Simple fairness: ending discrimination in health insurance coverage of addiction treatment.

Consider these facts: An often-debilitating brain disease afflicts millions of Americans. This disease is one of the country's greatest killers. Its victims frequently suffer from depression and many physical ailments, and often become unable to work effectively. The disease costs the U.S. economy hundreds of billions of dollars annually--more than cancer, more than heart disease. (1) Fortunately, although no cure exists, medical treatment can enable recipients to live normal, healthy, and productive lives. Treatment is cheap compared to many other common medical procedures and is highly cost-effective. (2) Now consider this: For the vast majority of victims of this disease, effective treatment is inaccessible. Most health insurance plans either do not cover it or put a variety of limits on coverage that do not apply to other diseases. Unless they can pay out of pocket, victims cannot get the treatment they need. To make matters worse, they are often told that their condition is not a real disease, or that it is their fault, or that suffering from it makes them a criminal.

The disease is drug and alcohol addiction, and the facts are real. Ubiquitous benefit caps on insurance coverage of substance abuse treatment put effective recovery out of reach for most addicts. In this Note, I assess the nature of this problem and some possible ways to address it. The general principle that I advocate is substance abuse treatment parity, which means that insurance plans should provide coverage for addiction treatment that is equivalent to that provided for analogous conditions. In some cases, failure to provide such parity should be considered illegal disability discrimination on the part of employers and insurers. Moreover, new laws should be adopted to require insurance parity explicitly.

In Part I, I review the current status of insurance coverage of addiction treatment and assess the scope of the shortfall and possible reasons behind it. In Part II, I set forth the case for insurance parity, including the nature and costs of the disease of addiction and the efficacy and cost-effectiveness of treatment, and consider some counterarguments. In Part III, I analyze the requirements of the Americans with Disabilities Act (ADA) as they pertain to insurance parity, drawing on the precedents set by recent challenges to other types of insurance discrimination. I conclude that the ADA should be interpreted to require parity in some cases, but that the potential effectiveness of this litigation strategy is limited--new reforms are necessary. In Part IV, I consider the strengths and weaknesses of current legislative proposals to accomplish insurance parity, and focus especially on the Substance Abuse Treatment Parity Act. Finally, in Part V, I offer my conclusions and recommendations for legal change and advocacy.

I. CURRENT SHORTFALLS IN INSURANCE COVERAGE OF ADDICTION TREATMENT

Most Americans benefit from health insurance plans provided by their employer or the employer of a family member. (3) Others are covered by Medicare or Medicaid, or purchase individual or family plans from a private insurer or health maintenance organization (HMO). (4) Among employers who provide insurance, some--generally very large companies--are self-insurers, meaning that they serve as their own insurance company, while others purchase group plans from third-party insurers. Approximately thirty-nine million Americans are uninsured.(5) Even among those who have insurance policies, however, coverage of alcohol and drug addiction treatment is often limited or absent. (6) In some cases, insurers entirely exclude coverage of addiction. More frequently, addiction treatment is subject to monetary caps and other limitations on coverage that do not apply to treatment of other diseases. (7) Frequently, coverage is limited to brief, one-shot treatment programs with no long-term maintenance care--a strategy with little chance of success. (8) The coverage gap is far more severe for employees of small businesses; although 90% of Fortune 500 companies have Employee Assistance Programs (CAPs), which are programs designed specially to deal with substance abuse treatment (although they may not provide insurance parity), few small businesses have them. (9) Substance abuse treatment may be most lacking for adolescents, for whom specialized care is rarely covered and whose parents may already have exhausted their family plan's lifetime addiction treatment allotment. (10) Overall, fewer than 10% of all American workers have a health insurance plan that treats addiction equivalently to analogous diseases. (11) Some estimates suggest that only 2% of substance abusers have health insurance plans that provide adequate coverage for treatment. (12)

These inequalities in treatment coverage have long been a part of the insurance landscape. Nonetheless, the gaps are becoming even wider as a consequence of the rise of managed care. Managed care has, over a short period of time, become the dominant force in American health care, representing a 29% share of the health insurance market for American workers in 1988 and over 80% today. (13) Unlike a traditional third-party insurance plan, an HMO (the most common form of managed-care plan) is both insurer and care provider--the doctors work for the same company that pays for treatment. This creates cost-cutting incentives that help control insurance premiums, but may compromise the quality of care. (14)

These cost pressures appear to have had an especially serious effect on addiction treatment. A Hay Group report showed that between 1988 and 1998, employer spending on substance abuse treatment benefits declined by 74.5%, while employers' overall health benefit spending declined by just 11.5%. (15) A 1996 Village Voice article reported that, as a result of these pressures, over half of the country's private treatment centers had closed over the course of a decade. (16) In addition, quality of care may be declining. Whereas a twenty-eight-day inpatient program used to be common, today inpatient stays average only 7.7 days, which may be insufficient to provide the necessary care. (17) Some studies have found that relapse rates have increased as a result. (18)

Despite the bleakness of this picture and the considerable obstacles to change, (19) recent years have seen the quiet growth of a pro-treatment movement. This movement is grounded in a coalition of people in recovery, treatment providers, medical experts, some sympathetic political leaders, and even a range of businesses that have found that providing treatment coverage makes economic sense. In terms of health insurance coverage, the focal point of this movement has been "treatment parity." The concept of "parity" implies equal treatment or, more simply, fairness. While allowing health insurance plans flexibility in terms of setting the actual amount of coverage for substance abuse treatment, parity advocates demand only that this amount be fair by comparison to coverage for analogous conditions. Parity advocates have won some victories in recent years. Eight states now have laws mandating some degree of substance abuse treatment parity, (20) although these laws have been criticized for their incompleteness. (21) In 1999, President Clinton announced that the Federal Employee Health Benefit Plan would adopt full parity for substance abuse and mental health treatment by 2001. (22)

II. THE CASE FOR INSURANCE PARITY

The issues surrounding insurance coverage of addiction treatment are complex, in part because of the very nature of the health insurance business. There is no such thing as truly "comprehensive" health insurance--i.e., insurance that covers every medical treatment. Insurance companies routinely determine what procedures they will and will not cover on the basis of considerations such as the seriousness of the underlying condition and the cost and effectiveness of the treatment. Because of the constantly changing landscape of medical science, decisions about whether to cover certain procedures are frequently controversial. It is thus understandable that politicians are often reluctant to jump into the fray by issuing coverage mandates that would take discretion over coverage decisions away from insurance professionals.

Nonetheless, considerable medical and economic evidence supports the argument that mandatory insurance parity for addiction treatment is an idea whose time has come. Studies have shown that treatment is both medically effective and cost-effective. The recovery movement's focus on parity sets a modest goal: requiring not that all insurers provide comprehensive coverage of addiction, but simply that they treat addiction equivalently to analogous diseases. This Part sets forth the medical, social, and economic arguments for insurance parity for drug and alcohol addiction treatment. (23)

A. Addiction Is a Disease

The first major barrier faced by proponents of drug and alcohol treatment coverage consists of convincing the public and policymakers that addiction is, in fact, a medical illness requiring medical treatment. Alive as it may be in the public mind, the idea that people are addicts by choice has long since lost currency among medical experts. Medical descriptions of addiction as a disease date to the eighteenth century. (24) The American Medical Association first officially recognized addiction as a disease in 1956. (25) Today, a remarkably uniform consensus in the scientific community supports this characterization. As Time recently reported: "Americans tend to think of drug addiction as a failure of character. But this stereotype is beginning to give way to the recognition that drug dependence has a clear biological basis. `Addiction ... is a disorder of the brain no different from other forms of mental illness.'" (26) This conclusion has been confirmed by a litany of scientific studies. (27)

Recently, scientists have begun to acquire a greater understanding of the neurological and chemical mechanisms by which addiction develops. For example, a 1997 study by Yale School of Medicine researchers Eric Nestler and George Aghajanian documented a range of changes in brain function resulting from repeated exposure to addictive drugs. Changes in the physical structure of certain neurotransmitter receptors, as well as changes in the rates of some forms of neurotransmission, create the chemical conditions that produce behavioral changes: dependence, tolerance, sensitization, and craving. (28) Nestler and Aghajanian explain:

   Addiction is a complex phenomenon with important psychological and social
   causes and consequences. However, at its core, it involves a biological
   process: the effects of repeated exposure to a biological agent (drug) on a
   biological substrate (brain) over time. Ultimately, adaptations that drug
   exposure elicits in individual neurons alter the function of those neurons,
   which in turn alters the functioning of the neural circuits in which those
   neurons operate. (29)

Some studies suggest that these changes may involve effects on a single neural pathway common to all or almost all forms of addiction: the mesolimbic reward system. (30)

Most drug and alcohol users do not become addicted. The likelihood of addiction is affected by a variety of factors that scientists are only beginning to understand. Strong evidence from a variety of studies demonstrates a genetic component of susceptibility to addiction. (31) In addition, a person's life experiences may influence his or her likelihood of becoming an addict. Physical abuse as a child increases the likelihood of later substance abuse. (32) Parental substance abuse and attendant abuse and neglect can also engender a cycle of future use in children. (33)

A National Institutes of Health (NIH) study found that 7.4% of adult Americans abuse alcohol and 1.5% abuse illicit drugs. A smaller percentage is actually chemically dependent: 4.4% on alcohol and 0.5% on illicit drugs. (34) Furthermore, addiction is a disease that does not discriminate in terms of race, gender, or social class; current and recovering addicts are found in all walks of life. (35) Seventy percent of drug abusers and 75% of alcoholics are employed. (36)

B. Health, Social, and Economic Costs of Untreated Addiction

The human and economic costs of untreated drug and alcohol addiction are staggering. Many of these costs are borne by employers and by society as a whole, but it is worth focusing first on the private suffering experienced by the addict and his or her family. Untreated addicts experience a wide range of direct health consequences from their illness; a 1993 Brandeis University study found that 25-40% of all people in general hospital beds are there because of alcoholism-related conditions alone. (37) Drug and alcohol addicts frequently experience depression and are at thirty times greater risk of suicide than nonaddicts. (38) The American Society of Addiction Medicine has stated that "the progression of untreated addictive disease has significant impact on the prevalence of a variety of physical and mental disorders and on the expenditures required" to treat them. (39) In particular, "the interface between addictive disorder and HIV infection is one of the major public health challenges facing the nation." (40) Unsafe use of intravenous drags is directly or indirectly responsible for more than one-third of all AIDS deaths to date in the United States. (41) Intravenous drag use also increases transmission of tuberculosis and hepatitis B and C. (42)

Substance abuse causes substantially increased use of the health care system, and particularly of expensive emergency room and trauma care services. Sixty-five percent of all emergency room visits are related to drag and alcohol problems. (43) Alcohol abusers spend four times the average number of days in hospital care, largely because of increased rates of injury. (44) Many types of accidents are highly correlated with alcohol abuse: 13-63% of falls, 33-61% of fatal fires, and one-third of drownings. (45) In addition, alcohol abuse causes close to half of all fatalities in motor vehicle accidents. (46) Alcoholism causes liver disease, which is often fatal, and many forms of neurological pathology. (47) In total, more than 100,000 people die in the United States each year because of alcohol-related problems, while nearly 16,000 deaths each year result directly from abuse of illegal drags, not including indirect effects such as accidents and drug-related disease transmission. (48) On average, those who die of alcohol-related conditions lose twenty-six years of life. (49)

Untreated addiction has catastrophic impacts on those close to the addict, especially family members. Health experts agree that drag and alcohol abuse are "inextricably linked" to problems including child abuse and other family violence. (50) Addiction, especially to alcohol, is highly correlated with domestic violence; in one study, 78% of female victims who survived domestic violence attacks reported that their attacker had been using drags or, more commonly, drinking. (51) Alcoholism is also strongly associated with rape and sexual assault. (52)

Addiction also leads to abuse and neglect of children; for example, one study found that nearly half of the fathers who sexually abused their children were alcoholics, (53) while another found that child abuse or neglect occurred 3.6 times more often in families with alcoholic members than in those without. (54) The number of children affected is by any measure enormous; though estimates range widely, up to ten million children may be living with chemically dependent parents at any given time. (55) Some forms of prenatal substance exposure have documented negative effects on fetal and child development; this may affect up to 20% of newborn children in America. (56) Abuse of alcohol by pregnant women causes one in every ten cases of mental retardation. (57) Substance abuse may lead to the breakup of families; for example, alcoholics have high rates of divorce, (58) and parental use of drags or alcohol contributes to at least three-fourths of cases where children are removed from their families and placed in foster care. (59)

Furthermore, untreated addiction inflicts many costs on U.S. employers every year. A 1996 study by the Bureau of National Affairs estimated that addiction costs U.S. businesses $200 billion every year. (60) This estimate encompasses half a billion lost workdays, premature deaths of workers including on-the-job accidents, and health treatment costs including fetal alcohol syndrome and AIDS. Other studies have further documented each of these factors. For example, researchers estimate that close to half of all workplace accidents are caused by alcohol abuse alone. (61) Indeed, it is alcohol, not illegal drugs, that costs U.S. businesses by far the most each year. (62) Substance abusers file five times as many workers' compensation claims as do nonabusers, have twice as many unexcused absences, and are tardy three times as often. (63)

In addition, the U.S. public absorbs a wide range of costs of addiction in the forms of lost productivity, social deterioration, and direct financial burdens on government programs. (64) A study by the HMO Kaiser Permanente found that treatment led to a 50% reduction in patients' days of hospital utilization. (65) Treatment of fetal alcohol syndrome and other prenatal drag exposure exacts billions of dollars in additional social costs. (66) Each premature death due to untreated addiction costs the economy almost $350,000. (67) A 2001 Brandeis University study estimated that, in total, untreated alcohol abuse costs Americans $166.5 billion every year, and drag abuse costs $109.9 billion. (68) An NIH study analyzing economic data from 1992 found that alcohol and drag abuse cost society $246 billion during that year. (69)

C. Treatment Works

One of the most intractable obstacles to insurance parity reform is the widespread but inaccurate perception that addiction treatment is ineffective. (70) In fact, studies show that treatment programs on average have an effectiveness rate of 30-60%, if effectiveness is measured in terms of the number of people who remain totally abstinent for one year or longer. (71) Individual treatment programs may achieve substantially higher abstinence rates. The Partnership for Recovery, a coalition of top treatment centers, cites success rates of 51-75%. (72) In 1989, AMR Corporation, the parent company of American Airlines, established an Employee Assistance Program (EAP) that encompassed insurance parity. In testimony to Congress, EAP manager John Saylor testified that AMR's program had a 75-80% success rate in keeping patients sober for a year or more. (73)

Moreover, a focus on total and permanent abstinence as the sole measure of effectiveness is misplaced. Addiction is a "chronic, relapsing disorder. Total abstinence for the rest of one's life is a relatively rare outcome from a single treatment episode." (74) Viewed in this light, success is marked by "a significant decrease in drug use and long periods of abstinence." (75) When this more realistic measure of success is used, success rates predictably increase; for example, the Hazelden Foundation boasts a 77% rate of keeping people "clean and sober" at the end of one year--54% having remained completely abstinent, and the remainder having had just one episode of use during that time. (76)

Indeed, although addiction treatment is not by any means universally successful, neither is treatment for many other conditions that are routinely covered by health insurance. Addiction treatment has similar success rates to those for treatment of hypertension, diabetes, and asthma. (77) Indeed, one study found that addiction treatment "ranked in the top 10 percent of medical treatments reviewed for savings in money and lives." (78)

D. Treatment Is Cost-Effective

Substance abuse treatment has a range of economic benefits for patients, health insurers, employers, and society as a whole. These benefits mean that investments in treatment more than pay for themselves. Treatment coverage may even decrease total health care costs because the cost of treatment is outweighed by decreases in other uses of the health care system. A study of patients receiving publicly funded treatment in Washington State found that five years after treatment, medical expenses were cut in half relative to costs of those who were untreated. (79)

More likely, however, slight increases in premiums would result from parity requirements, according to several recent studies. A study by RAND Corporation economist Roland Sturm estimated that, as compared to no coverage for treatment at all, unlimited addiction treatment coverage would add $5.11 to premiums per plan subscriber per year. (80) This amounts to a premium increase of approximately 0.3%. (81) The same study found that a $10,000 treatment cap saves members only six cents per year (versus unlimited coverage), while a $1000 cap saves $3.39 per year. (82) Even the starkest of these differences, however, still amounts to less than fifty cents per month. Similarly, an actuarial study by Milliman and Robertson, Inc., found that full and complete insurance parity provisions, if adopted nationwide, would raise average nationwide premiums by less than 0.5%, or less than one dollar per month; more limited parity provisions would raise rates only 0.1%. (83) Another actuarial study of provisions adopted by some states found that full parity would raise premiums by only 0.2%. (84) These estimates take into account only the cost of health insurance itself and not other costs to employers and society as a whole.

Providing employees with access to addiction treatment saves employers money in a variety of ways. A study by the Substance Abuse and Mental Health Services Administration determined that employees who maintained sobriety for one year after inpatient treatment reduced absenteeism by 85%, injuries by 50%, tardiness by 82%, and mistakes by 83%. (85) Treatment for alcoholism demonstrably decreases behavioral problems in the workplace. (86)

Individual employers' studies of their own programs support these findings. For example, at one plant, Oldsmobile found that in the year after undergoing treatment for alcoholism, disciplinary problems among its employees declined by 63%, workplace accidents by 82%, absences by 56%, and total lost person-hours by 49%. (87) Studies by the U.S. Navy and by Illinois Bell found that treatment reduced sick leave by 46% and 51.1%, respectively. (88) In addition, when treatment serves as an alternative to firing employees whose work is impaired by addiction, it saves considerable replacement costs, which for high-skill jobs can run up to $50,000 per vacancy. (89) In total, Chevron estimates it receives a ten to one payoff on its investment in treatment. (90) General Motors estimates that its Employee Assistance Program saves the company $37 million per year. (91)

Society as a whole benefits financially from treatment programs as well. The State of California estimates that its treatment program saved the public seven dollars for every dollar spent. (92) Crime was reduced by two-thirds following treatment, and hospitalizations were reduced by one-third. (93) Other states' studies confirm significant reductions in crime rates due to treatment. (94) In terms of reducing crime, studies have found that treatment is substantially more effective than the criminal justice approach taken in the war on drugs. (95) Treatment also slows the spread of HIV/AIDS. (96) A federal study of five states' parity provisions concluded that the provisions had little effect on premiums and any initial costs were offset by savings to society. (97) A RAND Corporation study found that providing treatment for every addict in the United States would cost $21 billion, but would save over $150 billion in social costs over fifteen years. (98)

Finally, insurance parity laws may help to make treatment even more cost-effective and medically efficacious than it is already. Studies have suggested that the lack of patients whose insurance fully covers addiction treatment operates as a financial disincentive for pharmaceutical companies developing improved treatments. Mandating parity would thus help to stimulate private sector investment in research and development to bring about cheaper and more effective treatments. (99) In addition, because substance abuse is often correlated with mental health disorders, parity may allow greater coordination of care that integrates the two concerns, increasing efficiency and effectiveness. (100)

E. Counterarguments

Although compelling, the case for insurance parity raises complex ethical and practical questions. Opponents of insurance parity mandates have raised a number of objections. In this Section, I consider and respond to some of these, and to other potential problems with the case I have set forth above.

1. The Voluntariness Argument

Some opponents of insurance parity have argued that addiction is different from other medical conditions because it stems ultimately from a choice to engage in unhealthy and socially undesirable behavior. (101) The common perception of alcohol and drug addiction as self-inflicted conditions accounts in large part for the social stigma that attaches to them, as I discuss further below. This perception is largely unjust; addiction is a disease with a documented biological mechanism. Tendencies toward addiction may also have a genetic component, which further undercuts the idea of voluntariness. (102)

Nonetheless, addiction undeniably has an underlying behavioral basis, and the initial behavior that triggers a cycle of addiction is generally voluntary. An individual may be genetically predisposed toward alcoholism, but she cannot become an alcoholic without ever taking a drink. This element of voluntariness may affect our assessment of whether those who suffer from addiction are morally or legally entitled to consideration analogous to that accorded to victims of other diseases.

In defense of insurance discrimination against addiction treatment, then, one could argue that the point of health insurance is to protect people from harms inflicted by chance, or at least by factors outside their control. Insurance is about spreading risk, not spreading the costs of self-inflicted injuries. When a person chooses to drink or use drugs, perhaps she assumes the risk of becoming addicted. Therefore, she should not expect sympathy or solicitude from society if that risk is actualized. (103) To take this analysis a step further, perhaps to cast drug and alcohol users as "victims" of addiction is misleading or even dangerous, because it strips away a sense of personal responsibility for their actions. Under this view, those who choose to drink or use drugs should pay the consequences.

For several reasons, however, the behavioral choice element of addiction does not justify insurance discrimination. As an initial matter, of course, the issue of moral responsibility and desert does not particularly bear on the bulk of the policy argument for insurance parity that I have laid out above. The social and economic costs of untreated addiction are an independent justification for insurance parity regardless of what we may think about whether addicts in some way "deserve" treatment. The majority of the costs of untreated addiction are, in fact, borne by persons other than the addicts themselves: family members, employers, and society as a whole. (104)

Second, the rationale that illnesses caused in part by voluntary conduct should not be covered by health insurance appears disingenuous, or at least hollow, when applied solely to addiction as opposed to other diseases. Many diseases that are routinely covered by health insurance stem at least in part from behavioral choices. The great majority of emphysema cases and a huge portion of lung cancer cases can be traced to cigarette smoking, which also increases the risk of a litany of other health harms. (105) Many illnesses, including heart disease, high blood pressure, and diabetes, are heavily influenced by "lifestyle" factors such as diet and exercise. (106) Also, accidents are frequently the fault of those who are injured in them.

But it is not within the realm of serious discussion to suggest that health insurers should exclude coverage for all these conditions just because they have a behavioral component. A vision of "personal responsibility" extended to such a degree would probably be considered heartless by most Americans. Most of us recognize that people sometimes make lifestyle choices that are in some way unhealthy: We drive too fast, or we smoke, or we do not exercise enough, or we eat too many carbohydrates. Why, then, single out addiction for the type of moral condemnation implied by the voluntariness argument? To deny treatment to drug and alcohol addicts on the basis, essentially, that they have committed a moral wrong or chosen their own fates is not only uncompassionate but also unjust, because it places a staggering burden on a group of people on the basis of an extremely arbitrary distinction between addiction and other behaviorally induced diseases.

Third, the specific behaviors that can eventually give rise to addiction--drug and alcohol use--are or have been engaged in by many, many more Americans than eventually become addicted. (107) In the case of alcohol use, the behavior is not even one that mainstream society condemns. Quantity of use clearly impacts the chance of becoming addicted, but there are many heavy users who do not become addicted. Use of drugs or alcohol is thus a necessary but not a sufficient factor in causing addiction. Genetic predisposition and unchosen life-history factors like physical or sexual abuse may play important roles. (108) Addicts are to some degree victims not just of their own deeds, but of some measure of bad luck. Given this fact, the instinct to blame the addict seems particularly arbitrary and unfair. Why should addicts alone be condemned--or, at least, alone bear the practical costs of this condemnation--when other users are not? The difference between the two groups of people is not that one group engages in a particular behavior while the other does not. Rather, it is simply that one, unlike the other, suffers from a disease. The underlying hypocrisy behind the blame-the-addict approach is most apparent with respect to alcohol, a product that is used openly and without shame by most adult Americans. The stigma of addiction thus does not attach merely to the behaviors giving rise to the illness, but to the illness itself. The alcoholic is viewed in some way as weak--that is, not capable of "handling" a substance that so many others have the ability to enjoy casually. (109)

Finally, the use of personal responsibility arguments to justify health insurance discrimination is particularly problematic, because addicts who seek treatment have, in most cases, made a decision to terminate their self-destructive behavior. Given the nature of the brain disease of addiction, (110) the difficulty of coming to and then following through on this decision should not be underestimated. The voluntariness argument endorses denying access to treatment to someone who is trying to take responsibility for their behavior in the face of an overwhelming biological compulsion, on the basis that he or she should be held responsible for past behavior. The concept of personal responsibility this reflects is harsh, unforgiving, and ultimately self-defeating.

2. The Illegality Argument

Another objection to insurance parity runs as follows: Even if requiring treatment for alcohol addiction is justifiable, addiction to illegal drugs stems from a decision to break the law. Requiring insurers to provide drug users with treatment rewards lawbreaking and is inconsistent with legislatures' and society's decision to condemn this behavior. (111)

The argument that drug addicts have chosen to break the law is in some ways tied to the issue of voluntariness and individual responsibility, and to that extent one may respond to this criticism with many of the same arguments outlined above. First, treatment coverage brings social and economic benefits that do not hinge on the legality of substance use. Second, a large percentage of Americans admit to having used illegal drugs at some point in their lives, (112) but relatively few become addicted. Third, it may be counterproductive to cite individual responsibility as a reason to deny treatment to addicts who have made the responsible and difficult choice to end their use of illegal drugs. Yet there is something to the "illegality argument" beyond the culpability of the individual addict. From a broader perspective, perhaps one could argue that requiring employers and insurers to bear the costs of treatment for illegal drug use is inconsistent with the overall logic of a legal system in which that drug use is a crime. As a practical matter, perhaps guaranteeing addicts access to drug treatment undercuts the central message of the so-called war on drugs: Drug use is a criminal act that has criminal consequences.

Clearly, this objection begs the question of whether drug use should, in fact, be considered a crime. This question is much too complex, and too much of a digression, to be addressed here; countless scholars and advocates have explored it thoroughly and insightfully. (113) Naturally, if one believes that drug prohibition is misguided or wrong, then one's views on insurance discrimination are unlikely to be swayed by an argument that relies on the internal logic of the drug war. Yet one need not start with this belief in order to conclude that insurance parity is morally and legally justified. Therefore, I will take as given two assumptions: that the use of certain drugs is and will remain illegal, and that, as a society, we wish to discourage the use of these drugs. Nonetheless, the current prevalence of health insurance discrimination against drug addicts must be seen in the context of broader drug policy; regardless of the validity of arguments for drug legalization per se, this discrimination is one of the disturbing excesses of modern America's drug war.

One response to the illegality argument is that regardless of what social sanctions we wish to attach to illegal drug use, the health insurance system is not the appropriate venue for those sanctions to be enacted. Decisions about the scope of health insurance coverage should be based on medical science and, as a secondary factor, cost-effectiveness. Questions about criminality belong to the sphere of criminal justice, which should properly remain separate from medical and actuarial decisionmaking. The criminal justice system sets certain penalties for drug use, and it provides procedural protections to the accused. Denial of access to medical treatment because of insurance discrimination, if conceived of as a punishment for drag use, is punishment that goes beyond those penalties prescribed by the criminal law and is administered without those procedural protections. In this sense, it is unfair to the "punished" addict, and it also distorts the objectives of the medical system by making it a tool of the criminal justice system.

Finally, the illegality argument raises the question of what, exactly, the drug war is supposed to achieve. If the goal is reducing drag use--as opposed to supporting the political exigencies of tough-on-drags politicians--then it is hard to see the justification for denying access to medical treatment for drag addiction. The goal of treatment, after all, is to terminate drag use. And treatment is demonstrably effective in reducing or terminating drug use--more so than criminal law enforcement. (114) It would be a sad irony if the characterization of addicts as criminals--one aspect of the stigma addicts face--actually served as a justification for denying them the opportunity to end the very behavior that is illegal.

3. Free Market Arguments

The category of "free market arguments" against insurance parity mandates includes several distinct but interrelated objections. The first set of these objections holds that the assessment of the medical value and cost-effectiveness of different forms of medical treatment is a responsibility best performed by health insurers themselves, who perform countless such assessments as a routine aspect of their business, or by the employers and individuals who select insurance plans with their own best interests in mind. Such an argument might be pragmatic--that is, market actors will actually make better decisions than the government will--or philosophical, grounded in a libertarian, laissez-faire political theory that emphasizes individual choice. In testimony against the Substance Abuse Treatment Parity Act, a representative of Americans for Tax Reform highlighted both aspects of this argument:

   [T]he American people ought to be free to decide what they want in their
   health policy coverage. If they want substance abuse treatment coverage in
   their health policies, they can buy it in the marketplace. Insurers will be
   more than happy to provide the coverage the market demands....

   ... [A] central planning approach is neither efficient nor does it have the
   proper respect for the freedom of choice that the American people should
   have. (115)

Another type of free market argument suggests that the significant coverage gaps that currently exist belie the supposed cost-effectiveness of insurance parity. That is, if addiction treatment really were so cost-effective, insurers and employers would choose to cover it voluntarily. Surely they would not simply ignore dramatic potential savings. (116) The fact that they have not covered treatment suggests that these savings do not exist. This argument is tied to a fourth, pragmatic formulation--that government mandates for health insurance coverage necessarily increase premiums and thereby increase the number of uninsured Americans by making health insurance less affordable. (117) This claim is the central argument of the Health Insurance Association of America (HIAA), the major insurance industry lobbying group, in opposition to the Substance Abuse Treatment Parity Act. (118)

The issue of whether the government should mandate insurance parity is complex in part because of the basic nature of the insurance industry. Health insurance plans cannot possibly cover all forms of medical treatment for all conditions. Some treatments are, for example, astronomically expensive or lack medical value. In order to keep premiums within the economic reach of most Americans, insurers must make some tough choices. Also, employers are able to assess the specific needs of their employees and the economic requirements of their business when they design compensation packages. Maybe, then, insurance professionals and employers are best suited to make these decisions, and the government should not interfere. Perhaps, even, requiring parity for treatment of a particular illness would set a precedent for excessive government entanglement in a business that is best left to the free market.

In response, it should first be noted that government-mandated requirements for insurance coverage of addiction treatment would not lack precedent; a number of laws currently constrain the coverage decisions of insurers and employers. To cite a few examples, federal law already requires mental health treatment parity, (119) many state laws require private insurance plans to cover oral contraceptives, (120) and the Pregnancy Discrimination Act (121) limits employers' ability to exclude maternity care from health insurance policies, (122) Indeed, the HIAA's concern about increased premiums appears to be linked not so much to the specific content of a substance abuse treatment parity mandate, but rather to the cumulative effect of many coverage mandates, most of which already exist, (123) As HIAA notes, "[T]he number of State mandates has increased twenty-five fold during the last two decades." (124) Against this background, concerns that substance abuse parity provisions would set a bad precedent or entangle the government in an area that is currently left to the free market seem off-base. The government routinely regulates the health insurance business.

Second, strong medical and economic evidence, discussed in Sections II.C and II.D above, suggests that the free market simply is not working. Employers and insurers for the most part continue to offer plans that cap insurance coverage for addiction treatment in ways that do not make economic sense. The disparity between what the studies on cost-effectiveness say should be done and what employers and insurers actually do, however, is so stark that it may demand an explanation in addition to correction. That is, it is worth asking why, if the studies are correct, the insurance coverage picture remains so bleak.

Drag and alcohol addicts face an extraordinary degree of social stigma that may prejudice even those employers and insurers who are generally attempting to engage in objective, profit-driven decisionmaking. Studies document the prevalence of these attitudes among the general public. (125) In introducing the Fairness in Treatment Act of 2001 (discussed below in Part IV), Senator Paul Wellstone stated the problem eloquently:

   For too long ... addiction has been viewed as a moral issue, rather than as
   a disease. Too often, a cloak of secrecy has surrounded this problem,
   causing people who have this disease to feel ashamed and afraid to seek
   treatment for their symptoms for fear that they will be seen as admitting
   to a moral failure, or a weakness in character. We have all seen portrayals
   of alcoholics and addicts that are intended to be humorous or derogatory,
   and only reinforce the biases.... I cannot imagine this type of portrayal
   of someone who has another kind of chronic illness. (126)

Stigma persists notwithstanding the existence and publication of scientific evidence demonstrating the nature of addiction as a brain disease. Dr. Alan Leshner of the National Institute on Drag Abuse explains that there is a "dramatic lag" between advances in scientific knowledge and public perception of those advances. (127) This is true as a general rule, but it is especially tree in the case of addiction because of the common perception of addicts as "weak or bad people, unwilling to lead moral lives and to control their behavior." (128)

The problem of stigma is compounded by market failure. The economic decisions of HMOs and other insurers do not take into account the full range of costs addiction imposes on employers and society at large: lost productivity, absenteeism, replacement costs, accidents, and so forth. (129) In determining whether a service is cost-effective, insurers consider only whether it will decrease medical costs overall: Is the cost of the service itself outweighed by its benefits in preventing costlier health problems in the future? Although the studies cited above in Section II.D demonstrate quite conclusively that coverage of treatment saves money for employers and for society as a whole, they do not demonstrate that covering treatment saves money on medical costs alone. Rather, using premium rates as an index of total medical costs, most studies find that covering addiction treatment would raise medical costs by a very small but appreciable margin (raising premiums, in one study, by half of one percent). The overall cost savings come from non-health-care factors, such as increased productivity.

In the highly competitive insurance industry, individual insurance companies may lose customers by raising their premiums by even a small amount. This risk is considerably amplified by the adverse selection problem, which is well documented in the literature on health care reform. (130) Insurance companies, like all businesses, respond to customer demand. Given the choice between one plan that covers addiction treatment and another, slightly cheaper plan that does not, which customers are likely to choose which? For people with no personal or family experience with addiction, choosing a plan with no coverage of addiction treatment makes economic sense. Thus, the only people who pick a plan that covers treatment are those who are probably more likely to need it. (131)

This adverse selection tendency triggers a vicious cycle: Premiums for the plan covering treatment increase more because the population they cover is likelier than the average population to need treatment. Moreover, addicts suffer more than nonaddicts from a range of other health complications, which greatly worsens the adverse selection spiral; health insurance for a less healthy population is simply more expensive. Because adverse selection results from individual choice in a competitive market, a solution is government regulation--making it mandatory for all plans to cover treatment. (132)

Employers internalize some costs that insurers do not, such as absenteeism, accidents, and lost productivity. The studies cited above demonstrate that when these costs are included, addiction treatment coverage becomes clearly cost-effective. Therefore, a rational employer purchasing a group insurance policy should demand broader coverage of addiction even at the cost of slightly higher premiums. But employers may not always act rationally; they are influenced by longstanding misconceptions both about the nature of addiction and about the effectiveness of treatment. Some employers may view it as easier simply to fire an employee rather than deal with her need for substance abuse treatment. In some cases involving low-skilled workers in a tight job market, this may actually be true, but in the great majority of jobs, worker replacement costs alone outweigh even the total cost of treatment. (133) Furthermore, firing probably makes the addict's chances for recovery even worse, amplifying the personal and social costs of addiction; many studies show that unemployment increases substance abuse rates and decreases the chance that treatment, even if it is accessible, will be effective. (134)

Stigma, along with inadequate medical information and failure to internalize relevant costs, may therefore explain the origins of coverage inequalities; it is also a factor impeding reform. In the political climate fostered by the drag war, politicians may be afraid to support pro-treatment policies for fear that they will be labeled "soft on drags." Furthermore, the stigma associated with addiction has impeded the development of pro-treatment social and political movements. (135) Addicts, notwithstanding their large numbers and prevalence in every sector of society, are a relatively politically powerless group, lacking the lobbying might to move legislation. According to William McColl of the National Association of Alcoholism and Drug Abuse Counselors, "[N]o one is stepping up and speaking out, so we haven't tapped into the potential power of people saying `this is a problem for us.'" (136)

Finally, even if all the studies demonstrating that addiction treatment coverage is cost-efficient for employers were wrong, a coverage mandate would still be justified. This is tree even strictly from an economic costs perspective, provided that all costs to society are taken into account. Just as employers internalize some costs that insurance companies do not, government internalizes some costs that employers do not. As discussed in Sections II.C and II.D, the costs of substance abuse, and the savings from treatment, extend beyond the employment realm. These costs include, for example, increased domestic violence, child abuse and neglect, and other crimes; automobile and other accidents; spread of infectious disease; and increased federal and state entitlement spending. Moreover, in addition to the quantifiable financial costs, addiction causes great suffering for its victims and their families. A humane and compassionate government policy ought to take these consequences into account; the complex ethical considerations surrounding health insurance policy cannot ultimately be reduced to a simple calculus of cost-efficiency.

On balance, the case for parity in insurance coverage of addiction treatment is strong. Treatment saves money and lives, and can help recovering addicts avoid tremendous suffering. Because of the large existing coverage gaps, recovery advocates need to look at legislative and litigative options for encouraging change. The remainder of this Note assesses some of these options. I begin by analyzing the currently most effective federal law protecting those who suffer from addiction, the Americans with Disabilities Act.

III. EMPLOYERS' OBLIGATIONS UNDER THE AMERICANS WITH DISABILITIES ACT

The Americans with Disabilities Act, passed in 1990, is landmark civil rights legislation restricting disability-based discrimination in employment, government services, and public accommodations. (137) Drug and alcohol addictions are disabilities under the ADA. This interpretation of the ADA stems from longstanding precedent under section 504 of the Rehabilitation Act, the ADA's antecedent legislation covering discrimination by recipients of federal funds. (138) The legislative history of the ADA clearly points to an assumption that addiction would be covered, and courts applying the ADA have recognized its application to the disability of addiction. (139) Drug addiction and alcoholism are specifically listed as "physical or mental impairments" in the ADA implementing regulations passed by the Equal Employment Opportunity Commission (EEOC). (140) The ADA, however, provides no protection for current users of illegal drugs; it does cover alcoholics and drug addicts who are already in recovery. (141) This exclusion might be understood to reflect Congress's adherence to the "illegality argument," discussed above in Subsection II.E.2, or a simple reluctance to be seen as soft on drug users.

Moreover, according to the EEOC, disability-based discrimination in employer-provided health insurance plans is prohibited by the ADA. In its 1993 implementation guidance on the application of the ADA to health insurance, the EEOC stated that its ADA implementation regulations explicitly covered discrimination in "fringe benefits," and that application to insurance benefits was specifically contemplated in section 501(c) of the Act. (142) The EEOC further noted that this application also prevents employers from discriminating in hiring: "[D]ecisions about the employment of an individual with a disability cannot be motivated by concerns about the impact of the individual's disability on the employer's health insurance plan." (143) Finally, "insurance offices" are also defined as "places of public accommodation" under Title III of the ADA. (144)

At least in theory, then, the ADA may prohibit some types of insurance discrimination against addiction treatment. This theory has not been tested by litigation. A realistic assessment of other strands of ADA jurisprudence, however, suggests that courts are increasingly reluctant to scrutinize closely the content of insurance policies, and that advocates pursuing an ADA claim may face an uphill battle.

A. Overview of ADA Litigation Strategies

A challenge to disability-based discrimination in private health insurance might proceed under Title I or Title III of the ADA. (145) Title I governs employment discrimination, and Title III governs discrimination in public accommodations. Title I would only apply to situations where the challenged health insurance policy is provided by an employer, although the lawsuit could be brought either against the employer or against the employer-contracted health insurance company directly. (146) In contrast, a Title III lawsuit alleging that the provision of health insurance is a "public accommodation" would be brought directly against the health insurance company. (147) Title III is thus a potential strategy for challenging health plans that are not employer-provided.

Title I of the ADA prohibits employers from "discriminat[ing] against a qualified individual with a disability in ... compensation ... and other terms, conditions, and privileges of employment." (148) This provision prohibits direct, disparate-treatment-type discrimination. So, for example, an employer may not refuse to hire a person because she is wheelchair-bound, unless her disability prevents her from performing the tasks of the job satisfactorily. In addition, however, Title I places upon employers the affirmative responsibility to provide "reasonable accommodations" for employees' disabilities. (149) An employer must make any adjustments in the working environment (including the physical facilities) that are reasonably necessary to enable a disabled employee to do the job. An employer is not required to provide an accommodation if to do so would cause "undue hardship" (an affirmative defense). (150)

Challenges to disability-based distinctions in insurance coverage have generally been framed as disparate treatment claims, and a Title I challenge to substance abuse treatment limitations could readily proceed along the same lines. A Title III claim would be similar, but rather than alleging discrimination in the compensation, terms, and conditions of employment, the claimant would argue that the health insurance company has an obligation to provide services in a manner that does not discriminate based on disability. The argument would be that the company's choice to provide less coverage for substance abuse treatment disadvantages those customers who suffer from the disability of addiction. This constitutes discrimination in "places of public accommodation," a term defined in Title III to include insurance offices. (151)

B. Legitimacy of Some Disability-Based Distinctions: ADA Section 501(c)

Not all distinctions in health care coverage based on disabilities are illegal. Health insurers, in fact, are in the business of making just such distinctions. In setting the scope of health benefits, insurers routinely distinguish among ailments, many of which constitute "disabilities" under the ADA. The question, then, is how to determine whether a distinction is legitimate and not discriminatory. Section 501(c) of the ADA provides some guidance. (152)

Section 501(c) provides a safe harbor from the ADA's requirements. It explicitly permits employers and insurers to make disability-based distinctions in health insurance plans under two conditions. First, the distinction must be "bona fide"--it must be based on "underwriting risks, classifying risks, or administering such risks that are based on or not inconsistent with State law." (153) This criterion is not strict; it essentially requires the health plan to prove that it is a real, existing, benefits-paying plan that is legally organized under applicable state law. (154) The heart of the section 501(c) inquiry is the second condition: The distinction in question must not be a "subterfuge to evade the purposes of [the ADA]." (155)

According to the EEOC, "subterfuge" is defined as "disability-based disparate treatment that is not justified by the risks or costs associated with the disability." (156) This cost-benefit analysis is conducted by courts on a case-by-case basis.(157) The employer has the burden of proving that the distinction is not a subterfuge. There are several ways the employer might prove this. First, it could argue that the distinction is "justified by legitimate actuarial data ... and that conditions with comparable actuarial data and/or experience are treated in the same fashion." (158) Second, it could argue that to eliminate the distinction--to provide the requested coverage--would be financially catastrophic, either causing the health plan to become insolvent or causing a "drastic" change in premiums or cutback in the remainder of the plan. Third, it could argue that the treatment in question can be shown by "reliable scientific evidence" to have "no medical value." (159)

The second and third of these arguments are not plausible in the addiction treatment context. Even if addiction treatment were not cost-effective--contrary to the evidence discussed above--it is inconceivable that the costs could be enough to cause insolvency or "drastic" premium increases. The best estimates place premium increases at less than 0.5%, (160) which is much less than any reasonable interpretation of "drastic." Furthermore, even the most pessimistic assessments of addiction treatment's effectiveness do not suggest that it has absolutely "no medical value"---even a thirty percent success rate would far exceed those of many commonly available types of medical treatment.

The first argument outlined above--that the distinction is consistent with actuarial data--is more complicated because it requires a close analysis of the particular plan in question. The crux of the issue is, is addiction treatment covered to a lesser extent than are comparably effective medical treatments for comparably severe conditions? The answer to this question is usually yes; many plaintiffs should be able to show (and it would be the employer's burden to disprove) that substance abuse treatment is singled out for discriminatory limitations. This singling out demonstrates that the disability-based distinction is a subterfuge, and therefore the section 501(c) exception to the ADA's requirements should not apply.

C. Courts' Differing Approaches: The Mental Health and AIDS Cap Cases

In recent years, federal courts have set some criteria for determining the legitimacy of disability-based distinctions in health insurance coverage, mostly in cases involving challenges to caps on coverage of AIDS-related or mental health conditions. (161) These cases involve limits on annual or lifetime insurance payments for AIDS-related or mental health conditions that do not apply to payments for other conditions. "AIDS caps," which were until recently common, are closely analogous to the provisions of most insurance plans regarding substance abuse treatment. Mental health caps were also common, but have been directly outlawed by the Mental Health Parity Act of 1996. (162) ADA challenges to both types of cap have proceeded under Titles I, II, and III.

The EEOC has clearly stated that it believes that AIDS caps are presumptively illegal under the ADA. (163) Nonetheless, many courts have disagreed. The first federal court of appeals to consider the issue was the First Circuit in Carparts Distribution Center, Inc. v. Automotive Wholesaler's Ass'n of New England. (164) In Carparts, the district court had rejected Title I and Title III challenges to an AIDS cap. The appeals court reversed and remanded on both issues, but its holdings were ambiguous as to the legality of the AIDS cap. On the Title I issue, the First Circuit held that the district court had used the wrong interpretation of the term "employer" in holding that an insurance plan was not an employer under the ADA. The appeals court instead followed the more expansive interpretation of "employer" found in case law interpreting Title VII of the Civil Rights Act of 1964, (165) and remanded to the district court to determine whether an insurance company fell within that definition. (166) Under this interpretation, an employer is "any party who significantly affects access of any individual to employment opportunities," regardless of whether that party would have been considered an employer at common law. (167) In the Title VII context, that interpretation has been held to encompass benefit program administrators. (168)

Turning to the ADA Title III claim, the Carparts court considered the issue of whether an insurance company could be a "place of public accommodation" even if it did not have a concrete physical structure like a store. The court declined to make more than a "threshold" ruling on this issue, but noted that the statutory language was ambiguous and that the legislative history demonstrated that Congress contemplated application to service companies that lacked "definite physical boundaries. (169) It remanded the case to the district court for further consideration and noted that the factual elements of the Title III claim were as yet poorly developed in that case. On remand, the district court denied the defendants' motions for summary judgment on both the Title I and the Title III claims. (170)

Notwithstanding these ambiguities, it is probably fair to read Carparts to stand for two main propositions, both favorable to plaintiffs. First, because insurance companies may be considered "employers" under Title I, victims of discrimination in insurance coverage may sue the insurers directly rather than suing their actual employers. Second, an insurance company, even one without a discrete "office" in which the discrimination took place, is a "place of public accommodation" under Title III. Note that neither of these holdings addresses the substantive issue of whether an AIDS cap constitutes disability discrimination under the ADA.

The Sixth Circuit came closer to resolving this question in the mental health cap context in Parker v. Metropolitan Life Insurance Co. (171) The Sixth Circuit panel held that discrimination in long-term disability insurance coverage was encompassed by the "place of public accommodation" provision of Title III, even though the plaintiff did not actually physically go to an insurance office. It went on to consider whether the mental health cap was a legitimate disability-based distinction or a "subterfuge" under the section 501(c) safe harbor provision. The panel reversed the district court's holding that "subterfuge" required an intent to discriminate, and held instead that subterfuge was simply a policy criterion that was "based on speculation, and not on sound actuarial principles, actual or reasonably anticipated experience, or bona fide risk classification." (172) It remanded the case to the district court to determine whether the policy met this definition of" subterfuge."

The Parker case, however, never reached the stage of applying that definition. The Sixth Circuit agreed to rehear the case en banc, and proceeded to reverse the panel holding in an 8-6 decision. (173) Specifically, the en banc court held that for the "place of public accommodation" requirement of Title III to be met, plaintiffs would need to have physically visited the office of the insurance company to request coverage. Thus, the court held that there was no Title III violation and that therefore the issue of the section 501(c) safe harbor was irrelevant. (174) The panel decision, including the discussion of the interpretation of "subterfuge," was vacated.

The Parker court went further, stating in dicta that the mental health cap did not actually discriminate because the same benefits package was offered to disabled and nondisabled employees. Because employees without mental illness were also denied coverage for mental health treatment, no discrimination occurred even though only the disabled employees would actually have needed treatment. (175) The court cited the Seventh Circuit's decision in EEOC v. CNA Insurance Cos., which followed a similar rationale with respect to limitations on mental health coverage in long-term disability insurance plans. (176) The court in CNA Insurance had cited the need to abstain from making a controversial ruling on insurance parity at the time that the Mental Health Parity Act was debated in Congress. (177) The main difference between CNA Insurance and Parker is that CNA Insurance was an interpretation of Title I, not Title III. The Third and Fourth Circuits subsequently made similar decisions in other cases involving mental health treatment limitations. (178)

The main dissent in Parker, joined by five judges, criticized the majority for ignoring legislative history that clearly showed an intent to encompass the content of health insurance policies, an intent reflected in the adoption of section 501(c) that the majority's interpretation would make largely irrelevant. (179) A separate dissent by Chief Judge Martin noted that the "purpose of Title III is to `bring individuals with disabilities into the economic and social mainstream of American life'"--a purpose that is inconsistent with the artificial, overly technical approach taken by the majority. (180) This criticism was directed at the majority's interpretation of the "place of accommodation" language, but it is equally applicable to the majority's dicta regarding the substantive issue of whether discrimination occurred. The Parker-CNA Insurance approach is wholly artificial; it is true but irrelevant that people who do not need mental health benefits are alSO deprived of them. This approach relies on the flawed assumption that the goal of antidiscrimination law is always like treatment, even in unlike cases. The ADA--which, after all, defines the failure to provide special accommodations as discrimination--rejected this narrow view in favor of a more realistic vision of what is necessary to level the playing field for disabled Americans.

The Seventh Circuit raised further barriers to ADA challenges to insurance coverage limitations in Doe v. Mutual of Omaha Insurance Co., an AIDS cap case. (181) Chief Judge Posner's majority opinion held that Title III, in general, does not reach the content of goods or services provided at a place of public accommodation. Specifically, so long as the same benefits package is available to all consumers, the ADA does not oblige an insurance company to provide any particular set of benefits. The court analogized the plaintiff's claim to a blind person's demand that a particular store sell books in Braille, or a one-legged person's complaint about a shoe store only selling shoes by the pair. (182) The court concluded, "Had Congress purposed to impose so enormous a burden on the retail sector of the economy and so vast a supervisory responsibility on the federal courts, we think it would have made its intention clearer." (183)

The Second Circuit, in Pallozzi v. Allstate Life Insurance Co., explicitly rejected the Seventh Circuit approach. (184) Citing the First Circuit's decision in Carparts, the court held that Title III must be understood to extend to the content of health insurance policies in order to effectuate Congress's "`clear and comprehensive national mandate'" against disability-based discrimination in all aspects of social and economic life. (185)

The rationales of the Sixth Circuit in Parker and the Seventh Circuit in Mutual of Omaha are applicable only to Title III suits against insurance companies. The problem in Title III litigation arises from the issue of what constitutes a "place" of public accommodation. Title I has no such language, and its ban on discrimination in the "terms and conditions of employment" seems clearly to extend to the content of benefits packages. Employees may therefore still be able to sue their employers under Title I.

In addition to CNA Insurance in the Seventh Circuit, however, recent precedents in other circuits cast substantial doubt on whether the Title I option remains viable, at least outside the First Circuit, in which Carparts remains good law as to both Title I and Title III. In Ford v. Schering-Plough Corp., the Third Circuit followed the rationale in CNA Insurance and Parker that no discrimination could have occurred where every employee "had the opportunity to join the same plan" and therefore received "equal treatment." (186) A previous Eighth Circuit opinion had applied a similar rationale to insurance caps on fertility treatment. (187) Similar decisions were subsequently reached in the Second, Fourth, Ninth, and Tenth Circuits. (188)

As Parker demonstrates, the substantive issue is basically the same under both Titles I and III: Does the provision of facially equal benefits to differently situated persons constitute discrimination? Recent trends demonstrate that under both Titles, courts are increasingly answering that question in the negative. The First Circuit remains a holdout, although even it has not resolved the substantive question but has simply removed some procedural bars to ADA challenges.

D. Assessment of ADA Litigation Potential in Addiction Treatment Cases

Although the circuits remain split to some extent, and some have not yet ruled on these issues, there is a decided trend among courts of appeals to reject ADA challenges to disability-based distinctions in health insurance, regardless of whether they arise from Title I, Title II (in government employment situations), or Title III. This is an unfortunate development. CNA Insurance and its progeny represent an artificial and overly formalistic view of equality that has no place in the interpretation of the ADA, which was designed as a "transformative statute, requiring not only formal equality, as the non-discrimination concept had traditionally been understood, but also structural equality--the accommodation of difference." (189) The ADA was not designed to enforce the principle that people with different needs must all be treated the same--quite the opposite. A better interpretation of the ADA, one more faithful to its purpose, would require not that all employees be offered the same benefits packages, but rather that the benefits packages they are offered be structured in a way that realizes the statute's vision of true equality.

The situation presented by the insurance cap cases is this: Two employees, one suffering from a disability and the other not, are both presented with an insurance plan that limits or excludes coverage for that particular disability. For the disabled employee, this is a tremendous limitation, one with enormous life consequences that may prevent her from effectively doing her job. For the nondisabled employee, the limitation is totally insignificant. In what meaningful sense can this insurance plan be said to treat the two employees equally? For one the insurance plan is sufficient, and for the other it is grossly inadequate, with the distinction solely based on disability.

Ultimately, the statutory plan of the ADA offers a way out of this dilemma: Nondiscrimination under the ADA means that the employer must either provide benefits that meet the reasonable needs of all employees, or provide special accommodations for its disabled employees--in this case, a different insurance plan. The CNA Insurance line retreats from this vision of a broader equality mandate in favor of the limited antidiscrimination principle the ADA was designed to transform. In fairness to the courts, perhaps this development is due to the pleadings of the litigants, who appear to have framed their claims as simple allegations of disparate treatment. (190) Future litigants should consider moving beyond this strategy.

Two specific clauses of Title I's definition of "discrimination" explicitly go beyond the formal equal treatment principle. Subsection 102(b)(3) prohibits "utilizing standards, criteria, or methods of administration ... that have the effect of discrimination on the basis of disability." (191) Subsection 102(b)(5)(A) requires that employers make "reasonable accommodations to the known physical or mental limitations of an otherwise qualified individual," unless such accommodations would impose an "undue hardship" on the employer. (192) Title III contains a similar disparate impact prohibition and also states that places of public accommodation must make "reasonable modifications in policies, practices, or procedures" when they are "necessary" to afford access to goods and services to disabled individuals, unless such modification would "fundamentally alter" the nature of the goods and services. (193) I discuss reasonable accommodation here, but similar analysis could apply to Title III's reasonable modification clause.

The existing challenges to disability-based insurance caps have not been crafted explicitly as disparate impact challenges or as requests for reasonable accommodation. A challenge that was so crafted might have a better chance of success. (194) Nonetheless, it may prove difficult to persuade courts to bring insurance and other benefits within the rubric of "reasonable accommodation," where they have not traditionally been found. Reasonable accommodation encompasses only those changes necessary to enable the employee to perform the job. (195) Benefits are typically understood to be outside this spectrum for logical reasons; the question of compensation is different from the question of job performance. In the case of substance abuse treatment, however, litigants in some cases may be able to prove that receiving treatment is necessary to enable them to perform the duties of the job. Such a conclusion would be consistent with existing case law, although the accommodation required by courts to date has been limited to matters such as sufficient leave time to obtain treatment on one's own, and has not extended to treatment funding. (196) If they can further prove that health insurance coverage is necessary to enable them to obtain that treatment, they may be entitled to a reasonable accommodation under the ADA. Alternatively, the employer could simply be required to pay for the treatment directly. Such an interpretation seems eminently reasonable and consistent with the nuanced notion of equality embodied by the ADA.

In any event, regardless of the potential of reasonable accommodation lawsuits in particular cases, they do not provide a broad solution to the problem of insurance discrimination against addiction treatment. The nature of reasonable accommodation is piecemeal--a particular employee gets a specific accommodation tailored to her interests. A court's finding that a reasonable accommodation was required for a specific employee would do nothing to change the employer's health insurance plan obligations toward the rest of its employees--health plans could continue to impose discriminatory benefit caps. Case-by-case litigation is an inherently inefficient way to achieve change, and it ensures that parity will only be provided to those plaintiffs who are able to meet the complicated requirements of the ADA on an individualized fact assessment. Furthermore, for a reasonable accommodation cause of action to lie, the plaintiff must formally request such accommodation from the employer and be denied. In many instances, addicts seeking treatment may not wish to make such a request because of privacy concerns?

Although ADA suits always turn on individualized facts, disparate impact litigation may hold the potential to bring about a broader remedy than would disparate treatment suits. Under Title I's disparate impact clause, a plaintiff would have to show that the defendant utilized "standards, criteria, or methods of administration ... that have the effect of discrimination on the basis of disability." (198) A holding that a given health insurance policy disparately impacted a disabled plaintiff might require that the policy itself be changed, rather than that the individual plaintiff simply be accommodated. It might require parity. Only qualified employees with disabilities under the meaning of the ADA, however, would have standing to challenge a company's failure to provide this parity.

Beyond the issues raised in the existing AIDS and mental health cap cases, there remain a number of hurdles for substance abuse treatment advocates to overcome in pursuing ADA litigation. For one thing, none of these cases has yet reached the substantive issue of what constitutes a legitimate actuarial decision by insurance companies as opposed to a subterfuge. This is an issue that may prove to be complex and require extensive factual scrutiny--a task in which courts may be reluctant to engage. Furthermore, no court has yet applied these precedents to addiction treatment, which may raise another issue: Plaintiffs will have to show that substance abuse treatment caps are, in fact, a disability-based distinction. Showing that addiction is a disability is fairly straightforward, given existing case law. (199) Not every person who seeks treatment for substance abuse, however, is necessarily an addict. (200) Some may have substance abuse problems that fall short of the statutory definition of a disability. There is no case law defining substance abuse itself as a disability.

Indeed, it may be that the determination of whether an individual claimant has a disability will be highly individualized regardless of whether that claimant is actually an addict. The ADA defines a disability as "(a) a physical or mental impairment that substantially limits one or more of the major life activities of [an] individual; (b) a record of such an impairment; or (c) being regarded as having such an impairment." (201) Courts determine whether this definition is met on a case-by-case basis, rather than identifying whole classes of ailments that count as disabilities per se. (202) An individual litigant (addict or not) who is suffering from a substance abuse problem has to show that the condition substantially limits one or more major life activities, while a person in recovery has to provide similar evidence of a past impairment.

Thus, although the best interpretation of the ADA would find a violation in some cases of insurance discrimination against addiction treatment, the trend in recent case law offers a fairly daunting, though not hopeless, outlook for advocates. Moreover, ADA cases are generally an uphill battle for plaintiffs, and courts are becoming increasingly hostile. Employers have won ninety-two percent of all ADA cases that have made it to court, and countless others have not made it that far. (203) The Supreme Court has begun to circumscribe the reach of the ADA as a whole by, for example, cutting deeply into its enforceability against state governments. (204) And, of course, the ADA provides no remedy for those who currently suffer from addiction to illegal drugs, because of the specific exception built into the law. Therefore, even if the courts interpret the ADA's application to health insurance policies in the ways advocated here, recovery advocates will need to think about new strategies beyond the ADA.

IV. ALTERNATIVES: CURRENT LEGISLATIVE PROPOSALS

A. Provisions and Advantages of Current Legislative Proposals for Substance Abuse Treatment Parity

In 1996, Congress passed the Mental Health Parity Act, which mandated insurance parity for mental health treatment, but excluded substance abuse treatment. (205) In the wake of this law, the Substance Abuse Treatment Parity Act was first introduced in 1997 and reintroduced in each subsequent Congress (1999 and 2001). The most recent version of the bill is called Fairness in Treatment: The Drug and Alcohol Recovery Act of 2001 in the Senate, (206) and the Harold Hughes-Bill Emerson Substance Abuse Treatment Parity Act of 2001 in the House. (207) Both were introduced in March 2001; as of April 2002, the Senate bill had five cosponsors and the House bill had thirty-nine. (208)

The Substance Abuse Treatment Parity Act would apply to all group health plans, not just to those provided by employers. As its name implies, its fundamental requirement is parity for substance abuse treatment. The Act does not require insurance plans to cover addiction treatment. Rather, it forbids plans that do cover addiction to impose discriminatory limitations on that coverage that do not apply to other conditions. (209) The bill exempts employers with fewer than fifty employees. The House version of the bill also exempts employers who would incur a premium increase of more than one percent by complying with the Act. (210)

Another bill currently pending in the House would have similar effects, although it would not include the one percent premium increase cap. In 2001, Representative Marge Roukema reintroduced amendments to the Mental Health Parity Act of 1996 first introduced in 1998 that would, among other changes, expand the legislation to include substance abuse treatment. (211) The functional effect on addiction coverage would be similar to that of the Substance Abuse Treatment Parity Act because that bill is in turn modeled on the Mental Health Parity Act. Representative Roukema's bill, however, contains substantially less detail regarding its application to substance abuse specifically--for example, what benefits are covered. No analogous legislation has been introduced in the Senate. (212) Nonetheless, this bill seems to be gaining strength in the House; in March 2002 it had 202 cosponsors, suggesting that its passage in the future is a strong possibility. (213) The analysis in the following Section focuses mostly on the Substance Abuse Treatment Parity Act, because of that bill's greater level of detail, but it is applicable to both proposals unless otherwise noted.

Both bills appear to apply a fairly broad parity standard to substance abuse treatment generally. This is commendable, although some clarification as to what treatments are covered may be necessary. (214) In any case, either insurance parity bill would represent an important step toward eradicating insurance discrimination against people who suffer from addiction. The fundamental concept of parity recognizes that addiction is a disease like any other and should be treated accordingly. It is because it captures the basic spirit of fairness that parity has become the focal point, in terms of federal lobbying efforts, of the recovery movement. (215) If passed, these bills will probably enable a large number of people to acquire treatment that might otherwise have been inaccessible.

B. Shortcomings

Although an important first step toward filling the gap in coverage of addiction treatment, the Substance Abuse Treatment Parity Act has a number of serious limitations that are the products of political compromise. These limitations may facilitate passage of the Act, but they will hamper its effectiveness to some extent. First, the Act does not require any employer or health plan to provide addiction treatment benefits in the first place. Rather, it simply mandates that if they do provide such benefits, plans cannot place monetary or other limits on them that do not apply to medical benefits generally. Thus, it would become illegal to cap substance abuse coverage at $5000 (unless all medical benefits had the same cap), but it would remain perfectly legal to exclude coverage entirely. Essentially, the worst plans in terms of parity are excluded from the Act's requirements.

This limitation may actually be less serious than it appears. The great majority of insurance plans currently do provide some coverage of substance abuse treatment--only with capped benefits--meaning that the provisions of the Act would apply to them. There is a danger, however, that the Act would create a perverse incentive for companies to slash treatment benefits entirely rather than increase them to the levels required by the parity principle. Although the extent to which this would take place is completely speculative, the danger that some employees would wind up worse off is a good reason to eliminate this loophole.

Second, the Act exempts employers with fewer than fifty employees. This is a serious limitation because most large businesses already provide addiction treatment coverage; it is small and mid-size businesses that do not. Although other antidiscrimination statutes also tend to exempt small businesses, the cutoff is generally lower: fifteen employees. (216) The small business exemption was put into the bill as a political compromise. (217) There are significant political obstacles to the imposition of any form of regulatory burden on small businesses, some of which are grounded in economic arguments that make some intuitive sense. Small businesses are less able to absorb the cost of regulatory changes; they cannot realize certain economies of scale (such as the creation of a self-administered EAP, which requires substantial resources within the company). They also do not have the inherent ability to spread risk that large businesses do. That is, suppose on average one in every twenty-five employees requires substance abuse treatment at some point. A company employing 400 people will have to treat sixteen, on average; if they are unlucky, they may have to treat twenty or even twenty-five, but there is unlikely to be a huge statistical aberration over such a large sample size. In a company with fifteen employees, however, the sample size is low, and it is certainly possible that the company will have to treat two or three employees, at a cost that may impose a serious financial burden.

Although these concerns are serious as applied generally to regulation of small businesses, there are at least two reasons why they do not justify excluding small businesses from the requirements of the Substance Abuse Treatment Parity Act. First, as discussed in Section II.D, a number of studies demonstrate substantial overall cost savings to businesses as a result of increased investment in substance abuse treatment. Larger businesses may benefit more from treatment coverage due to economies of scale and greater bargaining leverage in the insurance market, which may allow them to expand coverage less expensively. Still, the magnitude of the positive returns is so great that it seems impossible that small businesses would not also benefit. (218) In small businesses as well as large, employees who undergo treatment are dramatically more productive, healthier, and less likely to have accidents or take sick and injury leave than are untreated addicts. The fifteen-person business with three people who need treatment may therefore benefit the most from insurance parity. (219)

The second reason that the small business exemption is illogical relates to the basic purpose of the legislation: insurance parity. Every argument regarding the burden the Act places on small businesses can be applied to insurance coverage generally; for this reason, small businesses are less likely than large ones to provide health insurance benefits in the first place. Compared to the cost of treating many other medical conditions, addiction treatment is a bargain, even if the small business ignores its long-term benefits. A fifteen-person company that had to cover three employees, or even one, for cancer treatment would be in tremendous trouble. (220) The Substance Abuse Treatment Parity Act does not mandate that businesses provide health insurance, nor does it prevent them from placing annual or lifetime caps on health benefits. (The wisdom of such employer mandates is the subject of a much broader debate over health care reform.) Rather, it simply states that in whatever coverage employers do provide, they cannot discriminate against substance abuse treatment. If a small business can absorb the cost of providing comprehensive insurance coverage for hundreds of other conditions, many of which are more costly than addiction, there is no reason the cost of coveting substance abuse treatment would be uniquely prohibitive. Insurance parity is a very modest requirement, economically speaking; it demands only fairness.

Another problematic limitation on the Act's scope is a catchall provision that exempts any employers that would incur a premium increase of at least one percent if they complied with the Act. This attempt to minimize the Act's economic impact by legislative fiat, in theory, ought to be largely irrelevant. If the studies supporting adoption of the Act are correct, companies should be able to provide insurance parity for less than a one percent increase in premiums. (221) Nonetheless, the loophole is potentially worrisome because it allows for creative accounting. Companies know they are off the hook if they make cost projections that exceed one percent, so they have a perverse incentive to look for the least cost-effective plans--which they will then never have to adopt. Furthermore, it should be noted that if this provision is adopted, it would make the small business exemption wholly unnecessary, as any small business facing substantial cost increases would be exempted anyway.

The amendments to the Mental Health Parity Act of 1996 introduced by Representative Roukema would avoid this problem with the Substance Abuse Treatment Parity Act. The Mental Health Parity Act includes the same one percent increase cap. (222) The Mental Health and Substance Abuse Parity Amendments would abolish this cap in addition to expanding the Mental Health Parity Act to include substance abuse treatment. (223) Thus, this proposal would solve this problem with the Substance Abuse Treatment Parity Act, although it would still not solve the other flaws outlined here.

V. CONCLUSION AND RECOMMENDATIONS

Drug and alcohol addiction are common diseases that cause untold suffering in the United States each year, suffering that is exacerbated by the stigma and discrimination currently embedded in our social, economic, and legal structure. Simply providing access to medical treatment for these diseases could help to alleviate a large measure of this suffering. Encouraging treatment would also have significant economic and social benefits in terms of increased productivity; decreased absenteeism; decreased crime, accidents, and health care costs; and more intact families. Sadly, for the great majority of Americans, proper treatment is inaccessible because of inadequacies in health insurance coverage. Even though most health insurance plans nominally cover addiction treatment, almost all limit that coverage with an array of payment caps and other constraints that do not apply to treatment for other diseases. This is discrimination, pure and simple. The medical establishment has long since recognized the nature of addiction as a brain disease and the necessity for effective forms of treatment. Public perceptions may, ever so slowly, be starting to change. The time has come to make addiction treatment accessible, and the first step toward increased access must be parity in insurance coverage.

Because most insurance policies are provided by employers, lack of insurance parity is an employment discrimination issue. The Americans with Disabilities Act provides recovery advocates with one possible litigation strategy, both against employers and against insurance companies as "public accommodations." I conclude that, at least in many cases, the ADA should be interpreted to prohibit the caps on substance abuse coverage that pervade most insurance policies. Addiction is a disability, and the failure to treat addiction like analogous diseases constitutes discrimination on the basis of this disability, in particular if discrimination is understood to encompass policies with disparate impact or failure to provide reasonable accommodations for addicts. The federal courts have yet to hear ADA lawsuits to achieve parity in insurance coverage for substance abuse. Unfortunately, however, the courts have been increasingly unfriendly toward other litigation challenging various disability-based distinctions in insurance policies. Different litigation strategies might bring more success, but the task is formidable.

Ultimately, both the courts' intransigence and the inherent limitations of the ADA--for example, the exemption of current drag users and the piecemeal nature of case-by-case litigation--mean that a tree solution to insurance discrimination demands legislative change. Two currently pending bills, the Substance Abuse Treatment Parity Act of 2001 and the Mental Health and Substance Abuse Treatment Parity Amendments of 2001, provide potential solutions. Passing such legislation is an essential first step toward the achievement of parity, and Congress should do so without delay. Yet, certain limitations will hamper these bills' effectiveness, and advocates at both the federal and state levels should press for broader legislation. This legislation should apply to small as well as large employers, should mandate that addiction treatment be covered by any plan that covers treatment for diseases of similar severity, should clearly cover all medically effective forms of treatment including methadone maintenance, and should not contain loopholes insulating employers from any possible premium increases.

Recovery advocates should be creative in the search for solutions to the shortfall in insurance coverage of addiction treatment. Employer-provided health insurance is not the only solution; treatment could be funded through other private or public means. In fact, because of the public savings that treatment brings--for example, decreased crime, increased productivity, and decreased accidents--public funding of addiction treatment for the uninsured is justifiable based on cost alone. (224) On the other hand, a comprehensive funding effort may be politically implausible on the national level. National drug treatment programs, despite being long cited as a priority by politicians, remain woefully underfunded. (225)

While working for change in the legal system, advocates naturally should not forget that change can sometimes be brought about more quickly through education campaigns and social activism. Businesses are affected by prevalent social stigma, but they are also fundamentally self-interested creatures. They can be, and many have been, convinced to adopt parity in their own insurance plans simply on the ground that in the long run it will save them a lot of money. Advocates should work to change misconceptions about the effectiveness and cost-effectiveness of treatment and to transform the public misconception that addicts are criminals or derelicts who have brought their conditions on themselves by choice. Addiction is a serious disease, and its victims deserve understanding, respect, and compassion; they also deserve access to effective medical treatment. In the end, it is a question of simple fairness.

(1.) Shift in Congress on Drug Policy Reform: Legislation Eliminates Discrimination in Health Plans for Substance Abuse Treatment, U.S. NEWSWIRE, Mar. 22, 2001, LEXIS, Nexis Library, U.S. Newswire File (citing estimates by Physician Leadership on National Drug Policy).

(2.) See infra Section II.D.

(3.) U.S. Census Bureau, Health Insurance Coverage: 2000, http://www.census.gov/hhes/hlthins/hlthin00/hlt00asc.html (last modified Sept. 28, 2001) (stating that 64.1% of Americans are covered by employer-provided health insurance).

(4.) According to the U.S. Census Bureau, in 2000, 24.2% of Americans had government insurance while about 72.4% carded private insurance, whether obtained through an employer or otherwise. Fourteen percent of Americans are uninsured. Id. Note that the sum of these percentages (72.4% + 24.2% + 14%) exceeds 100% because many people carry two or more types of insurance. See id. (noting, for example, that 7.6% of Americans carry both private health insurance and Medicare).

(5.) Ronald Brownstein, A New Worry About an Old Concern: Health Care, L.A. TIMES, Jan. 7, 2002, at A14.

(6.) Throughout this Note, I use the terms "addiction treatment" and "substance abuse treatment" interchangeably unless the context demands otherwise. "Addiction" and "substance abuse" are not the same thing from a clinical or practical perspective, but the policy literature on insurance parity often uses the terms interchangeably, since treatment regimens for both are similar and are treated the same way by insurance policies. Cf. Nicholas F. Fleming et al., What Are Substance Abuse and Addiction?, in SOURCE BOOK OF SUBSTANCE ABUSE AND ADDICTION 3, 8 (Lawrence S. Friedman et al. eds., 1996) (defining "addiction" as "compulsive use and impaired control of intake of a drug despite its adverse consequences," and "substance abuse," from a diagnostic perspective, as "intermittent impaired control of substance use," and, as a colloquial term, as "drug use that violates social standards or causes self-harm").

(7.) In addition to monetary caps, these limitations include caps on the number of days of care or on the number of visits, higher copayments, and higher deductibles. Substance Abuse Treatment Parity: A Viable Solution to the Nation's Epidemic of Addiction?: Hearing Before the Subcomm. on Criminal Justice, Drug Policy, and Human Resources of the House Comm. on Government Reform, 106th Cong. 27 (1999) [hereinafter House Hearings] (statement of Rep. Ramstad). Copayments as high as 50% are common. David C. Lewis, Limits on Substance Abuse Benefits and the Quest for Parity with Other Chronic Diseases, BROWN U. DIG. ADDICTION THEORY & APPLICATION, Oct. 1996, at 12, 12.

(8.) Janet Firshein, PBS Online, Legislating Insurance Parity for Addiction Treatment, at http://www.pbs.org/wnet/closetohome/policy/html/insurance.html (last visited Feb. 27, 2002) (citing Norman Hoffman of the Center for Alcohol and Addiction Studies at Brown University).

(9.) Thomas R. Burke, The Economic Impact of Alcohol Abuse and Alcoholism, 103 PUB. HEALTH REP. 564, 567 (1988).

(10.) Am. Acad. of Pediatrics, Policy Statement: Improving Substance Abuse Prevention, Assessment, and Treatment Financing for Children and Adolescents, 108 PEDIATRICS 1025, 1026 (2001) (noting that families "rapidly exhaust their annual and even lifetime allotment of substance abuse benefits"); see also Am. Acad. of Pediatrics, Policy Statement: Insurance Coverage of Mental Health and Substance Abuse Services for Children and Adolescents: A Consensus Statement, 106 PEDIATRICS 860 (2000) (stating that children and adolescents lack necessary coverage for substance abuse treatment).

(11.) Lewis, supra note 7, at 12.

(12.) House Hearings, supra note 7, at 3 (statement of Rep. Mica).

(13.) Janet Firshein, PBS Online, How Managed Care Is Affecting Addiction Treatment, at http://www.pbs.org/wnet/closetohome/treatment/html/managed.html (last visited Feb. 27, 2002).

(14.) Dissatisfaction with HMO care appears to be a motivating factor behind current legislative reform efforts. For example, House Speaker Dennis Hastert has criticized HMOs for putting health care decisions into the hands of "bureaucrats" rather than doctors and patients. J. Dennis Hastert, Finally, Help for HMO Patients, CHI. SUN-TIMES, Aug. 6, 2001, at 31.

(15.) 147 CONG. REC. S2733 (daily ed. Mar. 22, 2001) (statement of Sen. Wellstone) (citing a 1998 Hay Group report).

(16.) Firshein, supra note 13 (citing a 1996 Village Voice article).

(17.) Id.; see also Managed Care Provider Magellan Terminates Caron Foundation Contract, CNET NEWS.COM, Nov. 5, 2001, at http://news.cnet.com/investor/news/newsitem/0-9900-1028-7786092-0.html (discussing a managed care company's practice of providing an average of three to five days of care).

(18.) Firshein, supra note 13 (citing University of Pennsylvania studies).

(19.) See infra Subsection II.E.3.

(20.) See House Hearings, supra note 7, at 35 (statement of Rep. Mica).

(21.) See id. (statement of Rep. Ramstad).

(22.) Letter from Barry R. McCaffrey, Director, Office of National Drag Control Policy, to Representative Patsy T. Mink (Oct. 20, 1999), reprinted in House Hearings, supra note 7, at 9.

(23.) In this Note, I focus on the treatment of addiction to alcohol and illegal drugs but not on addiction to nicotine. There are several reasons for this focus: First, it is consistent with the approach taken by most other literature assessing treatment access; second, nicotine addiction has a very different, though also very serious, set of social and economic consequences; third, treatment for nicotine addiction varies significantly from treatment for other forms of addiction, and therefore efficacy and cost-effectiveness assessments may be quite different. Nonetheless, one could easily make an argument, along the lines of the argument presented in this Note, in favor of requiring insurance coverage of smoking cessation programs. Smoking kills over 430,000 Americans each year, but most health plans do not cover smoking cessation programs. SCHNEIDER INST. FOR HEALTH POLICY, BRANDEIS UNIV., SUBSTANCE ABUSE: THE NATION'S NUMBER ONE HEALTH PROBLEM 6, 114 (2001).

(24.) John N. Chappel, Attitudes Toward the Treatment of Substance Abusers, in SUBSTANCE ABUSE 983, 984 (Joyce H. Lowinson et al. eds., 2d ed. 1992) (citing a statement of Benjamin Rush, the "father of American psychiatry"); see also DRUGS AND DRUG POLICY IN AMERICA 224 (Steven R. Belenko ed., 2000) (excerpting an 1894 statement of Dr. Paul Sollier). In the early 1920s, the short-lived narcotic clinic movement pushed for expansion of access to medical treatment for poor addicts. See DRUGS AND DRUG POLICY IN AMERICA, supra, at 226 (excerpting a statement of Dr. Willis P. Butler).

(25.) House Hearings, supra note 7, at 28 (statement of Rep. Ramstad).

(26.) J. Madeleine Nash, Addicted: Why Do People Get Hooked? Mounting Evidence Points to a Powerful Brain Chemical Called Dopamine, TIME, May 5, 1997, at 68, 70 (quoting Dr. Nora Volkow of the Brookhaven National Laboratory).

(27.) Matthew Antinosi, Note, Respect for the Law Is No Excuse: Drug Addiction History & Public Safety Officer Qualifications ... Are Public Employers Breaking the Law?, 60 OHIO ST. L.J. 711, 716-18 (1999) (reviewing the scientific literature, including reports by the National Institutes of Health).

(28.) Eric J. Nestler & George K. Aghajanian, Molecular and Cellular Basis of Addiction, 278 SCIENCE 58, 58 (1997).

(29.) Id.; see also Eliot L. Gardner, Brain Reward Mechanisms, in SUBSTANCE ABUSE, supra note 24, at 70, 86-88 (describing changes to neural reward pathways caused by repeated exposure to addictive substances).

(30.) Alan I. Leshner, Addiction Is a Brain Disease, and It Matters, 278 SCIENCE 45 (1997).

(31.) Alcoholism: Researchers Find Gene Mutation That Protects Against Addiction, GENE THERAPY WKLY., Jan. 10, 2002, at 9; Robert M. Anthenelli & Marc A. Schuckit, Genetics, in SUBSTANCE ABUSE, supra note 24, at 39, 39-40 (discussing studies comparing alcoholism rates in identical and fraternal twins that provide strong evidence of a genetic link); Fleming et al., supra note 6, at 9.

(32.) Melissa Bush et al., Substance Abuse and Family Dynamics, in SOURCE BOOK OF SUBSTANCE ABUSE AND ADDICTION, supra note 6, at 57, 63 (noting that among current substance users, "physical abuse occurred in 55% of the households in which they were raised").

(33.) Id. (noting that 83% of substance users have a history of substance abuse in their households as children).

(34.) HENRICK HARWOOD ET AL., NAT'L INST. ON DRUG ABUSE & NAT'L INST. ON ALCOHOL ABUSE & ALCOHOLISM, THE ECONOMIC COSTS OF ALCOHOL AND DRUG ABUSE IN THE UNITED STATES 1992, at 2-2 (1998).

(35.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 16.

(36.) House Hearings, supra note 7, at 3 (statement of Rep. Mica); see also Office of Nat'l Drug Control Policy, Statement on Parity for Substance Abuse Treatment 1 (Mar. 5, 1999), reprinted in House Hearings, supra note 7, at 10, 10.

(37.) P'ship for Recovery, The Impact of Treatment on Substance Abuse, at http://www.partnershipforrecovery.org/fctinfo2.htm (last visited Feb. 27, 2002) (citing a 1993 study by the Institute for Health Policy at Brandeis University).

(38.) P'ship for Recovery, Addiction Kills ..., at http://www.partnershipforrecovery.org/fctinfol.htm (last visited Feb. 27, 2002).

(39.) Am. Soc'y of Addiction Med., Public Policy of ASAM: Addiction Medicine and Health Insurance Reform (Apr. 1, 1992), at http://www.asam.org/ppol/Addiction%20Medicine%20and%20 Health%20Insurance%20Reform.htm.

(40.) Id.

(41.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 45 ("More than one-third of all AIDS deaths in the United States have occurred among injecting drug users and their sexual partners.").

(42.) HARWOOD ET AL., supra note 34, at 5-9.

(43.) See House Hearings, supra note 7, at 55 (statement of R. Michael Conley, Chairman of the Board of Trustees, Hazelden Foundation, on behalf of the Partnership for Recovery).

(44.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 58.

(45.) Id. at 52.

(46.) HARWOOD ET AL., supra note 34, at 5-6.

(47.) Donald W. Goodwin, Alcohol: Clinical Aspects, in SUBSTANCE ABUSE, supra note 24, at 144, 147.

(48.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 6, 54; see also HARWOOD ET AL., supra note 34, at 1-3 (stating that in 1992, 107,400 people died due to alcohol abuse and 25,500 died due to drug abuse).

(49.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 50.

(50.) Patti Juliana & Carolyn Goodman, Children of Substance Abusing Parents, in SUBSTANCE ABUSE, supra note 24, at 808, 808.

(51.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 62, 65.

(52.) Id. at 66 (noting that most cases of sexual abuse involve alcohol); Charles Winick, Substances of Use and Abuse and Sexual Behavior, in SUBSTANCE ABUSE, supra note 24, at 722, 726 (stating that up to half of all rapes are linked to alcohol use).

(53.) Bush et al., supra note 32, at 65; see also id. at 57, 63 (describing the cycle of family violence, neglect, and future substance abuse engendered by untreated substance abuse).

(54.) Hank Resnik et al., Child Welfare and Substance Abuse: Premises, Programs, and Policies, in SUBSTANCE ABUSE, FAMILY VIOLENCE, AND CHILD WELFARE 96, 99 (Robert L. Hampton et al. eds., 1998) (citing findings of the National Center on Child Abuse and Neglect).

(55.) Id. at 101 (reviewing a range of studies).

(56.) Brenda Jones Harden, Building Bridges for Children: Addressing the Consequences of Exposure to Drugs and to the Child Welfare System, in SUBSTANCE ABUSE, FAMILY VIOLENCE, AND CHILD WELFARE, supra note 54, at 18, 18-27.

(57.) Resnik et al., supra note 54, at 99.

(58.) Goodwin, supra note 47, at 146.

(59.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 64.

(60.) P'ship for Recovery, Workplace Surveys Show Benefits, at http://www.partnershipforrecovery.org/newsr3.htm (last visited Feb. 27, 2002) (citing a 1996 BNA study).

(61.) Mary Bernstein & John Mahoney, Management Perspectives on Alcoholism: The Employer's Stake in Alcoholism Treatment, 4 OCCUPATIONAL MED. 223, 223 (1989) (reporting that 40% of industrial fatalities and 47% of injuries are linked to alcohol).

(62.) See SUBSTANCE ABUSE & MENTAL HEALTH SERVS. ADMIN., SUBSTANCE ABUSE AND MENTAL HEALTH STATISTICS SOURCEBOOK 3 (Beatrice A. Rouse ed., 1995).

(63.) See Fred D. Hafer, PBS Online, The High Cost of Doing Nothing, at http://www.pbs.org/wnet/closetohome/policy/html/hafer.html (last visited Feb. 27, 2002); see also SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 45 (noting that substance abusers are more likely to skip work and to switch jobs often).

(64.) These costs are further explored in the discussion of cost-effectiveness of treatment. See infra Section II.D; see also HARWOOD ET AL., supra note 34, at 1-2 to-6 (stating that addiction imposes societal costs through increased health expenditures, premature death, decreased productivity, car crashes, crime, and increased entitlement spending).

(65.) House Hearings, supra note 7, at 70 (statement of Kenny Hall, Addiction Specialist, Kaiser Permanente).

(66.) HARWOOD ET AL., supra note 34, at 4-31 to-33.

(67.) Id. at 1-3. This figure is in 1992 dollars.

(68.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 18.

(69.) HARWOOD ET AL., supra note 34, at 1-1. This figure is also in 1992 dollars. Note that inflation and population growth have a substantial effect on cost projections from year to year. The authors of the NIH study estimated that substance abuse costs had increased 12.5% between 1992 and 1995 alone. Id. at 1-9.

(70.) See Janet Firshein, PBS Online, Introduction, at http://www.pbs.org/wnet/closetohome/treatment/index.html (last visited Feb. 27, 2002) (citing a study by the National Academy of Sciences).

(71.) Id. (quoting Dr. Patricia Owen of the Hazelden Institute).

(72.) See House Hearings, supra note 7, at 54 (statement of R. Michael Conley, Chairman of the Board of Trustees, Hazelden Foundation, on behalf of the Partnership for Recovery).

(73.) Substance Abuse: The Science of Addiction and Options for Treatment: Hearing Before the S. Comm. on Labor and Human Resources, 105th Cong. 62 (1998) (statement of John Saylor, Manager of Employee Assistance Programs, AMR Corp.). The cost of complete treatment averaged only $5000-$6000 per patient, the majority of whom received twenty-one-day inpatient care. Id. Similar success rates were reported by Mobil Corp. See Rimrock Found., A Corporate Response to Alcohol and Drug Addiction, at http://www.rimrock.org/html/corporate.htm (last visited Feb. 27, 2002) (citing the Mobil EAP's 70-85% success rate in avoiding relapse).

(74.) Leshner, supra note 30, at 46.

(75.) Id.; see also Substance Abuse: The Science of Addiction and Options for Treatment: Hearing Before the S. Comm. on Labor and Human Resources, 105th Cong. 34 (1998) (statement of Dr. Robert Morse, Professor of Psychiatry, Mayo Medical School).

(76.) Letter from Michael Conley, Chairman of the Board of Trustees, Hazelden Foundation, to Representative John L. Mica (Nov. 12, 1999), reprinted in House Hearings, supra note 7, at 97, 97.

(77.) See David McMaster, New Direction Needed in the Treatment of Alcohol Addiction, WIS. ST. J., Aug. 13, 2000, at 2B.

(78.) Shift in Congress on Drug Policy Reform, supra note 1 (citing a Harvard School of Public Health study).

(79.) Office of Nat'l Drug Control Policy, supra note 36, at 2, reprinted in House Hearings, supra note 7, at 11.

(80.) Roland Sturm et al., How Expensive Are Unlimited Substance Abuse Benefits Under Managed Care?, 26 J. BEHAV. HEALTH SERVICES & RES. 203, 206 (1999) (assessing the costs of broad parity requirements to managed care plans).

(81.) House Hearings, supra note 7, at 64 (statement of Roland Sturm, Senior Economist, RAND Corporation).

(82.) The reason the $10,000 cap saves so little is that only 1.3% of patients receiving treatment for addiction in the study had total treatment costs above $10,000; more than half were treated for $1000 or less. Sturm et al., supra note 80, at 207.

(83.) Stephen P. Melek & Bruce S. Pyenson, Milliman & Robertson, Inc., Premium Estimates for Substance Abuse Parity Provisions for Commercial Health Insurance Products (1997), http://www.health.org/workplace/insur/1.htm.

(84.) MERRILE SING ET AL., U.S. DEP'T OF HEALTH & HUMAN SERVS., THE COSTS AND EFFECTS OF PARITY FOR ]MENTAL HEALTH AND SUBSTANCE ABUSE INSURANCE BENEFITS (1998), http://www.mentalhealth.org/publications/allpubs /Mc99-80/prtyfnix.asp.

(85.) See P'ship for Recovery, supra note 60.

(86.) Diana Chapman Walsh et al., Treating the Employed Alcoholic, 16 ALCOHOL HEALTH & RES. WORLD 140, 147 (1992).

(87.) House Hearings, supra note 7, at 56 (statement of R. Michael Conley, Chairman of the Board of Trustees, Hazelden Foundation, on behalf of the Partnership for Recovery).

(88.) Bernstein & Mahoney, supra note 61, at 228.

(89.) Rimrock Found., The Bottom Line, at http://www.rimrock.org/html/info/bottomline.htm (last visited Feb. 27, 2002).

(90.) House Hearings, supra note 7, at 56 (statement of R. Michael Conley, Chairman of the Board of Trustees, Hazelden Foundation, on behalf of the Partnership for Recovery).

(91.) Id.

(92.) Office of Nat'l Drug Control Policy, supra note 36, at 3-4, reprinted in House Hearings, supra note 7, at 12-13.

(93.) Id.

(94.) See P'ship for Recovery, supra note 37 (citing Minnesota and Arizona studies). The Minnesota study showed a drastic reduction both in overall arrests and in DWI arrests. Id.

(95.) See, e.g., Janet Firshein, PBS Online, Our Current Policy, at http://www.pbs.org/wnet/closetohome/policy/html/current.html (last visited Feb. 27, 2002) (citing a RAND Corporation study finding treatment to be several times more effective in decreasing cocaine use than imprisonment including mandatory minimum sentences).

(96.) Robert L. Hubbard, Evaluation and Treatment Outcome, in SUBSTANCE ABUSE, supra note 24, at 596, 603.

(97.) Physician Leadership on Nat'l Drug Policy, Position Paper on Drug Policy 4 (n.d.), reprinted in House Hearings, supra note 7, at 106, 109.

(98.) Janet Firshein, PBS online, Does Treatment Work?, at http://www.pbs.org/wnetclosetohome/policy/html/treatwork.html (last visited Feb. 27, 2002).

(99.) See Office of Nat'l Drug Control Policy, supra note 36, at 2, reprinted in House Hearings, supra note 7, at 11; Physician Leadership on Nat'l Drug Policy, supra note 97, at 1, reprinted in House Hearings, supra note 7, at 106.

(100.) See House Hearings, supra note 7, at 66 (statement of Roland Sturm, Senior Economist, RAND Corporation). Mental health treatment is already subject to parity requirements. See infra note 205 and accompanying text.

(101.) See House Hearings, supra note 7, at 36 (statement of Rep. Barr) (stating that "a lot of people choose to use" drugs and alcohol); cf. id. at 33 (statement of Rep. Hutchison) (raising this argument as a possible objection to insurance parity, although not clearly endorsing it).

(102.) See supra Section II.A.

(103.) Cf. Richard H. Pildes & Cass R. Sunstein, Reinventing the Regulatory State, 62 U. CHI. L. REV. 1, 50-51, 57-72 (1995) (noting that voluntariness of a risk may legitimately be understood to discount its magnitude in risk analysis). But cf Frank B. Cross, The Subtle Vices Behind Environmental Values, 8 DUKE ENVTL. L. & POL'Y F. 151,160-69 (1997) (critiquing the value of the voluntary/involuntary distinction in risk analysis).

(104.) See HARWOOD ET AL., supra note 34, at 7-1 (stating that the nonabusing population bears at least fifty-six percent of the cost of drug abuse and fifty-five percent of the cost of alcoholism).

(105.) As Philip Morris publicly declared in late 1999: "`There is an overwhelming medical and scientific consensus that cigarette smoking causes lung cancer, heart disease, emphysema, and other serious diseases in smokers.'" Philip Morris Admits Tobacco Linked to Cancer, AGENCE FRANCE PRESSE, Oct. 13, 1999, LEXIS, Nexis Library, Agence France Presse File (quoting the company website).

(106.) See, e.g., Heart Association Dietary Guidelines, HARV. HEART LETTER, Feb. 1, 1997, LEXIS, Nexis Library, Harvard Heart Letter File.

(107.) SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 7.

(108.) See supra Section II.A.

(109.) See Luisa Dillner, When Life Is Just Poured Away ..., GUARDIAN (London), Sept. 17, 1996, at T4 (quoting a statement of Moira Plant of Edinburgh University that alcoholics are stigmatized as having "weak personalities").

(110.) See supra Section II.A.

(111.) Cf. House Hearings, supra note 7, at 46 (statement of Rep. Barr). Representative Bart stated:

   [Removing] the moral component[] completely may not be the best way to cast
   this argument, because we do want to send a message to people that alcohol
   is bad and the use of drags is bad, and not to say, well, it is OK, and we
   can't have any stigma at all attached to it.

Id. Note that Representative Barr's argument does not rely on illegality per se, but simply on the validity of attaching stigma to behaviors that society wants to condemn; he applies the argument to alcohol use as well.

(112.) According to Dr. Alan Leshner, director of the National Institute on Drug Abuse, "70 million adult Americans have used an illegal drug at some point in their lives." Alan I. Leshner, Take the Stigma Out of Addiction, FAM. PRAC. NEWS, Aug. 15, 2000, at 9, LEXIS, Nexis Library, Family Practice News File.

(113.) See, e.g., AFTER PROHIBITION: AN ADULT APPROACH TO DRUG POLICIES IN THE 21ST CENTURY (Timothy Lynch ed., 2000) (collecting essays debating issues related to drug legalization); STEVEN B. DUKE & ALBERT C. GROSS, AMERICA'S LONGEST WAR: RETHINKING OUR TRAGIC CRUSADE AGAINST DRUGS (1993); JAMES P. GRAY, WHY OUR DRUG LAWS HAVE FAILED AND WHAT WE CAN DO ABOUT IT: A JUDICIAL INDICTMENT OF THE WAR ON DRUGS (2001); HARM REDUCTION: PRAGMATIC STRATEGIES FOR MANAGING HIGH-RISK BEHAVIORS (G. Alan Marlatt ed., 1998) [hereinafter HARM REDUCTION] (collecting essays discussing "harm reduction," an alternative to current drug policy that would deemphasize criminal law enforcement in favor of efforts to alleviate the negative consequences of drug use).

(114.) See Firshein, supra note 95.

(115.) House Hearings, supra note 7, at 81 (statement of Peter J. Ferrara, General Counsel and Chief Economist, Americans for Tax Reform).

(116.) Id. at 82.

(117.) Id. at 81.

(118.) Id. at 85 (statement of Charles N. Kahn III, President, Health Insurance Association of America).

(119.) Mental Health Parity Act, 42 U.S.C.A. [section] 300gg-5 (West Supp. 2001).

(120.) See Lisa A. Hayden, Gender Discrimination Within the Reproductive Health Care System: Viagra v. Birth Control, 13 J.L. & HEALTH 171, 189 (1999) (reviewing state laws).

(121.) 42 U.S.C. [section] 2000e(k) (1994).

(122.) See Sylvia A. Law, Sex Discrimination and Insurance for Contraception, 73 WASH. L. REV. 363, 376-83 (1998).

(123.) HIAA's president conceded as much in his testimony, asking rhetorically, "[I]n isolation, who can argue against coverage for substance abuse?" House Hearings, supra note 7, at 86 (statement of Charles N. Kahn III, President, Health Insurance Association of America). Kahn did cite a study allegedly demonstrating that substance abuse treatment mandates would increase insurance premiums by an average of nine percent. Id. at 91. Subsequent committee debate challenged this finding. First, Representative Ramstad noted that that study, which was funded by HIAA itself, was inconsistent with the consensus of other studies. Id. at 101 (statement of Rep. Ramstad). Second, initial premium increases may have resulted from the effect of pent-up demand. In Kaiser Permanente's experience, many people who had never had access to treatment sought it in the program's first year. Once that demand leveled out, cost savings became more apparent. Id. at 95 (colloquy between Rep. Mica and Kenny Hall, Addiction Specialist, Kaiser Permanente).

(124.) Id. at 91 (statement of Charles N. Kahn III, President, Health Insurance Association of America).

(125.) Survey Reveals Bias Against Recovering Alcoholics and Addicts, ALCOHOL & DRUG ABUSE WKLY., Dec. 20, 1999, at 3 (demonstrating public unwillingness to hire recovering addicts).

(126.) 147 CONG. REC. S2732 (daily ed. Mar. 22, 2001) (statement of Sen. Wellstone).

(127.) Leshner, supra note 30, at 45.

(128.) Id.

(129.) Of course, if these costs translated into increased demand for treatment coverage among employers purchasing health care plans, insurers might well adjust the plans they offer in order to meet that demand. In doing so, they would be indirectly accounting for the range of costs addiction imposes on employers. But because employers themselves do not properly account for the costs of addiction and the savings brought by treatment, see infra notes 133-136 and accompanying text, this demand-driven market pressure is absent, and insurers have no reason to internalize these costs. Furthermore, insurers never have an economic incentive to factor in costs imposed on third parties--that is, on society as a whole.

(130.) See, e.g., Bryan Ford, The Uncertain Case for Market Pricing of Health Insurance, 74 B.U.L. REV. 109, 133, 140 (1994) (explaining adverse selection).

(131.) Adverse selection is likely to be a lesser problem in the employer-provided group health plan market than in the individual health insurance market, since the risk pool is averaged across all employees of that employer. See Daniel J.B. Mitchell, Perspective on Benefits: Humpty Dumpty Health Insurance, L.A. TIMES, Dec. 8, 1999, at B9 (noting that the employer-provided health insurance system is designed to avoid adverse selection). Nonetheless, adverse selection is broadly considered a serious problem in health insurance, even though most insurance is provided by employers. See Richard G. Frank et al., Solutions for Adverse Selection in Behavioral Health Care, HEALTH CARE FINANCING REV., Mar. 22, 1996, at 109 (noting that adverse selection is a particular problem with regard to pricing coverage of mental health treatment). Possible explanations include the following: Some employers offer individual employees choices between particular plans; for small or mid-sized employers, risk-pooling may not be sufficient; and for employers who do not self-insure, costs of medical plans are affected by the adverse selection that occurs among individual purchasers who are in the market for the same plans.

(132.) Cf. Toni Vranjes, Doctor Proposes Major Changes in Health Insurance System, MED. INDUSTRY TODAY, Jan. 19, 1999, LEXIS, Nexis Library, Medical Industry Today File (reporting Dr. David Levy's stance that standardizing coverage is the best solution to adverse selection).

(133.) See supra Section II.D.

(134.) See Kenneth Silverman & Elias Robles, Employment as a Drug Abuse Treatment Intervention: A Behavioral Economic Analysis, in THE ECONOMIC ANALYSIS OF SUBSTANCE USE AND ABUSE 279, 279-81 (Frank J. Chaloupka et al. eds., 1999) (reviewing these studies).

(135.) See Firshein, supra note 8.

(136.) Id.

(137.) 42 U.S.C. [subsection] 12101-12213 (1994).

(138.) Kenneth Vanko, Note, In Search of Common Ground: Leveling the Playing Field for Chemically Dependent Workers Under the Americans with Disabilities Act of 1990, 1996 U. ILL. L. REV. 1257, 1261.

(139.) See, e.g., Hoffman v. MCI Worldcom Communications, Inc., 178 F. Supp. 2d 152, 155 (D. Conn. 2001) (stating that "drug and alcohol addiction satisfies the disability prong" of the ADA); Thompson v. Borg-Warner Protective Servs. Corp., No. C-94-4015, 1996 U.S. Dist. LEXIS 4781, at * 22 (N.D. Cal. Mar. 12, 1996) ("Drug addiction and alcoholism are protected disabilities under the ADA.").

(140.) 28 C.F.R. [section] 41.31(b)(1) (2001).

(141.) 42 U.S.C. [section] 12114; see also Hoffman, 178 F. Supp. 2d at 155 (rejecting an ADA claim because the plaintiff was a drug user at the time of dismissal). One scholar has suggested that courts often wrongly extend this exception to current users of alcohol. Amy Hennen, Protecting Addicts in the Employment Arena: Charting a Course Toward Tolerance, 15 LAW & INEQ. 157, 172-73 (1997) (arguing also that the current-drag-use exception should be abolished).

(142.) U.S. Equal Employment Opportunity Comm'n, Interim Enforcement Guidance on the Application of the Americans with Disabilities Act to Disability-Based Distinctions in Employer-Provided Health Insurance, EEOC Notice 915.002, at 2-3 (June 8, 1993) [hereinafter EEOC Guidance on Health Insurance]. The EEOC's implementation regulations are found at 29 C.F.R. pt. 1630 (2001). Section 501(c) of the ADA, 42 U.S.C. [section] 12201(c), provides some guidance as to which disability-based distinctions in health insurance are acceptable; these are discussed further in Section III.B.

(143.) EEOC Guidance on Health Insurance, supra note 142, at 3.

(144.) Id.

(145.) Lawsuits against government employers or government health plans might also proceed under Title II, which covers government services. Cf. Williams v. Wasserman, 937 F. Supp. 524 (D. Md. 1996) (allowing a Title II lawsuit against the state for failing to place mental patients in community-based care facilities). The standards involved in Title II litigation are similar to those in Title I cases; I therefore do not discuss Title II in detail here.

(146.) See David Manoogian, With Suits Mounting, Courts Face the Question of Whether a Managed Care Organization Can Be an Employer Under the Americans with Disabilities Act, NAT'L L.J., Mar. 17, 1997, at B6 (stating that managed care organizations and other health insurers can be sued under Title I because they are "`organization[s] providing fringe benefits to an employee of a covered entity'" (quoting 42 U.S.C.A. [section] 12112(b)(2))); see also Carparts Distrib. Ctr., Inc. v. Auto. Wholesaler's Ass'n of New England, 37 F.3d 12 (1st Cir. 1994) (holding that managed care organizations are "employers" under the ADA). Manoogian notes that in the case of a bankrupt employer, a plaintiff might look directly to the insurance company to satisfy a judgment. Generally, however, the employer would probably be ultimately liable for discrimination on the part of the insurance companies with which it contracts. As the EEOC states, "[A]n employer will be liable for any discrimination resulting from a contract or agreement with an insurance company ... to provide or administer a health insurance plan on behalf of its employees." EEOC Guidance on Health Insurance, supra note 142, at 2-3.

(147.) Manoogian, supra note 146.

(148.) 42 U.S.C. [section] 12112.

(149.) Id. [section] 12112(b)(5)(A).

(150.) Id. See generally U.S. Equal Employment Opportunity Comm'n, Enforcement Guidance: Reasonable Accommodation and Undue Hardship Under the Americans with Disabilities Act, EEOC Notice 915.002 (Mar. 1, 1999) [hereinafter EEOC Reasonable Accommodation Guidance].

(151.) 42 U.S.C. [section] 12181(7)(F).

(152.) Id. [section] 12201(c).

(153.) Id. [section] 12201(c)(2).

(154.) EEOC Guidance on Health Insurance, supra note 142, at 4.

(155.) 42 U.S.C. [section] 12201(c).

(156.) EEOC Guidance on Health Insurance, supra note 142, at 12.

(157.) Id.

(158.) Id. at 13.

(159.) Id. at 14.

(160.) See supra Section II.D.

(161.) The EEOC, faced with a case overload and needing to prioritize its enforcement actions, has placed considerable emphasis on the eradication of AIDS discrimination, including discrimination in health insurance. See Manoogian, supra note 146:

(162.) Pub. L. No. 104-204, tit. VII, 110 Stat. 2944 (1996) (codified as amended at 42 U.S.C.A. [section] 300gg-5 (West Supp. 2001)).

(163.) EEOC Guidance on Health Insurance, supra note 142, at 8-9. Note that the EEOC's interpretive guidances, "`while not controlling on the courts by reason of their authority, do constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance.'" Meritor Sav. Bank v. Vinson, 477 U.S. 57, 65 (1986) (quoting Gen. Elec. Co. v. Gilbert, 429 U.S. 125, 141-42 (1976) (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944))).

(164.) 37 F.3d 12 (1st Cir. 1994).

(165.) 42 U.S.C. [section] 2000e(b) (1994).

(166.) Carparts, 37 F.3d at 21.

(167.) Id. at 17 (citing a Title VII case, Spirt v. Teachers Insurance & Annuity Ass'n, 691 F.2d 1054, 1063 (2d Cir. 1982)).

(168.) Barone v. Hackett, 602 F. Supp. 481,483 (D.R.I. 1984).

(169.) Carparts, 37 F.3d at 18 (quoting, but rejecting, the district court's criterion).

(170.) Carparts Distrib. Ctr., Inc. v. Auto. Wholesaler's Ass'n of New England, 987 F. Supp. 77, 78 (D.N.H. 1997).

(171.) 99 F.3d 181 (6th Cir. 1996), vacated, 121 F.3d 1006 (6th Cir. 1997) (en banc). For a discussion of the content of the vacated decision, see Kristin Kay Romero, Defining Discrimination in Doe v. Mutual of Omaha Insurance Co.: Determining if a Health Insurance Policy's AIDS Benefit Cap Violates the ADA, 9 GEO. MASON L. REV. 179 (2000).

(172.) Parker, 99 F.3d at 192.

(173.) Parker, 121 F.3d 1006.

(174.) Section 501(c) provides an exception to the ADA, and therefore only applies if an ADA violation is found.

(175.) Parker, 121 F.3d at 1015-17.

(176.) 96 F.3d 1039 (7th Cir. 1997).

(177.) Id. at 1044. That Act was passed the day before the release of the CNA Insurance opinion, but covered only health insurance, not long-term disability assistance. See Parker, 121 F.3d at 1017-18.

(178.) Ford v. Schering-Plough Corp., 145 F.3d 601 (3d Cir. 1998) (applying analysis similar to Parker's to Title I, and also holding that Title III did not apply because Congress intended Title I to be the exclusive remedy for employment discrimination); Rogers v. Dep't of Health & Envtl. Control, 174 F.3d 431 (4th Cir. 1999) (involving an ADA Title II claim related to government health benefits).

(179.) 121 F.3d at 1020-22 (Merritt, J., dissenting).

(180.) Id. at 1020 (Martin, C.J., dissenting) (quoting Carparts Distrib. Ctr., Inc. v. Auto. Wholesaler's Ass'n of New England, 37 F.3d 12, 19 (1st Cir. 1994) (quoting H.R. REP. NO. 101485, at 99 (1990), reprinted in 1990 U.S.C.C.A.N. 303,382)).

(181.) 179 F.3d 557 (7th Cir. 1999).

(182.) Id. at 559-60. In my opinion, this analogy is flawed. ADA analysis always turns on the reasonableness of a particular requirement and the burden it imposes--that is, a balancing of competing interests. A shoe store could likely show easily that the hypothetical single-shoe requirement would not be a "reasonable modification" or that it would "fundamentally alter" the nature of its business, while an insurance company--at least in the context of treatments that are proven to be cost-effective--likely could not make such a showing. For a discussion of reasonable accommodation requirements, see infra Section III.D. Mutual of Omaha has occasioned some scholarly criticism. See, e.g., Mary Crossley, Becoming Visible: The ADA's Impact on Health Care for Persons with Disabilities, 52 ALA. L. REV. 51, 84 (2000); Romero, supra note 171, at 191.

(183.) Mutual of Omaha, 179 F.3d at 560. The Mutual of Omaha court further held that an application of the ADA to the content of insurance policies would be inconsistent with another federal statute, the McCarran-Ferguson Act. Id. at 563. That Act directs courts not to interpret federal statutes in a way that would conflict with state regulation of the insurance industry unless the federal statute "specifically relates to the business of insurance." 15 U.S.C. [section] 1012(b) (1994). Because a section 501(c) analysis requires an inquiry into the actuarial soundness of the insurance policy, the court reasoned that it would substantially interfere with the work of state insurance commissioners and would thereby violate the McCarran-Ferguson Act.

In my view, inconsistency with the McCarran-Ferguson Act can only be found through extremely circular reasoning, since the very portion of the ADA that would require potential interference with state regulators--section 501(c)--clearly does "specifically relate to the business of insurance." The court gets around this objection by saying that the part of the Act that requires interpretation is section 302(a), not section 501(c), and that section 302(a) does not directly mention insurance. This is simply question-begging, however, since section 302(a) also does not require any interference with state administrators. The court separates the two portions of the Act for the purpose of determining whether the Act "specifically relates to the business of insurance," but conflates them for the purpose of determining whether they interfere with state administration of the insurance industry. Mutual of Omaha, 179 F.3d at 564.

(184.) 198 F.3d 28 (2d Cir. 1999).

(185.) Id. at 33 (quoting 42 U.S.C. [section] 12101(b) (1994)). Furthermore, the court held that the legislative histories and purposes of the ADA and the McCarran-Ferguson Act demonstrated no conflict between the two. McCarran-Ferguson was intended to prevent "inadvertent" federal intrusions on the field of insurance regulation; multiple specific references to insurance in the ADA demonstrate that the intrusion was not inadvertent but was clearly intended by Congress. Id. at 34-35.

(186.) Ford v. Schering-Plough Corp., 145 F.3d 601,608 (3d Cir. 1998).

(187.) Krauel v. Iowa Methodist Med. Ctr., 95 F.3d 674 (8th Cir. 1996).

(188.) EEOC v. Staten Island Sav. Bank, 207 F.3d 144, 148-49 (2d Cir. 2000); Weyer v. Twentieth Century Fox Film Corp., 198 F.3d 1104, 1116-18 (9th Cir. 2000); Kimber v. Thiokol Corp., 196 F.3d 1092, 1101-02 (10th Cir. 1999); Lewis v. Kmart Corp., 180 F.3d 166, 170 (4th Cir. 1999). The Second Circuit's decision in Staten Island Savings Bank, though it applied only to Title I, suggests that notwithstanding the encouraging precedent of Pallozzi, courts in the Second Circuit will not likely uphold plaintiffs' claims under Title III either. Pallozzi never reached the question of whether the insurance plan actually was discriminatory; under the rationale of Staten Island Savings Bank, it probably would not be. Staten Island Sav. Bank, 207 F.3d 144.

(189.) Linda H. Krieger, Foreword--Backlash Against the ADA: Interdisciplinary Perspectives and Implications for Social Justice Strategies, 21 BERKELEY J. EMP. & LAB. L. 1, 3 (2000).

(190.) Judging by the absence of any mention in the decisions of the various courts, reasonable accommodation does not seem to have been raised as an issue in any of the insurance cap cases.

(191.) 42 U.S.C. [section] 12112(b)(3) (1994).

(192.) Id. [section] 12112(b)(5)(A).

(193.) Id. [section] 12182(b)(2)(A)(ii).

(194.) See Keith Nelson, Comment, Legislative and Judicial Solutions for Mental Health Parity: S. 543, Reasonable Accommodation, and an Individualized Remedy Under Title I of the ADA, 51 AM. U. L. REV. 91 (2001) (proposing strategies for reasonable accommodation lawsuits in mental health cap cases).

(195.) Wendy E. Parmet et al., Accommodating Vulnerabilities to Environmental Tobacco Smoke: A Prism for Understanding the ADA, 12 J.L. & HEALTH 1, 20 (1997-1998).

(196.) Most existing ADA cases involving addiction have focused on what reasonable accommodations employers are required to provide to current or recovering addicts. Courts have held that employers are required to provide unpaid leave and reasonable adjustments in working hours in order to allow employees to take part in outpatient or inpatient treatment programs. See EEOC Reasonable Accommodation Guidance, supra note 150, at 26 (stating that reasonable accommodation includes unpaid leave for substance abuse treatment). Courts have usually held that employers may, in fact, make an employee's participation in treatment mandatory as a condition of continued employment--they may offer a firm choice between treatment or dismissal. See, e.g., Rodgers v. Lehman, 869 F.2d 253 (4th Cir. 1989). The issue of "forced accommodation" is complex and has occasioned some criticism from legal scholars. See, e.g., Tim Edwards, Constitutional Limits on an Employer's Right To Dictate the Terms of an Addict's Recovery Under the ADA: Some Sobering Concerns, 44 WAYNE L. REV. 1679 (1999). I do not focus on that issue here, however, since it is not germane to the insurance discrimination issue.

(197.) This, of course, presents a problem for any parity strategy that relies on case-by-case litigation for its enforcement, If parity were required in all cases, however--under a clear statutory requirement with which employers' insurance companies would rather comply than lose litigation--an addicted employee would be able to seek treatment without notifying her employer, and retain her privacy to the extent that all medical records remain confidential.

(198.) 42 U.S.C. [section] 12112(b)(3).

(199.) See supra notes 138-140 and accompanying text.

(200.) See supra note 6 (discussing the distinction between substance abuse and addiction).

(201.) 42 U.S.C. [section] 12102(2).

(202.) Albertson's, Inc. v. Kirkinghurg, 527 U.S. 555, 566 (1999).

(203.) Kevin Harris, Disability, American Style, CONVERIUM, Oct. 1, 1999, http://www.converium.com/.

(204.) Bd. of Trs. of the Univ. of Ala. v. Garrett, 531 U.S. 356 (2001) (holding that the Eleventh Amendment bars damage awards in ADA suits against unconsenting state governments).

(205.) Mental Health Parity Act, Pub. L. No. 104-204, tit. VII, 110 Stat. 2944 (1996) (codified as amended at 42 U.S.C.A. [section] 300gg-5 (West Supp. 2001)).

(206.) S. 595, 107th Cong. (2001). To avoid confusion, I refer to both the House and the Senate versions of this bill as the Substance Abuse Treatment Parity Act.

(207.) H.R. 1194, 107th Cong. (2001).

(208.) The bill's status is tracked at Partnership for Recovery, at http://www.capwiz.com/pfr/issues/bills/ (last visited Apr. 10, 2002). By the end of the 105th Congress, the 1997 bill had ninety-one cosponsors in the House, but never reached a floor vote. Lexis-Nexis Congressional Universe: Bill Tracking, at http://web.lexis-nexis.com/congcomp/ (last visited Apr. 10, 2002).

(209.) S. 595, sec. 2, [section] 2707(a).

(210.) H.R. 1194, sec. 3, [section] 9813(c)(2).

(211.) Mental Health and Substance Abuse Parity Amendments of 2001, H.R. 162, 107th Cong. (2001).

(212.) See Maria A. Morrison, Changing Perceptions of Mental Illness and the Emergence of Expansive Mental Health Parity Legislation, 45 S.D.L. REV. 8, 20 (2000) (discussing the original amendments); Nat'l Ass'n of Social Workers, Update on Mental Health Parity (Dec. 20, 2001), at http://www.naswdc.org/advocacy/updates/122001.htm (describing the current state of the bill and Senate legislation reanthorizing the Mental Health Parity Act but not applying it to substance abuse).

(213.) See Bill Summary & Status for the 107th Congress: H.R. 162, at http://thomas.loc.gov (last visited Mar. 13, 2002) (tracking the number of cosponsors).

(214.) In particular, any insurance parity legislation that is passed should include coverage for methadone maintenance treatment, a medication-based therapy that, although politically controversial, is by far the most effective treatment for heroin addiction. See Joyce H. Lowinson et al., Methadone Maintenance, in SUBSTANCE ABUSE, supra note 24, at 550, 556-57 (stating that methadone maintenance significantly reduces heroin use, decreasing crime and increasing productive employment of patients); Susan F. Tapert et al., Harm Reduction Strategies for Illicit Substance Use and Abuse, in HARM REDUCTION, supra note 113, at 145, 152, 154-60 (stating that methadone, though controversial, is effective in eliminating heroin use, reducing crime, increasing health, and reducing HIV infection). The version of the Substance Abuse Treatment Parity Act introduced in the House in 1999 appeared to exclude methadone by specifying that only "abstinence-based" services were covered, a term that was defined to encompass "non-narcotic medication-based therapy and appropriate transitional medication-based therapy." H.R. 1977, 106th Cong. [section] 3(a)(1)(A) (1999). Fortunately, the version of the Act introduced in the House in 2001 does not include the "abstinence-based" language, H.R. 1194, and the language was never in the Senate version, see Field Has Busy Legislative Agenda; Does Congress Have the Time?, ALCOHOLISM & DRUG ABUSE WKLY., Jan. 3, 2000, at 1, LEXIS, Nexis Library, Alcoholism & Drug Abuse Weekly File. In any event, the current version of the Substance Abuse Treatment Parity Act is fairly vague as to what treatments are covered, and the Mental Health and Substance Abuse Parity Amendments of 2001 are even more so. Further clarification may be necessary to avoid ambiguity in interpretation.

(215.) See, for example, the emphasis placed on the Substance Abuse Treatment Parity Act on the website of the Partnership for Recovery, a coalition of prestigious nonprofit treatment providers. P'ship for Recovery, at http://www.partnershipforrecovery.org (last visited Mar. 13, 2002).

(216.) E.g., Civil Rights Act of 1964 tit. VII, 42 U.S.C. [section] 2000(e)(b) (1994) (exempting employers with fewer than fifteen employees); Americans with Disabilities Act of 1990, 42 U.S.C. [section] 12111(5)(A) (same).

(217.) See House Hearings, supra note 7, at 34 (statement of Rep. Ramstad) (stating that the exemption was needed "to get the bill moving").

(218.) See supra Section II.D (citing payoff ratios of .up to ten to one). Although comprehensive studies focusing solely on small businesses are lacking, there is some evidence of cost savings among small businesses that have provided coverage of substance abuse treatment. For example, a study of hospitality industry employees, including employees of small businesses, showed reductions in injuries, tardiness, absenteeism, and errors following treatment. See House Hearings, supra note 7, at 57 (statement of R. Michael Conley, Chairman of the Board of Trustees, Hazelden Foundation, on behalf of the Partnership for Recovery); cf. Rimrock Found., Recovery from Alcohol and Drug Addiction: A Role for Small Business, at http://www.rimrock.org/html/role.htm (last visited Feb. 27, 2002); Rimrock Found., Addiction Treatment: A Small Business Investment with Unique Dividends, at http://www.rimrock.org/html/investment.htm (last visited Feb. 27, 2002).

(219.) Naturally, this raises the question of why treatment coverage should be made mandatory when it is in the businesses' self-interest; economic rationality ought to dictate that they provide coverage voluntarily. This issue is discussed above in Subsection II.E.3. Business decisions are not always purely rational; they may be skewed by social stigma or misinformation about treatment's benefits and effectiveness.

(220.) Obviously, this is why small businesses do not self-insure (as many large companies do) but instead buy into large insurance plans, if they provide insurance at all; it spreads the risk beyond the company itself. Still, a small business with one or two cancer patients may abruptly find its group premium rate hiked by the insurance company. This is a serious problem that may necessitate comprehensive health insurance reform, which is beyond the scope of this Note.

(221.) See, e.g., Melek & Pyenson, supra note 83.

(222.) Pub. L. No. 104-204, sec. 702, [section] 712(c)(2), 110 Stat. 2944, 2947 (1996) (codified at 42 U.S.C.A. [section] 300gg-5(c)(2) (West Supp. 2001)).

(223.) Mental Health and Substance Abuse Parity Amendments of 2001, H.R. 162, 107th Cong. (2001).

(224.) See supra Section II.D.

(225.) See Pamela L. Simmons, Solving the Nation's Drug Problem: Drug Courts Signal a Move Toward Therapeutic Jurisprudence, 35 GONZ. L. REV. 237, 251 (2000) (noting that even a modest 14.3% increase in treatment funding proposed by President Clinton in 1994 proved politically impossible, due to politicians' belief that treatment is perceived as soft on crime). Despite this inadequacy, governments actually bear the lion's share of the cost of treatment today, since private funding is even more lacking. Two-thirds of funding for addiction treatment today comes from federal, state, and local governments, while only fourteen percent is provided by private insurance. SCHNEIDER INST. FOR HEALTH POLICY, supra note 23, at 104.