passed similar continuation laws, the retirement rate increased dramatically, from 6.8 percent to 8.9 percent, a 2.1 percentage point or 31 percent increase. For those individuals, COBRA represented a new opportunity to continue their health insurance that did not previously exist. A more formal statistical analysis that accounts for differences in other factors such as marital status, age, race, education, industry, occupation, and pension coverage yields substantively similar results-COBRA has increased the probability that individuals aged fifty-five to sixty-four will retire by almost one-third.
A similar type of analysis reveals that COBRA has also affected the employment behavior of younger workers, increasing their job turnover rate in the workforce as a whole by between 10 percent and 15 percent. Moreover, the effects are similar for movements from one job to another, from employment to self-employment, and from employment to unemployment. Overall effects amount to a 40 percent to 50 percent reduction in the degree of job lock resulting from employer provision of health insurance. That is, for every one hundred individuals that would like to move to a different job but would not do so for fear of losing health insurance, forty to fifty now will make the move, thanks to COBRA.
How do we assess the value of the increased retirement and job mobility induced by COBRA? The individuals affected are clearly better off. The early retirees could have remained employed but chose to retire instead. Similarly, those who changed jobs could have remained with their former employer, but chose to leave instead. For the latter group, the value of increased labor mobility may exceed even the psychic benefits derived by the worker. Since individuals have different skills, moving to a job that is a better "fit" could mean that labor is being employed in a more efficient manner, contributing more to GDP. That benefits both workers and firms.
The magnitude of the benefits of matching a worker to a job in which his or her skills are better utilized is difficult to quantify. One indicator, however, is the effect of COBRA on earnings. By reducing the necessity of finding a job in order to obtain health insurance coverage, COBRA allows the unemployed to spend more time in their job search. Jonathan Gruber and I estimate that the reemployment earnings of those who lose jobs are 6 percent higher than they otherwise would be as a result of COBRA. While that figure is suggestive, further research is needed on other workers, such as those who have left their jobs voluntarily, before stronger conclusions can be drawn about the effects of COBRA on overall job quality.
EARLY RETIREMENT COSTS
Although there are substantial benefits derived by both individuals and firms from better job matches, some of those gains are partially offset by the societal costs that result from early retirement. They include reductions in both economic output and tax revenue, both of which impact society as a whole, not just the individuals who have retired.
A rough calculation indicates the possible magnitude of those reductions. In 1996, fifty-five to sixty-four years olds earned about $380 billion in employment income. Jonathan Gruber and I estimate that the COBRA-induced increase in the retirement rate will lead to a steady state decline in the labor force participation of older individuals of 3.3 percentage points. The resulting loss in economic output would thus be $12.5 billion annually. If we assume a combined average tax rate (federal plus state plus social security plus medicare) of 35 percent, COBRA-induced retirement would lead to a reduction in tax revenue of $4.4 billion.
However, the early retirees do receive income that is subject to taxation from pensions and the like. That income would not, however, be subject to social security taxation. Thus it would face a much lower average tax rate, probably around 20 percent or less. If we assume that taxable income falls by half, from $12.5 to $6.3 billion for COBRA-induced retirees, the 20 percent tax rate on the income would yield $1.3 billion in revenue. Thus the net reduction in tax revenue would be $3.1 billion.
President Clinton's plan to allow early retirees to buy into Medicare would likely have retirement effects that are similar to those generated by COBRA, with the attendant reductions in both output and tax revenue as well.
ADVERSE SELECTION COSTS
In addition to the societal costs of early retirement, another serious problem that COBRA has created for employers is the cost of adverse selection. The law does not allow a firm providing COBRA to differentiate the costs of premiums along dimensions other than those followed in the health plan for active employees. For example, if the firm requires different employee contributions for individual and family health insurance coverage in its plan for active employees, then it can only charge one COBRA premium for beneficiaries who elect individual coverage, and another premium for beneficiaries who elect family coverage.
But there is much more variation across individuals in average medical expenditures than that generated, say, by family situation. The average annual medical expenditures of fifty-five to sixty-four year-olds are three times those of twenty-five to forty-four year-olds. An individual with AIDS or cancer will have much higher anticipated medical expenditures than an individual of the same age who is in perfect health. But, unlike commercial insurers in the individual health insurance market, firms are constrained to charge the same COBRA premium to all of the individuals. The result is that the individuals who are most likely to elect COBRA coverage are those who have higher than average expected medical expenditures.
The adverse selection problem is exacerbated by one interesting feature of the law: terminated beneficiaries have sixty days to decide whether they want COBRA coverage, and if they elect COBRA, coverage is then retroactive. (Employees must, however, pay for retroactive coverage.) That means that any individual who initially decides against COBRA but who in the following sixty days experiences health problems that generate medical bills in excess of COBRA premiums would be foolish not to retroactively elect COBRA. That provision of the law is a certain recipe for adverse selection.
How severe is the adverse selection problem? Since 1989, the benefits publishing firm Charles D. Spencer & Associates has conducted an annual survey of employers regarding their COBRA experiences. The results of their survey reveal that the average COBRA beneficiary costs firms about 150 percent of what the average active employee does. Thus, if the average premium for employer-provided group health insurance is $400 per month, then the total costs for COBRA beneficiaries average $600 per month. But, by law, the premium paid by COBRA beneficiaries can be no more than 102 percent of the average per employee cost of providing health insurance to the firm's active employees. So firms on average subsidize about one-third of the cost of providing health insurance to their COBRA beneficiaries.
Those additional costs are borne by employers in one of two ways. First, many employers actually self-insure; meaning that they bear full responsibility for all medical expenditures in excess of those anticipated. (Similarly, they are the full beneficiaries if actual expenditures fall short of those anticipated.) Thus, if the expected costs for an active employee are $400 per month but the actual costs incurred are $600, the firm is completely responsible for the $200 excess. Under COBRA, the firm can charge COBRA recipients a premium equal to 102 percent of active employer costs, or $408 per month in this example. But if the adverse selection concerning who elects COBRA results in expenditures that are 150 percent of those incurred by active employees, or $600 per month, the firm will have to cover the entire $192 difference out of its own revenues.
Second, some enterprises, especially smaller ones, purchase coverage for their employees from an insurance company. The companies set rates based on the past and future expected medical costs of the enterprise's workers. For those firms, the disproportionately higher costs incurred by COBRA-covered former employees will be reflected in higher premium charges in the future.
COST VARIABILITY
In addition to adverse selection, employers face great variation in their COBRA costs. While the average ratio of COBRA to active employee expenses is 150 percent, one in five firms have COBRA costs in excess of 200 percent of active employee costs, while one in ten face costs that are more than 300 percent of active employee costs. Moreover, except in exceptionally large firms-those with more than ten thousand employeesthe ratio of COBRA to active employee costs appears to be more or less random. Stephen A. Huth of Charles D. Spencer & Associates has summarized the problem nicely: "Only the largest employers have COBRA experience approaching the average. For most other employers, providing COBRA coverage is much like rolling dicesometimes you win, sometimes you don't.... The low incidence of COBRA elections in any one company makes COBRA operate more like individual health insurance than like group insurance."
For the employers who "win," COBRA premiums cover the medical costs incurred by their COBRA beneficiaries. For the employers who "lose," COBRA premiums cover only a fraction of the costs incurred by those individuals and their dependents. The "excess" costs are imposed on firms by individuals who have no current attachment to the firm whatsoever, and, what irks many employers most, who may have left voluntarily to work someplace else. For small employers, COBRA simply exacerbates the struggle with the randomness inherent in the experience-rated premiums of the small group health insurance market.
ADMINISTRATIVE COSTS
Finally, there are administrative costs associated with COBRA. All employees and their dependent beneficiaries must be notified of their rights to continue health insurance coverage following an event such as a job change or divorce which would otherwise result in a loss of coverage. On average, about 10 percent of employees face such situations annually, but that fraction can vary widely across firms. For example, firms with high employee turnover may easily find that one-third or even one-half of their employees will need such notification. Moreover, some of the events which precipitate COBRA eligibility can make the responsibility of notification difficult. For example, job change and divorce are often accompanied by address changes as well.
Once an individual has elected COBRA coverage, the firm then has the burden of keeping track of those individuals until they either relinquish their coverage or it expires. For some beneficiaries, this period can last thirty-six months. Not surprisingly, a mini-industry has developed which will notify and track COBRA-eligible individuals and recipients on behalf of employers for a fee, of course.
Employers bear the burden of collecting and processing COBRA payments. The law allows employers to charge 102 percent of the employer's average cost of providing health insurance, with the extra 2 percent supposedly to cover administrative costs. Yet employers report average administrative costs that amount to about 6 percent of COBRA premiums, from calculations based on the Charles D. Spencer & Associates data. There is, however, wide variability in the administrative costs reported by employers: some employers report costs of as much as $1000 per year for each employee who elects COBRA. While some of the variation in administrative costs probably reflects differences in the quality of administrative services or the efficiency with which they are rendered, part of it undoubtedly reflects the randomness of dealing with whomever happens to be benefiting from continuation coverage in any given year.
Can we quantify the COBRA-related costs borne by firms? Charles D. Spencer & Associates estimates that in 1997, the annual per beneficiary medical expenditures for COBRA recipients exceeded those for active employees by $2,260. If that figure is multiplied by the number of former employees covered by COBRA, estimated at 5 million, the total direct excess costs to employers is about $11.5 billion per year.
That figure does not, however, reflect the costs of the uncertainty which COBRA generates for some employers, particularly for small firms that offer health insurance. It also does not reflect the legal costs incurred by employers (and taxpayers) generated by the lawsuits that have inevitably followed a mandate as complex and confusing as COBRA (as noted earlier one such lawsuit is actually being heard by the Supreme Court). Thus, the true cost to employers of COBRA is likely much higher than $11.5 billion per year. That places excess employer expenditures on COBRA in the same league as employer expenditures on unemployment insurance. (In 1996 the states paid out $20.6 billion in unemployment benefits.) Also not reflected in that figure is the potential cost borne by employees in firms that decide not to offer health insurance coverage because COBRA, in combination with the many other health insurance mandates that vary by state, makes health insurance too expensive to offer.
NEW HIPAA COSTS
The HIPAA restrictions on preexisting conditions exclusions are likely to compound all of the problems just described. Before the HIPAA, a firm might, for example, exclude coverage for a preexisting medical condition for twelve months for new employees. A primary reason that employers and other insurers impose preexisting conditions exclusions is to reduce the problems of adverse selection. Without such exclusions, individuals have an incentive to forego health insurance while they are healthy, waiting until an adverse medical event occurs before enrolling in a health plan. Most employers do not want to hire employees with large anticipated medical expenditures whose primary interest in a job is the great health insurance benefits that come with it.
Under the HIPAA, insurers must reduce any preexisting conditions exclusion by one month for each month that an individual can demonstrate a history of prior, uninterrupted health insurance coverage. Thus, for example, if an employer excludes coverage of some illness for one year, but a new employee can show that he or she had insurance for six months prior to taking the new job, the employer will have to reduce the preexisting conditions exclusion by six months. If the new worker had coverage for one year or more before coming to the new job, the employer could not impose a preexisting conditions exclusion at all.
As with COBRA, the costs of adverse selection that result from HIPAA are likely to be unpredictable and highly variable. The problems will be particularly acute for small firms, many of which are hard pressed to offer health insurance in the first place. (Indeed, most small firms do not offer health insurance at all). Small firms cannot take advantage of size in pooling risks, so the only weapon they have against adverse selection in an experience-rated health insurance market is preexisting conditions exclusions, and the power of that weapon has been severely blunted by HIPAA.
HIPAA also imposes additional administrative burdens on employers. Employers providing health insurance now have the responsibility of validating for new employers the coverage that former employees and other terminated beneficiaries have had, as well as verifying the coverage that new employees and their dependents have had with previous employers or other insurers. As with COBRA, the average administrative costs associated with HIPAA likely will fall harder on employers with high rates of job turnover.
IS THERE MARKET FAILURE?
If the value of better job matching brought about by health insurance portability is so significant, why does not the market for portable health insurance exist without costly government mandates? Is there a market failure that calls for government intervention?
It is interesting to note that many firms voluntarily offered post-retirement health insurance to their early retirees for decades before COBRA. Employer-provided retiree health insurance is a type of "continuation coverage" that is usually provided on terms much more generous than COBRA. So, there cannot be a complete market failure since that type of continuation coverage existed and, indeed, was widely available before COBRA.
The rationale for employers to provide retiree health insurance follows from the fact that many firms pay employees less than what they contribute to the company early in their employment tenure and then make up for it by paying them more than what they contribute later on. Firms compensate their workers in this way for many reasons: to reduce unwanted job turnover by bonding the worker to the firm; to discourage older workers from shirking; and to encourage selfselection of workers who are good matches into the firm. However, firms will eventually want their workers to retire because they cannot forever employ older workers who are contributing less to the company than they are being paid. But a pension alone might not be a sufficient incentive if early retirement leaves the worker without health insurance. So many firms offer post-retirement health insurance as well. As long as the cost of providing health insurance to the retired worker is less than the difference between what the worker is paid and what the worker contributes, it is a perfectly sensible thing for firms to do.
Why then do not all firms provide continuation coverage to all of their departing employees? Because much of the time it does not make economic sense, at least not for the employer. The primary reason is adverse selection. The employees who will most want continuation coverage are those with anticipated medical costs that exceed the cost of obtaining the coverage. Although the benefits of improved job matching might well exceed the cost to the firm of the resultant adverse selection, none of the benefits would accrue to the firm offering the continuation coverage. Because the firm cannot capture any of the benefits from job turnover that might accrue to departing workers and their future employers, it has no incentive to provide this type of portable health insurance.
Could not the firm capture the value of private benefits by incorporating them into the premium charged to departing employees? In principle, yes, but raising the premium for continuation coverage would simply serve to exacerbate the adverse selection problem. In short, there appears to be little reason to think that firms would voluntarily provide such coverage, at least not for younger workers.
ALTERNATIVE APPROACHES TO HEALTH INSURANCE PORTABILITY
If health insurance portability is a desirable thing, are there better ways to achieve it that do not require firms to bear the large and variable costs of adverse selection?
One alternative to the current system would make COBRA provision the responsibility of the government rather than the responsibility of firms. The government could provide time-limited health insurance under the same conditions as COBRA at a rate that reflects the average cost of health insurance across all firms. Taxpayers would then bear the costs of adverse selection and of COBRA administration. This option would clearly further increase the role of government in the provision of health insurance.
A second alternative would have the government subsidize employers for the costs of adverse selection that result from their provision of COBRA. For example, a crude approach would be to simply give all firms a tax credit equal to 50 percent of the active employee cost of providing health insurance, a figure that reflects the average costs of adverse selection. In a more targeted implementation, employers could be allowed to claim a tax credit equal to the amount by which their COBRA-related expenditures exceed the COBRA premiums that are paid and, conversely, employers with COBRA claims below premiums would pay increased taxes equal to this amount. This approach would also push the costs of adverse selection onto the government and taxpayers, although firms would continue to bear the costs of administering COBRA.
A third alternative would expand the HIPAA guarantees of access to private market health insurance for terminated beneficiaries at rates no greater than the average employer cost at the departing firm. It would shift the cost of continuation mandates from employers to the insurance companies that operate in the private market and their shareholders. HIPAA currently guarantees group-to-individual health insurance portability only after individuals exhaust COBRA coverage. HIPAA also does not regulate the rates that private insurers can charge individuals.
A fourth alternative would relax the requirement that health insurance coverage under COBRA be the same as that offered to active employees. For example, employers could be allowed to restrict COBRA health insurance coverage to vital medical services, excluding such things as vision care, dental care, routine physicals, and well-baby care. This would encourage COBRA beneficiaries to postpone their nonvital health care expenditures but not harm individuals who are receiving ongoing care or who need emergency care. Employers could also be allowed to require a higher deductible or copayment from COBRA beneficiaries. Although this approach would increase the administrative costs for the firm, if those costs were too great employers would simply not offer a differentiated COBRA plan.
Finally, there could be alternatives to government provision of health insurance that do not rest on health insurance being attached to jobs. For example, Congress has authorized a limited number of medical savings accounts that are similar to tax-exempt individual retirement accounts. It might be useful to explore the use of such an approach to cover health insurance transition costs between jobs.
The issue is ultimately one of who should bear the costs associated with the lack of health insurance portability inherent in a system of primarily employment-based health insurance. Without health insurance portability, there are private costs borne by individuals (the risks of being temporarily uninsured and the psychic costs of being stuck in a less desirable job) and social costs as well (lower economic output because some workers are constrained from moving to jobs where their productivity would be higher). But generating health insurance portability within the confines of a system of employment-based health insurance leads to adverse selection. The question of who should bear the costs of this adverse selection: firms and their shareholders, the government and the taxpayers, or private insurance companies, is one on which there is likely to be little agreement.
|
