Healthism and the Law of Employment Discrimination


For most of us, our health is a work-in-progress. A nagging voice in the back of our heads reminds us that we should shed a few pounds or give up smoking. Yet time and time again, many of us still reach for a second cupcake or light up a cigarette. But what if employers could use these seemingly personal choices related to our wellness to make decisions regarding whether to hire us, even if we never eat or smoke at work? Recently, employers around the country attempted to do just that.

In October 2012, Methodist Hospital System (“Methodist”) announced that starting in January 2013 it would no longer hire applicants who use nicotine.1 The new policy forbids hiring individuals who smoke or chew tobacco in an effort to establish the hospital as a “tobacco-free employer.”2 Methodist plans to enforce the ban with a urine test: Applicants who test positive for nicotine will lose their job offers, yet will have access to a free tobacco cessation program and may reapply once they have been nicotine-free for ninety days.3 In a similar move, Baylor Health Care System (“Baylor”) stopped hiring nicotine users in January 2012, also adding a test for the legal substance to its existing drug screening process.4 These new policies led to public outcry,5 given Methodist’s and Baylor’s statuses as among the largest employers in two greater metropolitan areas.6

Methodist and Baylor were not alone. In just the past two to three years, employers throughout the United States banned hiring nicotine users, including employers in Florida,7 Massachusetts,8 Michigan,9 Ohio,10 Pennsylvania,11 and Texas.12

Also in 2012, Citizens Medical Center (“CMC”), a county-run hospital, instituted a policy against hiring individuals with body mass indexes of thirty-five or over.13 Body mass index (“BMI”)—a medical heuristic—uses a person’s height and weight to approximate her amount of body fat.14 People with BMIs of thirty-five and above are considered “severely obese”15 and therefore at a heightened risk for weight-related health conditions such as high blood pressure and Type 2 diabetes.16 While initially defending its decision as an effort to promote a healthy image,17 CMC did away with the obesity ban in April 2012.18

Although the above employers adopted their policies recently, these bans are far from the first of their kind. Employers have been screening potential employees based on their perceived health for years.19 Over a decade ago, companies were already adopting hiring practices designed to regulate employee health.20 These policies prompted several states to pass legislation that prohibits employers from discriminating on the basis of legal lifestyle choices that do not directly affect performance in the workplace.21 While these types of statutes have gained significant traction at the state level, at present they have no federal counterpart.

This Article examines employment policies based on unhealthy conduct, asking if those practices constitute unjustifiable discrimination. Employers may opt to screen out applicants they consider unhealthy for a number of rational reasons: to cabin health-insurance costs, to increase workforce productivity, or to cultivate a particular business image. However, individuals considered unhealthy—particularly smokers and the obese—are also the objects of stigma. This Article queries whether this differentiation on the basis of health status constitutes legitimate grounds for employer decision-making, or whether the law ought to intervene on behalf of applicants and employees engaging in unhealthy conduct.

It ultimately concludes that while employers may have practically sound—and perhaps even socially desirable—reasons to screen for health in their employment decisions, policies like the ones described here have the potential to perpetuate existing health disparities, including those experienced by certain historically disadvantaged groups, mainly racial and ethnic minorities, people with disabilities, and the poor and near-poor. Ironically, these efforts at health promotion could ultimately lead to greater adverse health outcomes, by not only denying those populations access to wage work but also by impeding their ability to obtain health insurance and, as a result, access to needed health care.

This Article contributes to the practical and scholarly debates surrounding the regulation of employee health by recognizing valid concerns on both sides of this polarizing issue and proposing a middle legislative ground. While staunch privacy advocates disparage health-related employer screens as unduly intrusive into our private lives, proponents of employment at-will view them as just another useful metric upon which employers can base their choices. However, neither perspective captures the full complexity of the issues at play here. While these positions seem intractable, this Article proposes that well-drafted legislation can reconcile employer and employee concerns.

Further, health promotion, even if driven by cost concerns, could enhance the greater good. The Article thus explores nondiscriminatory ways in which employers could promote employee health, even in the wake of legislation. Encouraging workers to stop smoking or to maintain a healthy weight are laudable health goals and employers could—and perhaps should—play a significant role in their promotion. However, adverse employment actions, such as failing to hire or firing, based on health-related conduct that is not job-related and that occurs away from work are simply the wrong mechanisms to achieve this desirable end. Thus, the Article posits that employers may regulate employee health in ways that do not jeopardize a qualified individual’s ability to find or keep a job, thereby balancing employers’ valid interests in promoting the health of their employees with the rights of the workers themselves.

Finally, it contributes to the scholarly discussion surrounding the American antidiscrimination project by framing nicotine and obesity bans as discrimination on the basis of health status or “healthism.” The Article explores the normative justifications for and against understanding differentiation on the basis of health status as unfairly discriminatory, including the disparate impact employer healthism will have on vulnerable populations.

The Article proceeds in three parts. Part I casts these recent policies as healthism, or discrimination on the basis of health status, and explores both recent and past examples of employment policies that screen applicants based on health-related factors. Part II examines whether existing federal statutes would cover employer screens for nicotine use and obesity, concluding that they would not. This incongruity exists, at least in part, because the current laws governing health-related discrimination, such as the Americans with Disabilities Act and the Genetic Information Nondiscrimination Act, focus on protected traits, whereas the discriminatory actions at issue here differentiate on the basis of health-related conduct. Part III explores the need for legislation to protect against employer healthism. It begins by reviewing the existing state statutes that prohibit conduct-based discrimination. After exploring the arguments both for and against such legislation, the Article proposes that additional protections are necessary to protect workers, particularly members of certain historically disadvantaged groups, whom health-related employer screens would disproportionately harm. By denying access to wage work, employer-provided health insurance, and wellness programs, employer screens for health-related conduct could perpetuate existing health disparities, possibly leading to lower overall population health. These unintended—yet serious—negative effects warrant further regulation. Finally, the Article ends by asserting that lawmakers could draft legislation that would simultaneously protect employees while allowing employers to consider health-related factors when appropriate.

I.     Discrimination on the Basis of Health

Bans on hiring nicotine users and overweight individuals can be understood as employment practices that screen on the basis of health status. This Part catalogues prominent examples of these policies in the United States, exploring how employers have screened both prospective and current employees based on their health. It then takes the perspective of the employer, examining the reasons for choosing not to employ individuals classified as unhealthy. Finally, Part I ends by exploring the popular objections to health-related employment policies, framing differentiation on the basis of health as undue discrimination on the part of employers.

A.     Health-Related Employment Policies

Starting as far back as the late 1980s and early 1990s, various employers implemented policies designed to regulate the health of their employees.22 Early adopters of health-related employer screens include Alaska Airlines, Turner Broadcasting, Florida’s Gulf Power, and law enforcement agencies.23 Over the past five years, several kinds of companies have banned hiring smokers or other nicotine users, ranging from health-care providers,24 to state employers,25 to miscellaneous private employers, including a chain of casinos that opened its doors in 2012.26

Yet the regulation does not always stop with hiring bans: Some employers have likewise imposed conduct restrictions on their existing employees.27 For example, in 2004, Weyco, a health-insurance company based in Michigan, prohibited its employees from smoking.28 Weyco gave its workers who smoked fifteen months to quit, after which they would be required to take a breath test.29 At the end of the grace period, the company then terminated four employees who refused to submit to the test.30 Similarly, Best Lock, an Indiana-based manufacturing business with an alcohol-abstinence policy, fired an employee after he disclosed he had had drinks at a bar with friends a few years before.31

In 2006, Scotts Miracle-Gro adopted perhaps the most comprehensive—and notorious—initiative designed to promote employee health.32 Scotts’ policies targeted both tobacco use and obesity.33 The employees whom the company deemed to be at “moderate to high” health risk worked with individually assigned “health coaches” to set and achieve goals such as lowering cholesterol and losing weight.34 The workers who elected not to participate in the company’s program—or who failed to meet their coaches’ recommended health-management goals—faced increased health-insurance premiums.35 To support his employees in their efforts, Chief Executive Jim Hagedorn built a multi-million dollar gym and wellness center across the street from Scotts’ headquarters and offered free prescription drugs, medical care, and personal trainers.36

Scotts also instituted a hiring ban on nicotine users as part of its larger effort to improve employee health. Moreover, like Weyco, Scotts forbid current employees from using nicotine. In 2007, the company fired lawn care technician Scott Rodrigues after he tested positive for the substance.37 Rodrigues sued Scotts in federal court, alleging violations of the Employee Retirement Income Security Act (“ERISA”) and Massachusetts state law.38 Although the district court granted summary judgment in favor of the employer, the parties ultimately settled the case.39

Finally, Clarian Health Partners, a hospital chain, adopted a policy in 2009 that docked an employee’s pay $30 every other week for failing to meet certain health-related targets concerning blood pressure, weight, and cholesterol.40 However, after employees protested, the company revised its health-promotion policies by making them more incentive-based and less punitive.41 As in the case of Scotts, for many of these employers, the hiring bans evolved as part of a greater objective to improve the health and wellness of their workforce.42

As mentioned, Baylor began screening applicants for nicotine at the beginning of 2012. However, the new ban on hiring tobacco users was not its first move toward discouraging unhealthy behavior among its employees. Several years ago, the non-profit organization instituted a free program to help its workers quit smoking.43 Recently, Baylor has taken more drastic measures, such as banning smoking at all of its facilities and requiring employees who smoke to pay a health-insurance surcharge.44 The 2012 ban on hiring is, thus, Baylor’s most recent tool in its ongoing effort to regulate the health of its employees.45 Yet perhaps ironically, because the test detects nicotine from any source,46 Baylor’s screenings will likewise exclude individuals who are in the process of quitting or who have quit but rely on nicotine patches or gum.

B.     Motivations Behind Health-Related Employment Policies

Employers may have multiple motivations for taking an interest in the health of their employees. Those driving forces include: (1) costs related to employer-provided health insurance and worker safety and productivity, (2) stigma against smokers and overweight people, (3) promotion of a particular business image or workplace dynamic, and (4) paternalistic instincts.

1.     Cost

Employers often regulate their employees’ health to reduce costs.47 Through regulation, regulators hope to force individuals to bear their own costs, instead of passing those costs off to third parties.48 According to this rationale, regulation will lead to the optimal amount of the cost-generating activity by ensuring that people do not impose the cost of their choices on others.49 At least two kinds of cost-based concerns may drive employers to avoid hiring employees considered unhealthy: (1) higher health-insurance rates and premiums; and (2) productivity and safety concerns.50 Attempting to reduce these costs is thus economically rational behavior on the part of employers. In fact, health-care costs are arguably the greatest motivation for employers to regulate health-related conduct.51

The majority of Americans with health insurance hold group policies provided through their employers.52 Employers can either self-insure by independently covering the health-care costs of their employees,53 or purchase a health plan from a separate health-insurance company.54 While self-insured entities may feel the impact of their employees’ poor health more acutely, it is in the financial best interest of all employers to minimize the amount they spend on health insurance. Moreover, employees themselves stand to benefit from bans on unhealthy conduct, as low-risk employees must subsidize the costs of their riskier counterparts, either in the form of higher group insurance premiums or lower wages.55

Consequently, the increasing cost of health insurance (as well as anxieties regarding the unknown long-term effect of health-care reform) may have motivated recent hiring bans.56 The price tag on employer-provided insurance has more than doubled over the last ten years, giving employers a strong incentive to scale back costs by promoting employee health.57 As a group, smokers and other nicotine users face more health risks than people who do not use tobacco, thus making them more expensive to insure.58 Similarly, people outside of healthy weight ranges may also encounter more health risks, leading to higher insurance rates and premiums.59 Not surprisingly, smoking and obesity are often cited as two of the primary causes of high health-care costs.60 In fact, a 2007 poll revealed that 91% of employers thought they could lower their health-care costs by encouraging their employees to make healthier lifestyle choices.61

The relationship between a desire to cabin health-care expenditures and the regulation of employee health exceeds mere conjecture. Some companies explicitly referenced health-care costs to justify their decisions to adopt health-related employment policies.62 For example, Dr. Marc Boom, the president and chief executive officer of Methodist, cited cost as a factor in the decision to ban hiring nicotine users.63 Similarly, rising health-care costs in the early 2000s prompted Scotts to adopt its policy. From 1999 to 2003, the amount the company paid toward employee health care rose over 40%.64 After discovering that roughly a quarter of Scotts employees smoked and approximately half were overweight, Hagedorn adopted the contentious employee health program, specifically designed to lower those costs.65 Years earlier, Litho Industries Carolina, a North Carolina manufacturing company, adopted one of the first bans on hiring smokers; supporters of the policy cited health-care costs as a justification for the ban.66 Significantly, however, the impact of these policies on the actual cost of employer-provided health insurance remains unknown.67

The recent resurgence of health-related employment policies could be due in part to incentives created by health-care reform. David Gamage argues that the Affordable Care Act (“ACA”) creates effective taxes on employers who provide health insurance.68 In addition to the nondiscrimination rules discussed later in Part II, the ACA imposes penalties on employers who do not offer affordable health insurance to their employees under certain circumstances.69 Because the nondiscrimination rules prohibit differentiation based on individual health risk and the mandates require offering affordable nondiscriminatory policies, the ACA compels employers to provide insurance to potentially costly employees at reasonable prices. However, employers could circumvent this legislative double bind simply by not hiring employees who have high health risks in the first place. Thus, employers may be even more tempted to avoid hiring unhealthy employees post-health-care reform.70

Beyond insurance-related expenses, unhealthy workers may impose additional costs on their employers in the form of productivity and safety concerns.71 Workers at a lower level of relative health may require more time away from work to treat their conditions, thus leaving employers without necessary labor or in need of temporary help. Additionally, employees whose health is compromised may take longer to perform tasks, be more prone to accidents, or require more frequent or longer breaks.72 Both smoking and obesity are associated with longer and more frequent sick leave, as well as productivity loss on the job.73 Obese employees take from one to three additional sick days per year.74 When they are at work, moderately and extremely obese workers take more time to complete tasks and to perform physical job demands, thus decreasing their relative productivity as compared to lower weight workers.75 Smokers also tend to be more absent and less productive than nonsmokers.76 As a result, if an employer has the choice between equally qualified smoking and nonsmoking candidates, selecting the nonsmoker will almost always be the economically rational decision.77 Moreover, employees in poor health could also impose costs on their employers in the form of safety risks. For example, smokers have higher accident rates than their nonsmoking peers, perhaps due to their reduced reaction times.78 Similarly, obese people may find manual tasks more challenging based on their size or may more readily suffer fatigue-induced injuries.79 Thus, a variety of economically rational cost concerns could cause employers to implement practices designed to create a healthier workforce.80

2.     Stigma

General disdain for individuals deemed unhealthy may also underlie some health-related employment policies. “Stigma” is defined as a devalued social status that results from acknowledging and labeling human differences, linking negative stereotypes to those labels, then classifying—and subsequently discriminating against—individuals based on their labeled status.81 Stigma thus leads to negative differential treatment.

Stigmatized individuals are more likely to be rejected, excluded, and disliked.82 Importantly, the nature of stigma is contextual, as well as social: Whether a particular behavior or attribute becomes the object of stigma varies depending on the particular circumstances.83 Thus, a trait or behavior that engenders stigma in one social context may be considered neutral, or even positive, in another.84 That said, certain attributes are almost universally stigmatized.85

Smokers encounter discrimination and stigma. Research indicates that nonsmokers rate smokers lower with respect to any number of positive attributes, including health, competence, intelligence, and honesty, as well as attractiveness and moral virtue.86 In fact, some studies indicate that nonsmokers perceive smokers more negatively on practically every conceivable metric.87 Not surprisingly, a 2011 Gallup poll found that one in four Americans have less respect for smokers.88 The survey concludes that, in addition to heightened health risks and insurance costs, smokers face significant social disadvantage.89 These attitudes likewise affect their treatment in employment: Approximately one in four Americans would be less likely to hire an applicant upon learning she smokes.90

Similarly, research indicates that overweight people—especially the obese—generate negative affect. Twelve percent of Americans reported they had less respect for overweight people.91 Even children as young as three years old perceive being overweight as undesirable, classifying overweight people as ugly, stupid, and lazy.92 These unfavorable feelings are not confined to a single population: “Negative attitudes about the obese have been reported in children and adults, in health care professionals, and in the overweight themselves.”93 In a 2008 survey, 25% of respondents that self-identified as overweight reported negative attitudes toward significantly overweight individuals.94 Important to this Article, the stigma associated with being overweight leads to adverse treatment at work, as well as in other areas of life.95 According to a 2010 Gallup poll, almost one in five Americans would be less likely to hire an overweight person.96

In addition to social context, the degree of stigma may differ depending on the perceived nature of the defining characteristic. For example, stigmatized traits viewed as controllable trigger stronger negative reactions.97 Because both obesity and nicotine dependency are generally considered within a person’s control,98 smokers and the overweight may face greater stigma than comparable groups. For instance, opponents of nicotine bans point out that while many employers single out nicotine use for regulation, myriad other kinds of off-duty conduct—drinking alcohol, eating red meat, not getting enough rest, scuba diving, and even having multiple sexual partners—could also be potentially costly for employers.99 Thus, stigma tied to smoking and being overweight could drive employers to target nicotine use and obesity but not other risky behaviors.100

3.     Business Image

Relatedly, employers have a strong interest in the way in which customers perceive their businesses.101 Because attitudes undoubtedly affect people’s perceptions, stigma is also at play here. As noted above, stigma may affect decisions regarding whom to hire, fire, or promote.102 In those cases, the stigma stems from beliefs held by the decision-maker, i.e., the employer. Alternatively, stigma related to business image and workplace cohesion involves beliefs held by customers and co-workers, respectively. Consequently, the desire to promote a healthy business image—both externally and internally—may also drive employers to screen out applicants who appear unhealthy, regardless of the employer’s own attitudes. Importantly, so long as their actions do not violate the applicable antidiscrimination statutes, employers are free to adopt policies designed to promote their desired business image.103

Concerns related to business image incorporate multiple elements. To start, profitability may play a role. Customers may not want to interact with employees who look or act in a particular way. Given the stigma against smokers and the overweight, patrons may avoid companies that employ those individuals, leading the employers to lose money. These concerns may be particularly important to companies offering health-related goods and services.104 In addition to the bottom line, companies may also have health-related goals or missions. Thus, employing individuals who engage in unhealthy behavior may run counter to a company’s objectives. For instance, in justifying the Methodist policy, Boom explained, “This is part of a journey of wellness and making this a great place to work . . . . Employees work here to [take care of] patients. We can only do that if we’re leading by example.”105

Likewise, Baylor’s president and CEO Joel Allison defended the no-nicotine policy as part of the company’s underlying mission to promote health: “We want to be a good role model to our patients and the communities we serve . . . . For too long, we’ve been in the sickness business. It’s time to be in the wellness business.”106 Another representative of Baylor described the nicotine ban as an effort “to live by what we actually say.”107 Further, Allison predicted that other employers in the health-care industry might soon follow suit: “I think it is the future of health care . . . the focus on preventing illness.”108

It appears Allison was right. In addition to Baylor and Methodist, other health-related businesses have also recently stopped hiring nicotine users.109 A representative of Memorial Hermann Hospital, which has had its no-nicotine policy in place since June 2011, stated that the ban has already led to a “healthier workplace.”110

Similarly, in adopting its ban on hiring obese workers, CMC cited business image concerns. The policy explained that an employee’s appearance “should fit with a representational image or specific mental projection of the job of a healthcare professional.”111 CMC’s Chief Executive David Brown elaborated: “The majority of our patients are over 65, and they have expectations that cannot be ignored in terms of personal appearance . . . . We have the ability as an employer to characterize our process and to have a policy that says what’s best for our business and for our patients.”112 A local nutritionist echoed the hospital’s sentiments, indicating that he believed severely overweight employees could “set a poor example of health to patients.”113

It is worth noting, however, that employers may choose to use a health-oriented business image to mask their true motivations. Thus, companies could be primarily concerned about the reduced profitability associated with employing unhealthy people yet use their promotion of a healthy image as a more caring, altruistic justification for policies that may be viewed as draconian or coercive.

While the business image explanation appears most popular among health-care providers, other kinds of employers—such as gyms—may similarly want to hire employees whose appearance promotes health and wellness. Although not related solely to a health-oriented image, airlines famously restricted the weight of their flight attendants in the name of business image.114 Moreover, employers who work directly with clients and customers have chosen not to employ people who smoke because they believe the smell of cigarettes or other smoked tobacco products conveys an unprofessional image.115

The image a company presents can also affect its internal functioning. A workplace’s organizational culture plays a significant role in determining whether a new employee is accepted as a member of the team.116 People with similar attitudes and belief systems tend to work better together.117 Thus, certain employers—particularly small businesses—may attempt to create a unified company culture to facilitate workplace cohesion and employee morale.118 Inversely, employing workers who hold divergent beliefs could lead to intrapersonal conflicts, in turn compromising productivity and profitability.119 Workplaces may have cultural practices related to health. For instance, in the examples above, those employers expressed a commitment to presenting a healthy business image. Furthermore, given the documented hostility toward smokers and the obese, employers could choose not to employ those individuals simply to keep the peace.

4.     Paternalism

Finally, some employers may choose to regulate their employees’ health-related conduct simply because they believe that deterring unhealthy behavior and promoting wellness are morally desirable things to do.120 These entities are acting on a paternalistic impulse.

Paternalism occurs when an actor intervenes in another person’s life—without that person’s consent—because the actor believes her intervention will benefit the other person,121 either now or in the future.122 This sentiment can be summed up with the well-known parental axiom “I’m doing this for your own good.” Thus, advocates of health-related employment policies like nicotine and obesity bans may view those initiatives—not as untenable discrimination—but as benevolent attempts to improve employee health.123

Truly paternalistic actors regulate others regardless of costs.124 More commonly, however, regulators act with mixed motivations, combining both paternalistic urges and other interests.125 An employer who prohibits its employees from smoking or requires them to lose weight may simultaneously be acting to improve employee health out of concern for its employees’ wellbeing, as well as to reduce the potential costs of the unhealthy conduct.126 For instance, despite Hagedorn’s explicit invocation of ballooning health-care costs as his motivation, one journalist described Scotts’ wellness program as a “kind of companywide intervention that families use to help an addicted relative.”127 Like the referencing of altruistic company missions when discussing business image, in mixed motive paternalism cases, employers may choose to emphasize the morally upright nature of their policies and downplay their cost concerns to present a more favorable impression to the public. Not coincidentally, paternalism may be a particularly strong motivator for employers dedicated to providing health care or with health-related missions.

Notably, paternalistic impulses and stigma can sometimes be linked. It may be precisely because the regulator stigmatizes particular behaviors that it believes deterring the stigmatized conduct will have a positive impact. Thus, employers who find smoking or being overweight distasteful could paternalistically choose to discourage it.

C.     Health-Related Employment Policies as “Healthism”

The policies described here undoubtedly differentiate between applicants and employees on the basis of their perceived health. As demonstrated, this result is by design: Whether the reason is cost, business image, or simply negative attitudes, employers adopt these initiatives to distinguish between healthy and unhealthy workers. Opponents characterize these practices as unfairly discriminatory and unduly intrusive into workers’ private lives.

The word “discrimination” has more than one meaning. On a purely descriptive level, to discriminate means simply to differentiate.128 Discrimination is, therefore, standard in employment: Employers discriminate on the basis of education, work experience, and even appearance.129 Yet when given a normative overlay, discrimination takes on a distinctly pejorative tone.130 Used in this sense, to discriminate means to differentiate unfairly.131 This secondary meaning is the iteration most frequently used in the law.132

To be sure, sorting individuals based on their perceived health is nothing new.133 However, recently such practices have garnered attention as an undesirable variety of discrimination. In fact, the concern that employers and health insurers would use an individual’s health-related information to her detriment drove many of the conversations surrounding recent federal legislation such as the 2008 Genetic Information Nondiscrimination Act (“GINA”)134 and the ACA.135 Thus, labeling health-related policies as discriminatory—in the normative sense—implies that they are unjust and, consequently, an invalid basis for decision-making. “Healthism,” as used here, means unfair discrimination on the basis of health status.136

A 2011 survey reports that more than eight in every ten Americans believe that employers should not be allowed to refuse to hire an applicant solely because she smokes or is overweight.137 Yet what makes these policies unfair? Critics attack health-related employment policies for being too far attenuated from actual job performance. Michael Siegel, a professor of public health at Boston University, labeled the Methodist policy “a form of employment discrimination,” stating that the hospital was “making a hiring decision based on a group someone belongs to, not his or her qualifications for the job.”138

For many individuals, decisions that affect their personal health are made away from work and do not have an immediately perceptible effect on their ability to do their day-to-day jobs. Thus, a major critique lodged against employers’ use of health-related factors centers on the effect of the policies on conduct outside of the workplace. A representative of the American Civil Liberties Union (“ACLU”) of Texas, speaking out against Methodist’s decision to stop hiring tobacco users, commented: “We think this is an invasion of privacy and really overreaching.”139 Likewise, one newspaper editorial described Baylor’s no-smoking policy as “a step too far” in the employer’s efforts to combat smoking, asserting that the ban effectively puts smoking on par with illegal drug use.140 It criticized the ban as “extend[ing] far too deeply into the private lives of prospective workers.”141

Another major focus of these criticisms is the very legality of the underlying unhealthy conduct. For example, an editorial in USA Today asserted that Baylor’s hiring ban “punish[es] people for using a legal product on their own time.”142 The critique is thus two-fold: Employment decisions based on nicotine use—or weight for that matter—are undesirable (1) because they impact people’s non-work lives and (2) because they outlaw otherwise legal conduct.143

Nestled within these concerns is the fear that the current practices are merely the camel’s nose and that employers will grow increasingly interested in health-related factors as grounds for employment decisions. According to USA Today, Baylor’s initiative portends future restriction on the part of employers: “If smoker bans reduce health care expenses, cost-conscious employers might be tempted to stake out new and even more intrusive territory under the ‘wellness program’ banner.”144 The editorial asks its readers: “If employers routinely reject people who engage in risky, but legal, behavior on their own time, what about such things as overeating or drinking too much alcohol?”145

While the employers defend their policies as perfectly within the bounds of the law,146 critics nonetheless label the practices as morally unjustifiable. Commenting on CMC’s ban on obese workers, one state political candidate put it simply: “Even if it wasn’t illegal, [it] was discriminatory.”147

* * *

Over the past ten years, various employers have implemented policies geared toward regulating the health of their employees. These practices—which include hiring bans like those recently instituted by Methodist, Baylor, and CMC—are often part of a greater program intended to improve worker health. Employers may adopt health-related policies for various reasons. Yet regardless of the employer’s motive, health-related employment policies have garnered significant criticism as constituting healthism or unjust discrimination on the basis of health status.

II.     Healthism & Antidiscrimination

Critics of health-related employment policies denounce those practices as unduly discriminatory. Antidiscrimination laws protect individuals against certain kinds of differentiation deemed socially damaging.148 Many of these statutes apply to the workplace, thus limiting an employer’s ability to use particular kinds of information to make employment-related decisions, such as hiring, firing, and promotion. Among the existing laws that might apply to healthist employment actions are the ACA, GINA, the Americans with Disabilities Act (“ADA”), the Rehabilitation Act, and the Health Insurance Portability and Accountability Act (“HIPAA”). However, these statutes focus on an individual’s health-related traits, whereas the policies described in Part I center on health-related conduct. For that reason, current federal employment discrimination statutes offer little protection against those practices.

A.     Antidiscrimination & Protected Traits

While no federal protections explicitly bar employers from discriminating against unhealthy workers per se, existing legislation precludes discrimination on the basis of certain health-related attributes in both employment and in health insurance.

1.     Healthism & Employment Discrimination Law

Employers cannot discriminate on the basis of disability. Title I of the ADA provides that covered employers cannot discriminate against qualified individuals with disabilities “in regard to job application procedures, the hiring, advancement, or discharge of employees, employee compensation, job training, and other terms, conditions, and privileges of employment.”149 The Rehabilitation Act likewise prohibits discrimination on the basis of disability in federal employment.150 Both statutes define “disability” as “a physical or mental impairment that substantially limits one or more major life activities of such individual . . . a record of such impairment . . . or being regarded as having such an impairment.”151

While “health” and “disability” are not identical concepts, they are nonetheless related.152 The definition shared by the ADA and Rehabilitation Act tethers disability to the presence of “a physical or mental impairment.” Thus, to be a person with a disability for the purposes of those statutes, one must have a condition or trait that limits either physical or mental functioning. This requirement lends a decidedly medicalized air to the concept of disability within antidiscrimination law by linking those protections to an individual’s state of health.153 Further, people with disabilities often need more health-care services—and require those services more frequently—than people without disabilities.154 Finally, individuals that engage in unhealthy behaviors, like smoking or eating a high fat diet, may acquire disabilities as the result of their conduct.

Yet what happens if an employer fires someone for smoking or for being obese? Do these statutes offer any recourse? The ADA and the Rehabilitation Act undoubtedly apply to nicotine users and overweight people if those individuals develop impairments that substantially limit them, as required by the statutory definition of disability. For example, cancer is almost universally considered a disability.155 As a result, the ADA and the Rehabilitation Act protect the person who develops lung cancer after decades of smoking and the person who develops lung cancer because of a genetic proclivity equally. Similarly, if a severely obese person develops knee osteoarthritis as the result of her size and that condition substantially limits her ability to walk or perform other physical tasks, she enjoys protection. As such, certain individuals impacted by the policies described in Part I may also qualify as people with disabilities for the purposes of the ADA and the Rehabilitation Act. It is unclear, however, whether those statutes protect nicotine users and the overweight by virtue of their nicotine use or their weight absent any accompanying health conditions. Courts approaching these questions have reached different results.

Individuals at varying levels of overweight have challenged adverse employment actions as discriminatory on the basis of disability. While moderate obesity may not generally be considered a disability, severe obesity may qualify under some—albeit rare—conditions.156 Thus, often those cases center on whether the claimant qualifies as a person with a disability. As noted, the first prong defines “disability” as a substantially limiting impairment. Courts have thereby questioned both whether being overweight constitutes an impairment as well as whether it substantially limits the claimant in performing major life activities. They have similarly questioned whether weight-restrictive employment policies violate the Rehabilitation Act and the ADA by regarding overweight applicants and employees as disabled.

The lead case with respect to whether being overweight may qualify as a disability is the 1993 First Circuit case, Cook v. Rhode Island.157 In 1988, the Department of Mental Health, Retardation, and Hospitals refused to hire Bonnie Cook based on her severe obesity.158 Cook challenged the decision under the Rehabilitation Act.159 In deciding whether Cook qualified as an individual with a disability, the court noted that the Department specifically refused to hire her because the doctor charged with making that decision believed that Cook’s weight would affect her ability to perform certain physical tasks, such as “walking, lifting, bending, stooping, and kneeling,” to such a significant degree that she would be unable to adequately perform the job.160 Additionally, the court held that the mutability or voluntariness of a particular state was irrelevant to the inquiry as to whether the employer perceived that state as disabling.161 The First Circuit thereby concluded that Cook had provided enough evidence to prevail under either the first or the third prong of the definition.162

Importantly, however, the ADA’s definition of disability has been restricted—then subsequently re-expanded—in the ensuing years. Two major Supreme Court decisions limited the scope of disability under both the first and third prongs.163 Courts were likewise parsimonious with their coverage of obesity as a disability within that time period.164 Analyzing a claim under only the first prong, a Virginia district court found that to qualify as a disability, the claimant’s obesity must be caused by a physiological impairment and must substantially limit her across more than one kind of job.165 The court noted that “it remains unclear whether simple obesity falls within the broad sweep of the definition of physical impairment.”166 Similarly, in a 1997 case, the Sixth Circuit held that for an individual’s weight to constitute a disability the claimant must demonstrate a physiological cause because absent such a cause a person’s weight is “a mere, indeed possibly transitory, physical characteristic” and not in and of itself a disability.167 Ten years later the Sixth Circuit comparably limited the “regarded as” prong. Keeping with the Supreme Court’s restrictive interpretations of the definition of disability, the Sixth Circuit held that to be perceived as having a disability, the employer must mistakenly believe that the plaintiff has a disability that would garner protection under the statute.168 Because obesity absent an underlying physiological condition did not qualify as an impairment, the claimant could thus not be “regarded as” being disabled.169

However, in 2008, Congress passed the ADA Amendments Act (“ADAAA”).170 Among the statute’s changes was the re-expansion of the definition of disability.171 In particular, Congress explicitly rejected the notion that “regarded as” claims require the employer to hold a mistaken belief. Instead, the employer must only discriminate on the basis of a perceived impairment, regardless of whether the employer perceives that impairment to substantially limit a major life activity.172 On the other hand, plaintiffs suing under the first prong must still demonstrate that their weight substantially limits them but may have an easier time with the ADAAA’s broad rule of construction and new definition of major life activity. Yet although some commentators remain hopeful that the post-Amendments Act ADA may offer some refuge for employees and applicants who experience discrimination on the basis of their weight, the extent of the legislation’s application to those claims remains hazy.173

Nicotine users have enjoyed far less success with their disability discrimination claims than overweight individuals.174 Prior to the 2008 amendments, at least one federal court found that smoking did not qualify as a disability for ADA purposes.175 However, the statute does cover alcoholism under certain circumstances,176 indicating that it will at least sometimes protect employees from discrimination based on some kinds of addictions.177 Thus, if a claimant could establish that her nicotine addiction (in and of itself) substantially limits her in performing a major life activity, she could, in theory, have a viable claim.178 No court has ever made that determination positively,179 and commentators who have considered the question of the ADA’s application to various kinds of nicotine use have typically responded with a resounding “no.”180

As with overweight people, nicotine users may have their best shot at legal protection under the third prong of the definition of disability, especially after the ADAAA re-expansion. Specifically, if employers choose not to hire applicants merely by virtue of their status as nicotine users, arguably, those companies are “regarding” nicotine users as disabled.181 Yet while this application remains a possibility, given the previous interpretations, it seems likely that the ADA will offer little refuge against adverse employment actions on the basis of nicotine use, such as the policies described herein.182

Additionally, federal legislation also prevents employers from making adverse employment decisions on the basis of genetic information. GINA prohibits health insurers and employers from discriminating on the basis of genetic information.183 It also limits an employer’s ability to consider potentially useful information in their decision-making processes.184 While GINA might apply to non-medical genetic information,185 Congress clearly intended the statute to safeguard the results of health-related genetic tests.186

Health-related genetic data could interest employers for many of the same reasons as nicotine use and weight. All three kinds of information communicate potential health risks—and thus potential health costs—that could be relevant to an employee’s performance and consumption of health insurance. Moreover, like smoking and being overweight, certain genetic conditions could be the subject of stigma. Although studies have examined the link between genetics and obesity and addiction respectively, GINA probably offers no protection against the policies described in Part I. However, the statute does provide an example of a class of health-related information that, while of interest to employers because of its potential relevance to various costs, employers cannot consider due to legislative intervention. Thus, Title II of GINA can be understood as an antidiscrimination protection against a specific type of employer healthism.

2.     Beyond Employment: Healthism & Insurance Law

As noted, the majority of Americans with health insurance are insured through their employers. Consequently, although employment and health insurance are far from the same, they are nonetheless inextricably related. A discussion of healthism in employment would, therefore, not be complete without mentioning the legal protections that safeguard against discrimination on the basis of health-related factors in health insurance.

Health-related information occupies a central role with respect to health insurance. Given the nature of the American health-insurance system, health insurers have traditionally used health-related information to assess the potential risks of their insureds when determining eligibility and setting premiums.187 Thus, insofar as overweight people and tobacco users experience lower levels of relative health, they have also historically paid more for health insurance.

More recently, employers interested in regulating employee health have taken additional measures, such as insurance surcharges and other wellness incentives.188 Employers design their wellness programs to lower health-care costs and promote overall health.189 They may include positive incentives, like payment waivers, negative incentives, like insurance surcharges, or some combination of the two.190 Thus, health-related information may be used both when determining eligibility and rates and when offering incentives.

Certain laws restrict the factors health insurers may use when rendering their decisions. For instance, the ADA also has some limited applicability to employer-provided health insurance. In fact, the statute applies to employer wellness programs.191 However, the ADA includes an insurance “safe harbor” provision that allows health insurers to make decisions that relate to disability so long as those decisions are “based on sound actuarial principles or [are] related to actual or reasonably anticipated experience.”192 Given that both overweight and nicotine addiction do not likely qualify as disabilities and that the actuarial soundness of those factors as they pertain to health is well-documented, the ADA will offer little opportunity to challenge discrimination on the basis of weight or nicotine use in health insurance.

Outside the realm of antidiscrimination law, health-insurance statutes also constrain how insurers use certain kinds of health-related information. Importantly, the ACA—with its emphasis on preventing disease and promoting wellness193—specifically targets tobacco use and obesity as issues of public health.194

Yet even prior to 2010 health-care reform, Congress began regulating how group health insurers, like those responsible for employer-provided insurance, could consider certain kinds of information. HIPAA prohibits group health insurers (with some exceptions) from making decisions based on certain enumerated health-related factors.195 Specifically, HIPAA’s nondiscrimination provisions prevent those insurers from discriminating against individual group members when making decisions, but allow decisions made regarding the group as a whole.196 Despite the prohibitions on individualized eligibility and rating determinations, HIPAA and the ACA allow group health insurers to provide financial incentives through wellness programs.197 Moreover, the ACA includes sections explicitly designed to guide and support employers in instituting those initiatives.198

Wellness programs may offer participation incentives for merely enrolling,199 as well as attainment incentives for reaching one’s targeted health goals.200 Different rules govern these two kinds of incentives.201 Participation-based programs are automatically permissible.202 Conversely, attainment-based programs must meet five criteria: (1) the program must be “designed to promote health or prevent disease”;203 (2) it must reopen at least annually to allow for new enrollment; (3) its benefits must be available to all similarly situated individuals and the program must offer a reasonable alternative if participation is medically unadvisable or is unreasonably difficult because of a medical condition; (4) all materials must disclose the availability of reasonable alternatives or the possible wavier of standards; (5) finally, the financial incentives offered cannot exceed 30% of the cost of coverage.204 Further, the ACA’s final rules would allow attainment-based initiatives to offer incentives up to 50% for tobacco cessation programs.205 Thus, pursuant to HIPAA, the ACA, and their associated regulations, health plans offering wellness programs may differentiate on the basis of health, as well as on the basis of health-related conduct.206

Interestingly, people who are addicted to nicotine may be able to request a reasonable alternative in the context of attainment-based wellness programs. HIPAA’s regulations noted that a smoker could be entitled to an alternate program because it may be unreasonably difficult for that person to quit smoking because of “an addiction to nicotine (a medical condition).”207 Given the greater stigma against smokers, it is perhaps surprising that, while nicotine users might fall under an exception, more ambiguity exists with respect to whether overweight individuals would qualify for alternate programs or standards.208

Perhaps ironically, Congress drafted certain portions of the ACA to end healthism in health insurance.209 For example, the legislation eliminates pre-existing condition exclusions and health-based eligibility determinations.210 Moreover, as of 2014, the legislation prohibits health-status based rating in the individual and small-group markets.211 In lieu of health-related factors, those insurers must now set their premiums based only on the type of coverage provided (individual or family), geographic location, age, and—quite tellingly—tobacco use.212 Thus, while a person’s weight could perhaps be captured within one of the enumerated categories banned by the legislation, tobacco use is most assuredly not.

B.     Trait-Based vs. Conduct-Based Healthism

Existing federal legislation fails to provide comprehensive protection for people who engage in unhealthy conduct. While the statutes described above offer employees some recourse against discrimination on the basis of health-related factors, those protected statuses are based on relatively static factors like diagnosed conditions, genetic risk, and medical history. Conversely, the health-related employment policies described in Part I deal with statuses that are more dynamic. While a person often cannot un-acquire a disability—although it may be no small feat—she could quit smoking or lose weight. Nicotine and obesity bans thus screen for health-related behaviors. Because existing federal employment discrimination laws protect an individual from discrimination on the basis of her health status insofar as it relates to certain traits, yet not insofar as it relates to conduct, those statutes offer incomplete protection at best.

As used here “traits” are static characteristics that, once acquired, do not necessitate further conduct to maintain. Take disability as an example. While an individual may acquire a disability through voluntary—even morally reprehensible—conduct (such as getting into an automobile accident while drunk driving), once acquired, the status of disability is fixed.213 If an individual becomes paralyzed in a car crash, the continued state of paralysis mandates no subsequent conduct to maintain.

Conversely, “conduct” indicates a status that requires additional action to perpetuate. While conduct may range from voluntary to involuntary, what distinguishes conduct-based statuses from trait-based statuses is not the subject’s willingness to engage in the particular activity, but rather the need for additional action to maintain the status. Thus, under this definition even a status that results from addiction, other compulsive behavior, or factors otherwise outside of an individual’s immediate control would still be considered “conduct-based.”

The trait–conduct divide further illuminates why nicotine users and overweight people often do not enjoy the ADA’s protections by virtue of those statuses alone. For instance, to be “a smoker” someone must continue to smoke. However, if that same person develops lung cancer because of her smoking, once she quits she will still have lung cancer. The health-related status of nicotine use necessitates continued behavior, although the health-related status of lung cancer does not. Likewise, if an individual acquires diabetes through her eating habits, changing those eating habits—though capable of controlling her condition—may never completely eliminate it. Yet the line between traits and conduct blurs when it comes to a person’s weight. While people must continue to eat to maintain their current weight, it takes far longer for their status to alter, particularly when the individual is severely obese. Arguably, from a theoretical perspective, someone “becomes” a nonsmoker at the moment she has her final cigarette. On the contrary, a person does not stop being obese at the moment she decides to make eating or lifestyle changes. Only over time will her decisions aggregate to have an impact. That is not to say the individual’s weight is fixed: the status simply requires longer-term conduct to change. Put simply, statuses based on traits are static, whereas statuses based on conduct are dynamic.214

Traditionally, antidiscrimination law has focused on trait-based status, especially when that trait is considered immutable.215 Race, sex, and national origin are among the paradigmatic antidiscrimination categories. While scholars have highlighted the fluidity of these traits, those classifications are still considered by most to be relatively rigid.216 In the context of healthism, traits include inert statuses such as genetic information, disability, and medical history. Thus, the existing statutes that target healthism described in the preceding Subpart protect on the basis of specific traits, not conduct.

Inversely, antidiscrimination law has historically had far less sympathy for protected statuses resulting from conduct, particularly when that conduct is viewed as within the individual’s control.217 For example, courts have relied on the distinction between traits and conduct when deciding Title VII cases related to employer grooming policies. Perhaps the most notorious invocation of the trait–conduct dichotomy in that particular context occurred in Rogers v. American Airlines Inc., when a federal district judge dismissed a challenge to the company’s ban on all-braided hairstyles, a style choice worn predominantly by black women, writing:

[An all-braided hairstyle] is not the product of natural hair growth but of artifice. An all-braided hairstyle is an ‘easily changed characteristic,’ and, even if socioculturally associated with a particular race or nationality, is not an impermissible basis for distinctions in the application of employment practices by an employer.218

Moreover, the court explicitly acknowledged that American Airlines adopted the challenged policy to “project a conservative and business-like image.”219 Courts have thus been willing to uphold employment policies that differentiate on the basis of conduct seen as voluntary or otherwise within the employee’s control.

Important to the subject of this Article, unlike race or even disability, both weight and nicotine addiction are often considered the consequences of personal choices220 and thereby within the ambit of individual control.221 The tendency of antidiscrimination statutes to protect unchangeable traits explains why the ADA, GINA, and the various health-insurance laws allow bans on obesity and nicotine use.

The trait–conduct divide is also seen in the health-insurance context with the traditional dilemmas of adverse selection and moral hazard. Insurers rely on accurate risk assessment when issuing policies and setting premiums.222 People at higher risk are more likely to need health care and thus seek and purchase health-insurance policies. However, they are also more likely to use those policies, forcing insurers to pay out. Adverse selection describes the tendency of poorer risks to seek insurance more actively.223 Insurers insulate themselves against the costs of adverse selection by not issuing policies to risky applicants and by charging higher premiums when they do. Adverse selection can be understood as a trait-based phenomenon because it involves static factors, such as medical history, disease, or disability. Alternatively, moral hazard describes when an individual behaves in a way she otherwise would not because she knows she will not bear the costs.224 For example, in the context of health insurance an individual might opt to use health care, even in circumstances when she might not actually need it.225 This concern is thus conduct-related.

In short, the kinds of policies described in Part I discriminate on the basis of health-related statuses associated with conduct, not traits. It, therefore, comes as little surprise that the current federal laws designed to target healthism offer nicotine users and overweight people little recourse. To target employment policies designed to regulate unhealthy conduct, Congress must therefore draft a new kind of antidiscrimination law, one that focuses specifically on protected conduct not protected traits.226 The following Part explores how a majority of the states have done just that and considers the need for additional legal intervention.

* * *

Several statutes currently restrict the information employers may use in making job-related decisions. While the ADA and GINA, as well as health-insurance laws like HIPAA and the ACA, may prohibit discrimination on the basis of certain health-related traits, those laws ultimately fail to proscribe employer decision-making based on statuses associated with unhealthy conduct. Thus, if Congress or state legislatures wish to prevent employers from using an individual’s health-related conduct when deciding whether to hire, fire, or promote someone, they will need to draft a different kind of antidiscrimination statute.

III.     Outlawing Healthism

Current federal employment discrimination statutes do not proscribe employment decisions based on unhealthy conduct. This Part argues that additional protections are necessary to combat the unintended adverse effects of healthism. Specifically, it proposes that employer screens for health could disproportionally harm certain vulnerable populations, in particular racial and ethnic minorities, people with disabilities, and the poor and near-poor by simultaneously restricting their access to wage work and to employer-provided benefits. Given the relatedness of income, health insurance, and access to health care, healthist employment policies could perpetuate existing health disparities. Thus, this Part considers the kind of legislation necessary to outlaw the types of employment polices outlined in Part I. It begins with a summary of current state laws that restrict an employer’s ability to use health-related conduct to make job-related determinations. It then examines the arguments for and against antidiscrimination protection for unhealthy behaviors. Finally, it concludes that Congress or state legislatures should draft conduct-based antidiscrimination legislation—or amend their existing protections—to avoid harming historically disadvantaged groups. However, in so doing lawmakers must incorporate numerous safeguards to protect the valid interests of employers.

A.     Existing State Lifestyle Discrimination Statutes

A majority of states, as well as the District of Columbia, offer some kind of legislative protection against employment policies that discriminate on the basis of legal—but unhealthy—off-duty conduct.227 As a class, they are commonly referred to as lifestyle discrimination statutes.228 These laws, however, vary greatly in both their breadth and their depth of coverage.

One source of variation is the protected conduct. Some statutes offer very broad protections for lawful off-duty conduct or for the consumption of lawful products generally, whereas others specify particular behaviors and activities, often tobacco or alcohol consumption.229 Lawful product and conduct states include California, Colorado, Illinois, Minnesota, Montana, Nevada, New York, North Carolina, North Dakota, Tennessee, and Wisconsin.230 States that single out tobacco use for protection are Connecticut, Indiana, Kentucky, Louisiana, Maine, New Hampshire, New Jersey, New Mexico, Oklahoma, Rhode Island, South Carolina, South Dakota, Virginia, West Virginia, and Wyoming, as well as the District of Columbia.231

Although many states have specific prohibitions related to tobacco, fewer specifically deal with food and issues of weight. Minnesota, one of the states that protects “the use or enjoyment of lawful consumable products . . . off the premises of the employer during nonworking hours” lists food along with tobacco and alcoholic (as well as non-alcoholic) beverages as examples of such products.232 Similarly, Montana defines a “lawful product” as “a product that is legally consumed, used, or enjoyed and includes food, beverages, and tobacco.”233 Interestingly, Michigan is the only state that explicitly outlaws discrimination on the basis of weight,234 yet it does not have a lifestyle discrimination statute.

Other areas in which these laws differ are the employers they cover and the conduct they prohibit employers from taking. For instance, some states will not impose lifestyle discrimination protections on non-profit employers whose mission involves discouraging the protected conduct, such as foundations devoted to ending tobacco dependency or promoting health.235 Additionally, certain state lifestyle discrimination statutes protect only current employees from adverse employment actions.236

Further, many of these statutes include exceptions to allow employers to make conduct-based decisions that are job-related. Several statutes have exceptions for bona fide occupational requirements that relate to the job in question,237 while Missouri’s statute conditions its application on the protected conduct’s failure to interfere with job performance.238

Although many states have protections already in place, courts have limited the applicability of lifestyle discrimination statutes, calling the extent of their coverage into question. For example, California’s lifestyle discrimination statute, which offers facially broad protection “for lawful conduct occurring during nonworking hours away from the employer’s premises” has been interpreted narrowly.239 At least one California appellate division has held that the state’s lifestyle discrimination statute is merely a procedural enactment designed to enforce existing rights with no independent substantive protections of its own.240 Thus, even with broad conduct-based antidiscrimination statutes on the books, employees may still not enjoy adequate protection should courts choose to interpret those provisions restrictively.

B.     Additional Legislation

As summarized, the majority of states have opted to safeguard employees against at least some discrimination on the basis of unhealthy conduct. While the remaining states and Congress could follow suit and enact their own lifestyle discrimination legislation, the question remains whether the kinds of policies described in Part I are suitable territory for legislative intervention. Several norms govern American antidiscrimination law. Thus, before a legislature acts, it must consider whether screening applicants and employees based on their perceived health constitutes untenable discrimination worthy of antidiscrimination protection or a valid basis for differentiation on the part of employers. The following Subpart explores both the reasons for and against lifestyle discrimination legislation.

1.     Reasons for Regulation

Supporters of legislative interventions cite a number of arguments in favor of lifestyle discrimination legislation, including privacy, autonomy, and anti-paternalism, as well as immutability and the absence of choice when engaging in unhealthy conduct.

a.     Privacy, Autonomy, & Anti-Paternalism

Privacy, personal autonomy, and anti-paternalism represent three separate but interconnected justifications for stopping employers from making employment decisions based on unhealthy behavior.

Privacy is generally thought of as our “right to be let alone.”241 Being free from outside scrutiny is essential to our ability to maintain our independence and sense of self while simultaneously living communally.242 Relatedly, autonomy is the ethical principle that one’s personal choices should not be constrained by outside preferences or decisions.243 It respects an individual’s right to make decisions about herself and her wellbeing unencumbered. Autonomy arguments—often popular with Americans—tend to be tied up with issues of self-determination and personal identity.244 Likewise, the concept of anti-paternalism rejects the notion that third parties should be able to make decisions for the good of others. Anti-paternalism and autonomy are therefore two sides of the same coin: Whereas autonomy asserts that one’s personal choices should remain unconstrained by outsiders, anti-paternalism holds that outsiders—however well-intentioned—should not constrain one’s personal choices.245 Both autonomy and anti-paternalism champion individual liberty and freedom of choice. Privacy, thus, may function to preserve a person’s autonomy and to combat paternalism by carving out a sphere where one is answerable only to herself and her conscience for her decisions.

This right to be let alone is so central that it enjoys numerous kinds of legal protection.246 Individuals may sue in tort law for invasions of privacy247 or breaches of confidentiality.248 Likewise, certain aspects of our private lives rise to the zone of constitutional protection. For example, within criminal law, whether one has a reasonable expectation of privacy governs the ability of law enforcement to search and seize.249 Likewise, the substantive due process right to privacy (sometimes with its close cousin, the right to bodily integrity) has been cited as the basis for reproductive freedom250 and the ability of consenting adults to choose their sexual partners.251

Privacy in the workplace is of particular importance because of the high value placed on employment.252 Given the centrality of work to most Americans, employer infringements on privacy and personal autonomy are arguably just as constraining as similar intrusions by the government.253 Thus, opponents of employment policies restricting off-duty conduct assert that such practices present applicants and employees with a Hobson’s choice: either alter your behavior away from work or lose your job.254 According to this reasoning, employers are using economic means to unjustly control the personal choices of their workers, while simultaneously violating those workers’ personal autonomy and intruding into their private lives.255 In particular, critics of conduct-based employment policies focus on the off-duty nature of the prohibited behavior and the absence of an immediate impact on job performance.256 Some opponents have gone as far as asserting that workers hold positive “rights” to engage in legal off-duty conduct.257 The appeal of these kinds of arguments is so strong that even some proponents of tobacco control have expressed discomfort regarding bans on nicotine use.258

Moreover, justifying conduct-based employment policies because the targeted behavior relates to employee health opens the door for a broad range of restrictive practices. As Lewis Maltby, the president of the National Workrights Institute, points out, almost everything we do holds the potential to impact our health:

The more we learn about the relationships between behavior and health, the more we realize that everything we do in our private lives affects our health. If employers are permitted to control private behavior when it is related to health, virtually every aspect of our private lives is subject to employer control.259

Health could therefore be used to justify restrictions on all sorts of behaviors, including what we eat, how much we sleep, and the kinds of sports we play. Further, once an employer has access to an employee’s sample taken to test for nicotine or other substances, she could screen for any number of other kinds of health-related conditions or behaviors.260 Hence, arguments related to privacy, autonomy, and other anti-paternalistic ideals often invoke the slippery slope. If the law permits employers to regulate the nicotine use or the weight of their employees in the name of health, what will be next?

Privacy, autonomy, and anti-paternalism thus capture several concerns related to employment policies designed to regulate off-duty, health-related conduct. Specifically, those practices affect the choices workers make about their bodies and their lives, even when they are off the clock. This influence is viewed as unduly restrictive on private decision-making. Legislation designed to prohibit employers from making employment decisions based on unhealthy conduct could, therefore, ground its protections in the rhetorics of privacy and autonomy, distinguishing the conduct-based antidiscrimination claims from the more traditional status-based prohibitions described in Part II.261

b.     Immutability and the Absence of Choice

As mentioned in the proceeding Part, traditionally antidiscrimination law has not been favorable to statuses that are considered voluntarily chosen or within an individual’s control. Further, although this norm has existed for decades, the role of immutability appears to have taken on a renewed importance in the world of employment discrimination following GINA and the ADAAA.262

Frequently, nicotine addiction and obesity are understood as the consequences of voluntary unhealthy conduct, mainly tobacco use and poor eating choices. However, certain studies have called the level of voluntariness of those statuses into question. For example, research has indicated that genetic factors, which are largely determined before birth and exist outside a person’s control, may contribute to an individual’s propensity to gain weight.263 A genetic link to obesity would therefore call the mutability of that condition into question. Additionally, the ecological branch of public health indicates that environmental factors, which are largely out of an individual’s control, may contribute to her being overweight.264 Similarly, many smokers wish to quit.265 Opponents of nicotine bans have used this fact to argue that it is unfair to penalize people for conduct that they wish to abandon, thereby questioning the voluntariness of continued tobacco or other nicotine use.266 Thus, while conduct-based, these statuses are arguably not completely voluntary.

Importantly, the underlying assumption behind allowing employers to regulate health-related behavior is that the targeted conduct is something the applicant or employee can change if given the proper motivation. However, critics of employer wellness programs point out that this assumption comes into direct conflict with practical experiences related to lifestyle change. For example, many people regularly try to lose weight, struggle, and fail, despite their genuine intent and sincere effort to diet.267

Yet even assuming the voluntariness of unhealthy behaviors, what may have initially been a decision within an individual’s personal control may at some point become an undesired status. Our current selves inevitably make decisions that affect our future selves.268 While a fifteen-year-old may choose to light up a cigarette, the thirty-year-old whom she becomes may be desperate to quit smoking. Likewise, once a person reaches a particular degree of overweight, it may grow increasingly difficult to attain a healthy BMI. In short, our past decisions impact our current health, perhaps even more than our present ones. Moreover, many individuals who have already acquired unhealthy habits may actively want to change their behaviors but struggle—sometimes for reasons beyond their control269—with making those aspirations into a reality.270 Consequently, health-related employment policies beg the question, how accountable do we wish to hold people for the poor health-related decisions of their prior selves?

Thus, apart from the protection of privacy and autonomy, legislatures could also justify conduct-based antidiscrimination statutes by framing the protected behaviors as outside of individual control. As explored here, this argument could be made in at least two ways: (1) advocates could argue that the status, i.e., obesity, is not the result of voluntary conduct but rather of factors beyond an individual’s control, like genetics; (2) alternatively, advocates could argue that what began as voluntary conduct has now transformed into an un-chosen and perhaps immutable state, like in the case of addiction. Both lines of reasoning effectively collapse the distinction between trait-based and conduct-based statuses, eliminating the role of personal choice. Immutability arguments are therefore at odds with personal autonomy and privacy justifications. One champions the role of individual choice, while the other eliminates it.

c.     Harm to Historically Disadvantaged Groups

Additionally, employment policies that discriminate on the basis of health-related conduct may disparately impact historically disadvantaged groups, most notably racial and ethnic minorities, people with disabilities, and the poor. Individuals in those social categories are more likely to use tobacco and to be overweight, and simultaneously encounter more structural and institutional barriers to adopting healthy habits. As a result, adverse employment actions based on unhealthy conduct could further harm individuals who already experience discrimination on the job.

Both tobacco use and obesity are more common among certain disadvantaged populations, thereby amplifying the effect of employment policies designed to regulate health-related conduct on those groups. Racial and ethnic minorities are more likely to use tobacco;271 thus nicotine bans like the one adopted by Baylor will have a greater impact on non-whites.272 Similarly, minority group members on average tend to be heavier.273 As a result, weight-related policies will also have a disproportionate impact on racial and ethnic minorities. At first blush, it may seem that Title VII could offer some protection because courts have applied the statute to prohibit facially neutral policies with discriminatory effects. However, as noted, courts have been reluctant to extend those protections to situations in which the challenged policies differentiate on the basis of a conduct-based attribute thought to be within the control of the employee. Thus, the unwillingness to protect employees with braided hairstyles implies a similar unwillingness to protect employees who smoke or are overweight. Moreover, Title VII allows employers to adopt policies that produce a disparate impact if those policies are job-related and consistent with business necessity.274 Hence, if an employer can justify nicotine or obesity bans in terms of job-relatedness or business necessity, Title VII would not apply, regardless of the detrimental effect on racial and ethnic minorities. Given that smoking and being obese have measurable negative impacts on productivity and may undermine an employer’s business image, Title VII does not provide a solid basis for challenging health-related employment policies.

In addition to members of racial and ethnic minorities, people with disabilities are more likely to smoke and to be overweight than their non-disabled counterparts.275 Health-related employment policies will therefore also screen out individuals with disabilities in greater numbers. As discussed, while the ADA covers discrimination on the basis of smoking- and obesity-associated health conditions, courts have been reluctant to find that discrimination based on those states alone warrants protection.

Finally, the poor may also be more greatly impacted by the policies described in Part I than higher-wage workers. Lower-income individuals as a group weigh more and use tobacco more frequently.276 They also have less access to health care and less opportunity to adopt healthy lifestyle choices.277 Moreover, although the poor have not been expressly recognized as a distinct antidiscrimination category, they experience serious health disparities, which could be exacerbated by widespread employment policies that differentiate based on health.278

In addition to the fact that certain historically disadvantaged groups are more likely to engage in unhealthy conduct, those groups may likewise be less able to engage in the kind of lifestyle changes health-related employment policies are designed to promote. Critics of attainment-based wellness programs have discussed the concern that those initiatives could have a discriminatory impact.279 As one author points out, disadvantaged populations may encounter very real structural barriers to improving their health:

A law school graduate from a wealthy family who has a gym on the top floor of his condominium block is more likely to succeed in losing weight if he tries than is a teenage mother who grew up and continues to live and work odd jobs in a poor neighborhood with limited access to healthy food and exercise opportunities. And he is more likely to try.280

Thus, policies that encourage or reward healthier lifestyle choices assume the voluntariness of those changes without taking significant account of the impediments certain groups may face in actually implementing those behaviors.

Given the disparate impact of health-related employment policies on certain historically disadvantaged groups—mainly racial and ethnic minorities, people with disabilities, and the poor—and the absence of strong legal protections against those discriminatory outcomes, additional lifestyle discrimination statutes may be necessary at both the state and federal levels to eliminate the possibility of additional harm to those individuals.

2.     Reasons Against Regulation

As noted, employers discriminate all the time. In fact, to make employment decisions—such as whom to fire or promote—employers must be able to distinguish between candidates. Thus, not all differentiation in the workplace rises to the level of actionable discrimination. Moreover, employers have historically been able to hire and fire employees as they please. Restricting an employer’s ability to act on health-related factors thus eliminates one potential ground for differentiation between its workers and thereby conflicts with the doctrine of employment at-will. Additionally, critics of conduct-based regulation have maintained that the market itself will penalize employers who over-regulate and that adding nicotine use and weight to the canon of antidiscrimination law will erode the social relevance of traditional antidiscrimination categories.

a.     Freedom of Contract and Employment At-Will

One critique of legislation that would limit the kind of information employers may use to make their decisions cuts to the core of the very nature of employment in the United States. If employers are free to discriminate based on any number of things—education, work experience, and (sometimes) appearance—why not health?281

American employment law gives employers broad discretion in selecting their employees. This practice is commonly referred to as “employment at-will.” According to the employment at-will doctrine, both employees and employers are free agents, able to negotiate contractual terms to their mutual liking; consequently, either party—employer or employee—is free to unilaterally end that agreement at any time and for any reason.282 The vast majority of states adhere to this principle.283 In fact, only Montana has replaced employment at-will,284 instead adopting a statutorily outlined “good cause” standard.285 In 1991, the National Conference of Commissioners on Uniform State Laws proposed the Model Employment Termination Act, which included a provision modeled on the Montana approach, but it was rejected by the forty-nine other states.286 Hence, absent some kind of statutory exception to the common law, most American employers can hire and fire employees at their discretion. Thus, a statute that would limit an employer’s ability to make decisions based on an individual’s unhealthy conduct would violate employment at-will.

Employment at-will, however, is not absolute. In particular, antidiscrimination and wrongful discharge laws represent statutory exceptions to this doctrine.287 These laws are diverse in character and protect employees on the basis of both traits and conduct.288 Additionally, several states will override employment at-will for public policy reasons.289 The question remains, however, whether health-related conduct warrants such an exception.

b.     Market Effects

Critics of conduct-based antidiscrimination laws also argue that these statutes unduly interfere with the functioning of the market. As explained in Part I, absent some kind of policy designed to differentiate on the basis of employee health, employers—and fellow employees—subsidize the unhealthy choices made by their employees and co-workers respectively in the form of higher insurance premiums, while absorbing the associated costs of increased absenteeism and decreased productivity.290 It is therefore economically rational on the part of employers to adopt practices designed to reduce that cost-producing conduct. For example, banning smoking could not only lead to a healthier set of workers but also increased productivity.291 As a result, passing lifestyle discrimination legislation interferes with the ability of employers to make sound, economically rational decisions related to the expense of running their businesses.

To be sure, several trait-based antidiscrimination statutes interfere with economically rational employer decision-making. In fact, certain Supreme Court opinions have implied that the only kind of discrimination that the law should prohibit is “irrational” discrimination.292 Race discrimination is often cited as the paradigmatic example here: One’s skin color has no effect on her ability to perform a particular job. Thus, to differentiate solely on that basis is irrational. However, employers may have economically rational reasons for choosing not to hire women (who may be more likely to take family leave), people with disabilities (who may require accommodation), individuals with certain religious beliefs (who may require time away from work to worship), and persons with a genetic proclivity for serious disease (who may increase insurance costs if the related condition manifests), yet the law forbids such differentiation.293

With respect to lifestyle discrimination, a statute that prohibits employers from adopting policies to regulate costly employee conduct forces them to subsidize the expensive habits of their workers by internalizing the associated costs.294 M. Todd Henderson describes this result as “perverse,” explaining that “[t]he government subsidizes employer-provided health insurance (through numerous tax benefits for firms and employees) but then restricts the ability of corporations to regulate behaviors that will impact those health-insurance costs.”295

Given that worker health affects the bottom line, the issue of conduct-based employment policies could be left to the market, not to legislators.296 Without legislative intervention, employers would be free to adopt policies to regulate the health-related conduct of their employees. Under such a regime, Stephen Sugarman explains that employees who do not wish to be subject to such regulation could seek alternate employment or demand higher wages.297 According to this theory, employers who regulate too extensively—that is, beyond the optimal level—will suffer, thus causing them to adjust their policies to reach an efficient level of regulation.

Take recently adopted policies to screen for nicotine users. Henderson explains that a ban on hiring smokers could force the employer to offer increased wages, as desirable employees may want greater compensation to offset the intrusion into their private lives or to counterbalance other negative associations potential workers may have with the policy.298 In a market-regulated system, employers must calibrate their decisions to ensure the practices they implement will actually lower costs or increase profits. Thus, in the case of the nicotine ban, if the expense of higher wages exceeds the benefits of lower health-care costs or increased productivity, the employer will find itself in an economically worse position than before it instituted the ban.299 Leaving the issue of health-related employment policies to the market would thus not give employers free reign to restrict the off-duty behavior of their employees, but would rather push them to adopt an efficient level of regulation.

In sum, additional lifestyle discrimination legislation may not be necessary—or attractive—as the market could itself determine the extent to which employers can influence the off-duty, health-related conduct of their employees based on economic efficiency. However, there may be certain considerations aside from efficiency that warrant legislative intervention. Thus, even if the market could facilitate reaching an optimal level of health regulation, that economically optimal level could be undesirable for other reasons, such as its adverse impact on certain disadvantaged groups. Hence, legislation may still be necessary, not to promote efficiency, but to prevent adverse unintended consequences that would be costly in other ways.

c.     Erosion of Existing Protected Categories

A lifestyle discrimination statute may also impact the field of antidiscrimination law as a whole. Passing additional legislation to prohibit employers from making decisions based on health-related conduct could erode the power of existing antidiscrimination categories by placing that conduct on par with statuses like race, sex, or disability.300

Traditionally, statuses receive antidiscrimination protection in response to severe and on-going subordination.301 However, the social disadvantage on the basis of lifestyle choices is not currently perceived as a serious problem.302 Moreover, given that even the earliest bans came into being no more than a decade or two ago, there is not the history of discrimination that was present when Congress passed Title VII or the ADA.

Critics of lifestyle discrimination protection contend that passing such laws raises the protected conduct to the level of a civil right,303 thus creating a legally recognized “right” to using tobacco or to being overweight.304 Moreover, such protection would place nicotine use and obesity alongside race, sex, and disability as federally recognized protected statuses,305 yet absent an accompanying widespread public perception of discrimination or a history of subordination. Some authors have even argued that an antidiscrimination statute protecting unhealthy behavior undermines the very goals of the American antidiscrimination project itself, putting self-destructive behavior under the umbrella of civil rights, which should only protect constitutionally recognized liberties and immutable traits.306 Thus, opponents of lifestyle discrimination statutes assert that adding protected—yet voluntary—health-related conduct to the canon of antidiscrimination law elevates the status of that conduct to a civil right, thereby downplaying the significance of other protected groups.

d.     Benefits of Paternalism

Finally, lawmakers could choose not to outlaw health-related employment policies because those practices may have a positive social impact. Despite its negative effects on privacy and personal autonomy, paternalism can be welfare-enhancing. As Jeremy Blumenthal and Peter Huang have pointed out, paternalistic policies “may help people save money, live safer, be healthier and make better decisions for themselves.”307 In particular, employers who outlaw or discourage unhealthy conduct are also promoting personal, as well as public, health.308 Regardless of their effect on individual businesses, such policies hold the potential to further the greater good.309

Workers themselves agree. Employees at Scotts reported positive health-related outcomes, such as quitting smoking and losing substantial amounts of weight.310 Scotts’ policies promoted a positive work place culture.311 Thus, in the case of health-related employment policies, even the objects of paternalism themselves may actively acknowledge the benefits they obtain from complying.

C.     Going Forward

As explored in Part II, the current federal protections against healthism in employment fail to extend to health-related conduct. Although a majority of states have some protections on the books that deal with lifestyle discrimination, strong arguments exist both for and against this variety of legislation. It is, however, possible for legislatures to simultaneously honor the concerns on both sides of the issue in drafting additional conduct-based antidiscrimination legislation.

Additional conduct-based antidiscrimination protection would ensure that health-related employment policies do not substantially disadvantage already vulnerable populations. People need jobs. Basing hiring and firing decisions on health-related conduct holds the potential to harm those individuals who are in the most need of income, health insurance, and access to wellness programs. This Article proposes that policies involving adverse employment actions, such as those described in Part I, are simply the wrong mechanism for regulating employee health. However, even with additional protections in place, employers can still protect their interests and facilitate the health of their employees.

This Article, thus, proposes a middle intervention in an attempt to protect both workers and employers. Lifestyle discrimination legislation would offer some limited protection against adverse employment actions on the basis of unhealthy conduct to streamline protection across the several states yet would employ certain statutory safeguards to facilitate both the profit-maximizing and welfare-enhancing goals of employers.

On the federal level, legislation could be beneficial because, at present, protection varies wildly from state to state.312 The newly instituted policies described in Part I, while permissible in certain states, are outright banned in more than half of the country. Given the national—even global—nature of modern employment, such variation complicates the lives of both employers and employees: What may be acceptable conduct in one office of a national company may be prohibited in another. Furthermore, employers may not know the kinds of practices they can lawfully adopt and employees may not be aware of the protections available to them. In the past, Congress has intervened when states have varying levels of antidiscrimination protection to create a unified federal standard.313 A federal lifestyle discrimination statute would lend consistency to the current pastiche of state law. Yet even absent federal protections, passing or amending statutes at the state level could safeguard against the possible unintended negative effects of employer screens for health-related conduct described below.

1.     Protecting Employees

Health-related employment policies could do more harm than good in terms of their ability to improve American health as a whole. While screens for obesity or nicotine use might create a healthier set of employees, those policies achieve that end by excluding individuals who do not conform to norms of healthy behavior. Thus, workers who either cannot conform for structural reasons, or choose not to conform for autonomy reasons, may find themselves un- or under-employed. This outcome could in fact perpetuate population-wide health disparities, as those people who are screened out find themselves even more restricted in their ability to access health care, including preventive services, or to adopt healthier behaviors. Health-related employment policies could—perhaps counter-intuitively—generate a healthier workforce but a less healthy total population. Hence, employer healthism could increase the cost of the ACA on the federal government by raising the cost of subsidies for individuals seeking insurance through the exchanges.314 Moreover, as explained above, certain historically disadvantaged groups are statistically more likely to engage in unhealthy behavior. Thus, healthist employment policies will ultimately perpetuate or even exacerbate the current health disparities experienced by racial and ethnic minorities, people with disabilities, and the poor and near-poor.

Employer screens for health-related conduct could perpetuate existing health disparities, as well as have a negative impact on overall population health, for at least three reasons: (1) Those policies further entrench the structural barriers to accessing care and making healthy choices encountered by historically disadvantaged populations through restricting their access to needed wage work; (2) they deny the individuals engaged in unhealthy conduct the opportunity to participate in potentially beneficial workplace wellness programs; and finally (3) employer screens for unhealthy behavior limit the capacity to obtain employer-provided benefits, the primary source of health insurance for most insured Americans. The combination of these factors could have the unintended, yet strongly detrimental, effect of impeding the impacted populations’ ability to secure necessary care, thus ultimately leading to poorer overall health outcomes for those already disadvantaged groups.

To start, employer screens limit the number of jobs available to individuals who engage in unhealthy behaviors.315 By restricting the available work, employers who adopt those policies are likewise restricting access to wages for nicotine users and overweight people. However, finances are often necessary for obtaining health care and making healthy choices. People in poor health often pay more for health care. Those individuals tend to need more health care than their healthier counterparts, as well as to pay more out-of-pocket for the services they consume.316 Thus, wage work is particularly essential for people in poor health because they need an income to finance their care. Further, as discussed previously, making healthy decisions may be more challenging for poorer people. Fresh nutritious foods, like produce, cost more than processed foods.317 Additionally, lower-income individuals encounter greater barriers to exercising.318 Thus, people at lower levels of health need money both to pay for health care and to facilitate healthy decision-making. As a result, cutting off access to wage work via screens for unhealthy behavior is particularly devastating to those individuals.

Moreover, health-related employer screens cut off unhealthy people’s access to potentially valuable workplace wellness programs.319 Under such policies, potential or current employees who could have benefited from tobacco cessation or other employer wellness initiatives will be screened out prior to employment, or in cases like Scotts’, even fired. Ironically, the employees who will have access to employment-based health initiatives under such a system will be employees who are already engaging in healthy conduct. Likewise, the employees who could most benefit from such programs will be screened out and thus unable to utilize them.

Finally, and perhaps most importantly, screening for health-related behavior at the employment level could deprive health-insurance coverage to those who need it most. As noted, the American health-insurance system relies heavily on employer-provided benefits.320 Thus, for many Americans, being unable to work will also mean losing access to health insurance. Furthermore, health care in the United States is astronomically expensive and costs are still on the rise, even as compared to other industrialized nations,321 making health insurance a prerequisite to obtaining health care under many circumstances.322 Consequently, there is a link between being un- or under-insured and poorer health outcomes, as those individuals may go without needed care because they cannot afford it.323 Sadly, the uninsured are thus more likely to experience the most tragic negative health outcome: premature death.324 Given the reliance on employer-provided benefits, when employers screen out unhealthy workers, those individuals not only lose access to jobs but also to the associated health-insurance coverage, thereby cutting off their ability to obtain needed health care and leading to even poorer health outcomes. As some low-income individuals will qualify for Medicaid or the low-income subsidy under the ACA, the population that will be most impacted will likely be the near-poor.325 Importantly, Congress intended health-care reform to improve access to health care in part by reducing the number of uninsured Americans. However, when employers screen for health, they are restricting access to an important source of health insurance and for individuals who may have greater health-care needs.

To sum up, employer screens for health-related conduct simultaneously deny the impacted groups access to (1) wage work, (2) employer wellness programs, and (3) the leading source of health insurance in the U.S. The net effect is that those individuals, who are likely to be in poorer health, will be unable to obtain needed health care, leading to even lower levels of health and even worse health outcomes. However, it does not end there. As discussed, certain historically disadvantaged populations are more likely to use tobacco and to be overweight.326 Perhaps not surprisingly then, those same groups are also more likely to experience health disparities, as well as barriers to health-care access. Racial and ethnic minorities, who will be disproportionately impacted by screens for health-related conduct, already experience health disparities.327 In addition to being more likely to have a lower income, they make up a substantial portion of the uninsured.328 Thus, access to employment for the purposes of obtaining care is particularly essential for these groups. Screening them out of employment opportunities for engaging in unhealthy activities that are not relevant to job performance imposes yet another barrier to attaining needed health care.

Similarly, as mentioned, people with disabilities are more likely to smoke or to be overweight, making them probable targets for health-related employment policies. They are also often un- or under-insured.329 Moreover, despite using more health-care resources more frequently, individuals with disabilities face significant health disparities.330 They report lower levels of overall health, a higher rate of secondary conditions, and less access to preventive measures, leading to poorer health outcomes.331 Hence, employer screens for tobacco use and obesity could further limit their ability to access health care.

Lastly, people with low incomes, who are again more likely to engage in unhealthy behavior and thus be affected by the policies described in this Article, currently experience health disparities. Workers in poor and near-poor families are significantly less likely to have access to employer-provided health insurance than their wealthier counterparts.332 Perhaps relatedly, about a third of the poor and near-poor do not have health insurance.333 People with lower incomes are thus more likely to be in fair or poor health.334 Screens for health at the employment level foreclose additional opportunities for securing health care.

While seemingly discrete, the issues of income, insurance, and health are intimately intertwined.335 Given their disproportionate impact on populations already facing poor health and significant barriers to accessing care, health-related employment policies could exacerbate the existing health disparities experienced by these historically disadvantaged groups. As a result, employer screens for health could lead to poorer overall health outcomes, particularly for disadvantaged populations. These policies deny people their livelihood in an effort to promote healthier conduct. Although using tobacco and being overweight are conduct-based statuses, the underlying choices are not simple ones: The lack of access to healthy foods and time to work out, or a long-standing addiction to tobacco may be difficult obstacles to overcome without some help. Arguably, nicotine and obesity bans help workers by pushing them to be healthier. Yet those policies could also cause significant harm by restricting their access to wages, wellness programs, and employer-provided health insurance. Current and potential employees thus need protection from the unintended negative effects of health-related employment policies.

2.     Protecting Employers

However, employer screens for health-related conduct may also have positive impacts. As outlined in Part I, employers have a variety of reasons for wanting to regulate the unhealthy conduct of their employees. Those reasons include: (1) stigma against unhealthy behavior; (2) costs to the employer; (3) business image; and (4) paternalism. Some of these justifications are not only rational but also altruistic and could ultimately benefit society as a whole. Thus, the Article addresses each of them in turn, examining how lifestyle discrimination legislation might offer security for employees while simultaneously protecting the valid interests of employers.

Take stigma. In the case of irrational stigma, one benefit of additional legislation prohibiting employment discrimination on the basis of unhealthy conduct would be to ensure that employers do not bypass good workers for irrational reasons. Thus, such a law could actually be efficiency enhancing. Insofar as the stigma has a rational basis, it would tie in with the cost-related concerns described below. Moreover, antidiscrimination protection could reduce the stereotypes faced by those stigmatized groups.336

As explored in Part I, the rising costs associated with health care and the American dependence on employer-provided insurance have led employers to regulate the health of their employees. While geared toward protecting employees and applicants,337 a lifestyle discrimination statute could likewise make allowances for employer concerns related to cost. For example, in lieu of outright hiring bans, employers could instead adopt health-insurance surcharges or premium reductions.338 Instead of basing these adjustments on mere behavior, employers could use them as an incentive to participate in wellness programs designed to support employees in making healthier lifestyle choices.

Health-care reform recognizes this possibility. As discussed in Part II, when an attainment-based wellness program’s goal is tobacco cessation, the incentive can be up to 50% of the cost of coverage.339 The result is that tobacco users in the small group market could find themselves paying up to 50% more for health-insurance coverage. As a result, the ACA’s final rules require that small group insurers only institute the tobacco-related surcharge in conjunction with a wellness program.340 Alternatively, a focus on participation—and not goal attainment—with respect to wellness programs reduces the chances of disparities resulting from the structural barriers faced by historically disadvantaged populations.341 Thus, while premium adjustments could still have a disproportionate impact on certain vulnerable populations, employers could adopt policies that minimize that effect such as offering participation-based wellness programs as an alternative to the surcharge. Moreover, the negative impact of a premium adjustment will be less drastic than an outright denial of employment.

The statute could also consider employers’ concerns related to safety and productivity by limiting its protections to “qualified individuals.”342 Carefully drafted descriptions of job requirements could ensure that employers can take health and fitness into account when making employment decisions regarding jobs that involve strength, agility, or physical endurance. Moreover, like the ADA, legislation protecting health-related conduct could allow post-offer medical examinations to ensure an applicant’s ability to perform job-related tasks, so long as those examinations are confidential and administered to all employees in the same job category.343 Limiting protection to qualified individuals and allowing nondiscriminatory, post-offer medical examinations could thereby capture concerns related to safety and productivity.

Lastly, employers might also want to screen workers based on health to promote a particular business image. A lifestyle discrimination statute could permit such differentiation by providing a defense for employers, particularly when health is related to the products and services being offered. Like Title VII, legislation that prohibits discrimination on the basis of unhealthy conduct would allow employers to construct business images (and workplace cultures) that would make them market competitive. Thus, employers with health-related missions, such as Baylor, Methodist, or CMC, could argue that their health-related employment policies help establish their desired business image.344

Apart from efforts to reduce costs or maximize profits, employers may also regulate employee health due to paternalistic impulses. As noted, paternalism can foster good outcomes. Although it may be difficult to separate pure paternalism from mixed motives, lifestyle discrimination legislation could nonetheless permit—or even encourage—employers to adopt altruistic policies that do not infringe too deeply into workers’ private lives or harm socially disadvantaged groups. In particular, employers could adopt initiatives designed to promote a healthy workplace culture. In addition to the insurance-related wellness programs described above, employers could eliminate unhealthy foods in vending machines and at the cafeteria, set individual or group fitness goals, like weight loss or exercise targets, then reward good outcomes with non-job-related awards or privileges, such as firm-wide recognition of success or allowing individuals to a have a casual dress day.

In fact, some commentators have even proposed that changes to workplace culture provide a more effective means to promote employee health than wellness programs themselves.345 Legislators could thus draft the law to simply prohibit adverse employment actions, such as failing to hire, firing, or not promoting. Thus, so long as employers keep their incentives unrelated to compensation or employment status, their actions would be outside the scope of the legislation. However, employers must likewise be careful to structure their non-employment-related incentives so as not to create additional disadvantage. For example, Whole Foods garnered significant criticism when it adopted a policy giving employees an additional 10% sales discount based on health-related factors, including BMI.346 Critics asserted that the policy would either reward naturally thin employees347 or end up benefiting already healthy employees without making an overall improvement in employee health.348 Ironically, the employees who were arguably most in need of access to the kinds of healthy foods sold at Whole Foods—those with higher BMIs—would have to pay more for that food than their thinner counterparts.

By taking into account the interests of employers, additional protections against discrimination on the basis of unhealthy conduct at the state or federal levels could also bypass several of the critiques outlined above. While any intrusion into employer decision-making by its nature infringes on employment at-will, a lifestyle discrimination statute, if drafted properly, could bypass at least two remaining critiques: the erosion of existing antidiscrimination categories and the positive effects of paternalism. Congress or state legislatures could frame the statute’s protections as privacy- or autonomy-based protections of lawful conduct so as not to create a substantive “right” to smoke or to be overweight that could trickle into other areas or erode existing antidiscrimination categories. Finally, even limited wellness incentives would facilitate positive paternalism, especially if employers could use those programs to create a healthy culture at work. It is thereby possible to maximize the benefits and limit the harms of additional protective legislation for unhealthy conduct.

* * *

As established in Part II, existing antidiscrimination protections against employer healthism will not reach the kind of bans on unhealthy conduct described in Part I. Thus, to prohibit these practices, Congress or state legislatures must draft legislation that shields employees not just on the basis of their health-related traits, but also on the basis of their health-related behaviors. A majority of states have already taken such measures but the result has been an inconsistent hodgepodge of protections. By adopting federal legislation, Congress could streamline the law in this important area. Moreover, both state and federal lawmakers could structure lifestyle legislation in such a way as to acknowledge the concerns of employers and to protect their interests. Done properly, these statutes could be a win for employers and employees alike, while simultaneously promoting health and wellness for all.


With America’s continued dependence on employer-provided health insurance and the full implementation of health-care reform looming on the horizon in 2014, now more than ever employers have myriad reasons to take an interest in the health of their employees. Within the past few years, several employers adopted policies designed to regulate the health-related conduct of their employees. These policies are not, however, unique. Employers of many kinds have taken similar initiatives in the last two decades. Critics of these policies characterize them as unduly discriminatory.

While existing federal employment discrimination statutes offer some protection against discrimination on the basis of health, the health statuses they protect are uniformly trait-based. Conversely, bans on hiring smokers and obese people are matters of health-related conduct. Trait-based statutes, therefore, fail to provide adequate protection against those employer practices.

At present, there is no federal conduct-based antidiscrimination statute to protect applicants and employees who engage in unhealthy behaviors off the clock. Yet despite the lack of a federal law, the majority of states have enacted legislation to shield workers from just these kinds of conduct-based employment decisions. However, these statutes vary greatly in terms of both the employee conduct they protect and the employer actions they forbid.

Employer screens for unhealthy conduct could have an intensely adverse impact on groups who already experience significant health disparities. Thus, additional lifestyle discrimination protections may be in order. In drafting such laws, Congress or state legislatures could carve out certain exceptions to address the employer concerns of the cost of unhealthy conduct, the need to promote a certain business image, and the desire to further the wellbeing of their workers. However, additional legislation would not foreclose an employer’s ability to take an interest in the health of its employees. A conduct-based employment discrimination law could protect only qualified individuals and prohibit only specific adverse employment actions. Employers could thus continue to promote the healthy lifestyle choices of their employees through rewards programs that do not relate directly to employment status or compensation.

In an era when health is on the mind of most Americans, employers will continue to adopt policies related to employee health. While no solution is perfect, a carefully crafted statute could at once protect the private lives of employees and allow employers to act in both an economically and socially beneficial manner.

  1. [1]. Renée C. Lee, Hospitals Turn Away Applicants Who Smoke, Hous. Chron. (Oct. 29, 2012, 2:45 PM),

  2. [2]. Id.

  3. [3]. Id.

  4. [4]. Gary Jacobson, Dallas-Based Baylor Health Care System to Stop Hiring Smokers, Dall. Morning News (Sept. 23, 2011, 11:20 AM),; see also Editorial: Not Hiring Smokers Crosses Privacy Line, USA Today (last updated Jan. 29, 2012, 6:56 PM), [hereinafter Privacy Line]. The newly added nicotine screenings are expected to cost Baylor an additional $60,000 per year. Jacobson, supra.

  5. [5]. For a discussion of the critiques of these hiring bans as unfairly discriminatory, see infra Part I.C.

  6. [6]. Methodist currently employs more than 13,000 people. Lee, supra note 1. Baylor currently employs approximately 20,000 people and received around 3,800 job applications just last year. Jacobson, supra note 4.

  7. [7]. See Delray Beach, Florida Bans Hiring of Smokers or Tobacco Users to Save on Health Insurance, HuffPost Miami (Oct. 03, 2012, 6:03 PM),
    delray-beach-florida-bans_n_1933172.html [hereinafter Delray Beach].

  8. [8]. See Jim Kinney, Are Companies’ ‘No Smoking’ Policies Turning into ‘No Smokers Need Apply’?, The Republican (Feb. 4, 2013, 5:10 AM),

  9. [9]. See Richard Burr, Hospitals to Quit Hiring Smokers, Detroit News, Sept. 29, 2012, at A6.

  10. [10]. See James Ritchie, Cincinnati Hospitals Set Pace with Smoking Bans, Cincinnati Bus. Courier (May 18, 2012, 6:00 AM),
    2012/05/18/cincinnati-hospitals-set-pace-with.html?page=all; see also Two New Casinos in Ohio Say They Won’t Hire Smokers, Las Vegas Rev.-J. (Jan. 28, 2012, 2:02 AM), http://www.

  11. [11]. See Cindy Stauffer, LGH Won’t Hire Smokers; Job Applicants and Prospective Volunteers Will Have to Take a Screening Urine Test that Will Detect Nicotine, Intelligencer J. New Era, Oct. 5, 2012, at A1.

  12. [12]. Methodist and Baylor are in Houston and Dallas, respectively.

  13. [13]. Emily Ramshaw, Victoria Hospital Won’t Hire Very Obese Workers, Tex. Trib. (Mar. 26, 2012),

  14. [14]. The BMI metric is designed to assess whether an individual is within a healthy weight range. To calculate her BMI, one can multiply her weight in pounds by 703, divide by her height in inches, and divide again by height. The resulting number is her BMI. BMIs from 18.5–24.9 are considered healthy. See Nat’l Insts. of Health, Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults xiv (1998) (explaining BMI calculations and classifying BMIs as underweight, normal, overweight, obese, or extremely obese).

  15. [15]. Id. at 72. There are three classifications within the “obese” range: Obesity I (BMI 30-34.9), Obesity II (BMI 35-39.9), Obesity III/Extreme Obesity (BMI of 40 and above). Id. at xiv tbl.ES-2.

  16. [16]. Id. at 12–19 (explaining the health risks associated with being overweight and obesity); see also Erin Pradia, Citizens Medical Center Reverses Ban on Hiring Obese People, Victoria Advoc. (Apr. 12, 2012, 10:45 AM),
    citizens_health_041212_173212/ (citing nutritionist Tim Holcomb).

  17. [17]. For a discussion of the role of business image in hiring for health-related employers, see infra Part I.B.3.

  18. [18]. Pradia, supra note 16.

  19. [19]. For example, Henry Ford notoriously regulated the conduct of his employees. See Lewis Maltby, Whose Life Is It Anyway?: Employer Control of Off-Duty Smoking and Individual Autonomy, 34 Wm. Mitchell L. Rev. 1639, 1639 (2008) (describing the invasive nature of the Ford Motor Company’s policies); see also James A. Sonne, Monitoring for Quality Assurance: Employer Regulation of Off-Duty Behavior, 43 Ga. L. Rev. 133, 140 (2008) (describing Ford’s “Sociology Department”).

  20. [20]. Employers have been adopting policies related to their employees’ health since the early 1990s. See, e.g., Jessica Jackson, Colorado’s Lifestyle Discrimination Statute: A Vast and Muddled Expansion of Traditional Employment Law, 67 U. Colo. L. Rev. 143, 143 n.1 (1996) (citing Paula Span, Smokers’ New Hazard: No Work; Health Costs Behind Job Bias Issue, Wash. Post, Nov. 12, 1991, at A1).

  21. [21]. Currently, depending on how widely the term is defined, 30 states have legislation prohibiting lifestyle discrimination in employment. See Maltby, supra note 19, at 1647 n.43. For an in-depth discussion of the state legislation, see infra Part III.A.

  22. [22]. This Article describes efforts of modern employers to regulate the non-job-related health of their employees. Yet, depending on where one draws the line, employers have been regulating health for far longer, such as the company towns of the late 1800s.

  23. [23]. Privacy Line, supra note 4.

  24. [24]. Jacobson, supra note 4.

  25. [25]. Delray Beach, supra note 7; Privacy Line, supra note 4.

  26. [26]. Delray Beach, supra note 7; Privacy Line, supra note 4; Two New Casinos in Ohio Say They Won’t Hire Smokers, supra note 10.

  27. [27]. See M. Todd Henderson, The Nanny Corporation, 76 U. Chi. L. Rev. 1517, 1517–18 (2009).

  28. [28]. Id. (describing Weyco’s policy).

  29. [29]. Id.

  30. [30]. Id.

  31. [31]. He later sued and won because the employer’s policy was not reasonably related to a business interest. See Best Lock Corp. v. Review Bd., 572 N.E.2d 520, 521–22 (Ind. Ct. App. 1991); see also Ann L. Rives, Note, You’re Not the Boss of Me: A Call for Federal Lifestyle Discrimination Legislation, 74 Geo. Wash. L. Rev. 553, 553 (2006).

  32. [32]. Reportedly, Jack Welch told Scotts’ CEO Jim Hagedorn that Hagedorn had “balls of steel.” Michelle Conlin, Get Healthy—Or Else: Inside One Company’s All-Out Attack on Medical Costs, BusinessWeek, Feb. 26, 2007, at 60. Hagedorn himself stated: “This is an area where CEOs are afraid to go. A lot of people are watching to see how badly we get sued.” Id. For additional description of the Scotts policy, see Henderson, supra note 27, at 1548 (describing the Scotts policy and its reception); James Ruffin Lawrence, III, “Let Us Now Try Liberty”: Freeing the Private Sector to Tackle North Carolina’s Tobacco Addiction by Reinstating Employment Freedom of Contract, 90 N.C. L. Rev. 510, 511–15 (2012) (describing the Scotts policy and its reception).

  33. [33]. Lawrence, supra note 32, at 512; see generally Conlin, supra note 32 (detailing Scotts’ wellness policies).

  34. [34]. Conlin, supra note 32, at 58 (detailing a phone conversation between a Scotts employee and his health coach).

  35. [35]. Id. at 64 (explaining that Scotts employees who refuse to take health-risk assessments paid $40 per month more in premiums and those who did not comply with their health coaches’ plan paid an additional $67 per month).

  36. [36]. Id. (describing the wellness center).

  37. [37]. Conlin, supra note 32, at 60, 69; see also Lawrence, supra note 32, at 514.

  38. [38]. See Rodrigues v. EG. Sys., Inc., 639 F. Supp. 2d 131, 132 (D. Mass. 2009); see also Lawrence, supra note 32, at 514 (describing the Rodrigues litigation).

  39. [39]. Lawrence, supra note 32, at 514.

  40. [40]. Brendan W. Miller, Note, Your Money or Your Lifestyle!: Employers’ Efforts to Contain Healthcare Costs—Lifestyle Discrimination Against Dependents of Employees?, 5 Ind. Health L. Rev. 371, 377 (2008) (describing the Clarian policy).

  41. [41]. Id.

  42. [42]. Micah Berman & Rob Crane, Mandating a Tobacco-Free Workforce: A Convergence of Business and Public Health Interests, 34 Wm. Mitchell L. Rev. 1651, 1653-54 (2008) (noting that for Scotts, the Cleveland Clinic, Union Pacific Railroad, and Alaska Airlines “the tobacco-free workforce policy is part of an overall workplace wellness program”); Christopher Valleau, If You’re Smoking You’re Fired: How Tobacco Could Be Dangerous to More than Just Your Health, 10 DePaul J. Health Care L. 457, 458 (2007) (noting that Weyco’s tobacco cessation program included acupuncture and hypnosis).

  43. [43]. Jacobson, supra note 4; Privacy Line, supra note 4. Baylor will pay for both its employees and their spouses to go through the smoking cessation program. Jacobson, supra note 4.

  44. [44]. Jacobson, supra note 4; Privacy Line, supra note 4.

  45. [45]. For more on the nicotine hiring ban and its role in Baylor’s efforts to curb tobacco use by its employees, visit Baylor’s website at, Baylor Health Care System to Implement Nicotine-Free Hiring Policy in 2012, Baylor Health Care Sys., (last visited Nov. 19, 2013).

  46. [46]. See Jacobson, supra note 4 (explaining that “[t]he test will detect nicotine from cigarettes, cigars, pipes, chewing tobacco, gum, and even patches,” but according to Baylor, “[t]he threshold level for a positive test should be high enough to exclude exposure from second-hand smoke”).

  47. [47]. Henderson, supra note 27, at 1522 (identifying cost internalization as “the primary basis for public and private regulation”).

  48. [48]. Id. (describing the cost-internalization theory of regulation as “forcing individuals or firms to bear the costs they impose on others”).

  49. [49]. Id. at 1522–23.

  50. [50]. Advocates of tobacco bans often cite cost to justify the policies. See, e.g., Berman & Crane, supra note 42, at 1652 (arguing that “making the transition to a tobacco-free workforce may be an easy and cost-effective way for businesses to substantially reduce healthcare costs and increase productivity”). M. Todd Henderson calls the cost-benefit analysis behind decisions to prohibit unhealthy conduct “simple.” Henderson, supra note 27, at 1517–18 (“The logic behind Weyco’s decision is simple—firms bear some of the costs of individuals smoking (including higher health-insurance costs, lower productivity, increased absenteeism, and so on) and therefore have an incentive to reduce these costs.”).

  51. [51]. Henderson, supra note 27, at 1542 (describing health-care costs as “the biggest driver of modern corporate nannyism”); see also Stephen D. Sugarman, “Lifestyle” Discrimination in Employment, 24 Berkeley J. Emp. & Lab. L. 377, 392 (2003) (noting that employers who do not hire employees who engage in unhealthy behavior “seem to be most concerned about potential health care costs”). Authors have been attributing such regulation to the cost of health care for almost two decades. See, e.g., Lewis L. Maltby & Bernard J. Dushman, Whose Life Is It Anyway—Employer Control of Off-Duty Behavior, 13 St. Louis U. Pub. L. Rev. 645, 645 (1994).

  52. [52]. As of 2009, 87.2% of privately insured Americans procure their health insurance from their employers. Carmen DeNavas-Walt et al., U.S. Dep’t of Commerce, Income, Poverty, and Health Insurance Coverage in the United States: 2009, at 71 tbl.C-1 (2010), available at; see also Miller, supra note 40, at 380–82 (describing employer-provided health insurance).

  53. [53]. The Kaiser Family Found. & Health Research & Educ. Trust, Employer Health Benefits: 2012 Annual Survey 160 (2012), available at http://kaiserfamilyfoundation. (defining a “self-funded plan”). While employers that self-insure cover costs from their own budgets, they may hire third-party administrators to manage the plans. Id.

  54. [54]. Id. (defining a “fully insured plan”).

  55. [55]. Henderson, supra note 27, at 1518. In fact, Henderson argues that employees and owners not only stand to benefit from employment policies regulating employee conduct both on and off the job, but also demand it. See generally id. (arguing that the question is not whether third parties should regulate conduct but rather which third parties are the most effective and appropriate regulators—governments or private firms). Similarly, Conlin notes that “people could start blaming unhealthy colleagues for helping push up premiums.” Conlin, supra note 32, at 60.

  56. [56]. See Henderson, supra note 27, at 1529 (noting that “[b]oth firms and the government, which pay most healthcare expenditures, are increasingly active in policing health and safety practices in an attempt to reduce healthcare costs”); see also Berman & Crane, supra note 42, at 1651–52 (describing controversies surrounding previous tobacco bans as “the conflict between company executives determined to cut healthcare costs and ‘privacy advocates’ (or in some articles ‘civil rights activists’)”).

  57. [57]. Privacy Line, supra note 4. Employers offering health-insurance benefits annually pay about $11,000 for family coverage and $4500 for individual coverage. See The Kaiser Family Found. & Health Research & Educ. Trust, supra note 53, at 70; see also Miller, supra note 40, at 373–75 (describing the effects of the rising cost of health care on employers and employees); Sonne, supra note 19, at 152 (describing rising health-care costs for employees).

  58. [58]. Recent estimates indicate that each smoker costs his or her employer an additional $4000 each year. See Harald Schmidt, Kristin Voigt & Ezekiel J. Emanuel, The Ethics of Not Hiring Smokers, 368 New Eng. J. Med. 1369, 1369 (2013). But see Jill R. Horwitz, Brenna D. Kelly & John E. DiNardo, Wellness Incentives in the Workplace: Cost Savings Through Cost Shifting to Unhealthy Workers, 32 Health Aff. 468, 472 (2013).

  59. [59]. Pradia, supra note 16.

  60. [60]. See, e.g., Henderson, supra note 27, at 1544 (citing smoking and obesity as “[t]wo of the biggest causes” of health-care costs).

  61. [61]. Michelle M. Mello & Meredith B. Rosenthal, Wellness Programs and Lifestyle Discrimination—The Legal Limits, 359 New Eng. J. Med. 192, 192 (2008). However, despite the widely held belief on the part of employers, of late, scholars have challenged the notion that employees who smoke or are obese actually spend more on health care. See, e.g., Horwitz, Kelly & DiNardo, supra note 58, at 471.

  62. [62]. However, employers may only be tempted to consider health and health-related behaviors in making employment decisions when there is a surplus of qualified candidates. See Sugarman, supra note 51, at 398.

  63. [63]. Lee, supra note 1.

  64. [64]. Lawrence, supra note 32, at 511–12.

  65. [65]. Conlin, supra note 32, at 61–62 (noting that after “watch[ing] health-care costs explode” and being disappointed by the lack of response by the government and health insurers, Hagedorn instituted the policy when he discovered that “[h]alf of his 6,000 employees were overweight or morbidly obese [and] a quarter of them smoked”); see also Henderson, supra note 27, at 1548 (noting that Scotts adopted its policies in response to growing health-care costs); Lawrence, supra note 32, at 512 (explaining that the high percentage of Scotts employees facing significant and potentially costly health risks prompted the policy).

  66. [66]. See Jim Morrill, Bill Bars Bosses from Controlling Workers’ Off-Hours Habits, Charlotte Observer, June 24, 1992, at 1A (quoting Roger O’Quinn, executive vice president of the North Carolina chapter of the American Cancer Society, as saying “[i]f the company is paying for your health or your family’s health care, I think there’s a trade-off there” (internal quotation marks omitted)); see also Lawrence, supra note 32, at 529.

  67. [67]. See Lee, supra note 1 (noting the reporter’s inability to confirm whether tobacco bans have generated savings for the instituting employers).

  68. [68]. David Gamage, Perverse Incentives Arising from the Tax Provisions of Healthcare Reform: Why Further Reforms Are Needed to Prevent Avoidable Costs to Low- and Moderate-Income Workers, 65 Tax L. Rev. 669, 670 (2012).

  69. [69]. Id. at 692–98. Specifically, the ACA penalizes employers if any of their employees receive subsidies on one of the health-insurance exchanges.

  70. [70]. Alternatively, employers could also attempt to avoid the cost of insuring riskier employees post-ACA by attempting to force those people into the individual health-insurance market. See Amy Monahan & Daniel Schwarcz, Will Employers Undermine Health Care Reform by Dumping Sick Employees?, 97 Va. L. Rev. 125, 128 (2011). For more analysis of the impact of health-care reform on employer motivations and the general effect of employer-provided health insurance on access to care, see Jessica L. Roberts, The Privatization of Health Policy (unpublished manuscript) (on file with author).

  71. [71]. Berman & Crane, supra note 42, at 1655–56; Henderson, supra note 27, at 1543.

  72. [72]. See, e.g., Berman & Crane, supra note 42, at 1655 (explaining that “studies have consistently demonstrated that employees who smoke are less productive than employees who do not”); id. at 1656 (“Smoking employees are also more likely to suffer work-related disability and on-the-job accidents, injuries, and facilities.”).

  73. [73]. Suzan J.W. Robroek et al., The Role of Obesity and Lifestyle Behaviours in a Productive Workforce, 68 J. Occupational & Envtl. Med. 134, 137 (2011).

  74. [74]. K. Neovius et al., Obesity Status and Sick Leave: A Systematic Review, 10 Obesity Revs. 17, 21–22 tbl.2 (2008).

  75. [75]. Donna M. Gates et al., Obesity and Presenteeism: The Impact of Body Mass Index on Workplace Productivity, 50 J. Occupational & Envtl. Med. 39, 42 (2008).

  76. [76]. See, e.g., William B. Bunn III et al., Effect of Smoking Status on Productivity Loss, 48 J. Occupational & Envtl. Med. 1099, 1105 (2006); Michael T. Halpern et al., Impact of Smoking Status on Workplace Absenteeism and Productivity, 10 Tobacco Control 233, 235-36 (2001).

  77. [77]. Karen L. Chadwick, Is Leisure-Time Smoking a Valid Employment Consideration?, 70 Alb. L. Rev. 117, 135 (2006).

  78. [78]. Martin J. Lecker, The Smoking Penalty: Distributive Justice or Smokism?, 84 J. Bus. Ethics 47, 48 (2009).

  79. [79]. When commenting on CMC’s hiring ban, nutritionist Tim Holcomb explained that severely overweight workers “may be hampered in performing tasks at the hospital because of the physical nature of the work.” Pradia, supra note 16.

  80. [80]. See Henderson, supra note 27, at 1519 (“Contrary to rights-based accounts, nannyism by firms is generally not premised on malice, invidious discrimination, or exploitation of unequal bargaining power between managers and employees. It is, in fact, inevitable in cases where third parties bear some costs of others’ behavior.”); see also id. at 1528 (“[I]t would be irrational for the third parties not to try to influence the behavior of the individuals imposing the costs, since it would be subsidizing socially inefficient conduct.”).

  81. [81]. See Bruce G. Link & Jo C. Phelan, Conceptualizing Stigma, 27 Ann. Rev. Soc. 363, 366–75 (2001); see also Webster’s Third New International Dictionary 2243 (1993) (defining stigma as “a mark of shame or discredit” and “a mark or label indicating deviation from a norm”).

  82. [82]. Brenda Major, Stigma, in Encyclopedia of Social Psychology 944, 944-47 (Roy F. Baumeister & Kathleen D. Vohs eds., 2007).

  83. [83]. Id.

  84. [84]. Major explains that certain historically devalued statuses, such as being gay, are less stigmatized in certain social contexts, e.g., in a gay bar. Id.

  85. [85]. Id.

  86. [86]. Bryan Gibson, Psychological Aspects of Smoker–Nonsmoker Interaction: Implications for Public Policy, 49 Am. Psychologist 1081, 1081 (1994).

  87. [87]. Id. But see Scott Burris, Disease Stigma in U.S. Public Health Law, 30 J.L. Med. & Ethics 179, 187 (2002) (noting the argument that smoking “[s]atisfies all the criteria of stigma in a formal way, but that in none of the domains is the effect serious enough to rise to the level of stigma.”).

  88. [88]. Lydia Saad, One in Four Americans Have Less Respect for Smokers, Gallup (Aug. 5, 2011), Interestingly, Americans have more respect for the overweight, relative to smokers. Id. (finding that 25% of respondents had less respect for smokers, whereas only 12% had less respect for overweight people).

  89. [89]. Id.

  90. [90]. Elizabeth Mendes, Health Situation May Present Barrier for Many Job Seekers, Gallup (Aug. 20, 2010),

  91. [91]. Saad, supra note 88.

  92. [92]. See Sophie Lewis et al., How Do Obese Individuals Perceive and Respond to the Different Types of Obesity Stigma that They Encounter in Their Daily Lives? A Qualitative Study, 73 Soc. Sci. & Med. 1349, 1350 (2011).

  93. [93]. Nat’l Insts. of Health, supra note 14, at 20 (citations omitted). But see Saad, supra note 88 (finding that bias against the overweight is most prevalent in high-income, college-educated Americans).

  94. [94]. Frank Newport, Impact of Smoking, Being Overweight on a Person’s Image, Gallup (July 21, 2008),

  95. [95]. Nat’l Insts. of Health, supra note 14, at 20; see also Major, supra note 82, at 944.

  96. [96]. Mendes, supra note 90.

  97. [97]. Major, supra note 82, at 945 (“People with stigmas that are believed to be controllable are more disliked, rejected, and less likely to receive help than are people whose stigmas are perceived as uncontrollable.”).

  98. [98]. Id. (“Obesity, drug addiction, and child abuse are examples of marks generally perceived to be controllable; whereas skin color and physical disability are examples of marks generally thought to be uncontrollable.”).

  99. [99]. See, e.g., Maltby, supra note 19, at 1640–41 (noting the medical risks associated with those and other behaviors). Some employers have regulated various kinds of employee conduct. See id. at 1641–42 (describing employer bans on skiing, motorcycle riding, alcohol consumptions, and having high cholesterol).

  100. [100]. See, e.g., Lee, supra note 1 (stating Michael Siegel, a Boston University public health professor, considers the Methodist policy to be “a value judgment against smoking”).

  101. [101]. Sugarman, supra note 51, at 383; Alan Wertheimer, Jobs, Qualifications, and Preferences, 94 Ethics 99, 101–03 (1983).

  102. [102]. See supra Part I.B.2 (exploring the effect of stigma on decision makers).

  103. [103]. The role of business image in antidiscrimination law is discussed at length later in this Article. See infra Part II.A.

  104. [104]. Berman & Crane, supra note 42, at 1656 (“[I]ntangible costs associated with a smoker’s personal presentation to customers or the public, especially in health-related industries.”).

  105. [105]. Lee, supra note 1.

  106. [106]. Jacobson, supra note 4 (internal quotation marks omitted).

  107. [107]. Lee, supra note 1 (internal quotation marks omitted) (quoting Becky Hall, Vice President of Health and Wellness at Baylor Health Care System).

  108. [108]. Jacobson, supra note 4 (internal quotation marks omitted) (quoting Joel Allison, President & CEO of Baylor).

  109. [109]. Id. (“In recent years, many hospitals and health care-related enterprises have adopted the nicotine-free policy, including Memorial Hermann in Houston and the Cleveland Clinic in Ohio.”).

  110. [110]. Id.

  111. [111]. Ramshaw, supra note 13 (quoting the CMC policy) (internal quotation marks omitted).

  112. [112]. Id. (internal quotation marks omitted).

  113. [113]. Pradia, supra note 16 (citing nutritionist Tim Holcomb).

  114. [114]. Such restrictions were the subject of infamous Title VII cases. See, e.g., Frank v. United Airlines, Inc., 216 F.3d 845, 847 (9th Cir. 2000); Gerdom v. Cont’l Airlines, Inc., 692 F.2d 602, 603 (9th Cir. 1982); EEOC v. Delta Air Lines, 441 F. Supp. 626 (S.D. Tex. 1977), rev’d on other grounds, 619 F.2d 81 (5th Cir. 1980).

  115. [115]. Berman & Crane, supra note 42, at 1656 & n.30.

  116. [116]. Edgar H. Schein, The Corporate Culture Survival Guide 54 (2009) (describing the role of organizational culture in defining group boundaries).

  117. [117]. Jeanette W. Gisdorf & Fraya Wagner-Marsh, Organizational Culture, in Encyclopedia of Management 624 (Marilyn M. Helms ed., 5 ed. 2006) (“A strong [organizational] culture tends to increase behavioral consistency and reduce turnover.”).

  118. [118]. Id.

  119. [119]. Sugarman, supra note 51, at 383.

  120. [120]. Id. (noting that “the personal values of upper management may also be at stake in employment decisions” especially in the case of “small or family-owned firms and non-profit groups”); see also Berman & Crane, supra note 42, at 1652 (explaining that advocates of bans on tobacco use believe that “such policies, rather than injure smokers by infringing on their rights, help them by encouraging them to quit”). But see Henderson, supra note 27, at 1543 (asserting that at present, corporate nanny activities “are focused only on cost internalization”); Maltby, supra note 19, at 1641 (stating that “[e]mployers don’t ban off-duty smoking because they are anti-smoking; they ban off-duty smoking to increase the bottom line”).

  121. [121]. Henderson, supra note 27, at 1523 (defining “paternalism”); see also Jeremy A. Blumenthal & Peter H. Huang, Positive Paternalism, Nat’l L.J., Jan. 26, 2009 (distinguishing between positive and negative paternalistic measures).

  122. [122]. Paternalistic policies may be designed to prohibit the current self from making decisions, such as smoking, that will harm the future self. Henderson, supra note 27, at 1524. Harms that occur solely within the individual are called “internalities.” Id. As Henderson notes, “internality-based arguments are often simply cloaks for pure paternalism.” Id. at 1526.

  123. [123]. See, e.g., Berman & Crane, supra note 42, at 1670 (asserting that “[f]ar from discriminating against employees who may face higher health costs, these employers have actively sought to help them reduce their health risks”).

  124. [124]. This type of paternalism is called “pure paternalism.” Henderson, supra note 27, at 1523 (internal quotation marks omitted).

  125. [125]. See id. at 1526; see also id. at 1528 (explaining that “pure paternalism is generally disfavored in a free society”). These motivations, therefore, are called “‘impure paternalism’ since the regulator is motivated in part by a consideration of the well-being of the individual being regulated (that is, pure paternalism) and in part by a consideration of the costs the person being regulated is imposing on others (that is, cost internalization).” Id. at 1526.

  126. [126]. See, e.g., Berman & Crane, supra note 42, at 1653 (explaining that the CEO of Scotts “cited the rising cost of healthcare coverage and the desire to have a healthy workforce as reasons for the tobacco-free workforce policy”).

  127. [127]. See Conlin, supra note 32. Thus, while Hagedorn explicitly cited cost concerns as driving the policy, he also expressed a desire to “help” his employees. Id.

  128. [128]. See The Oxford English Dictionary 757–58 (2d ed. 1989) (defining “to discriminate” as “to divide, separate, distinguish”); see also Robert K. Fullinwider, The Reverse Discrimination Controversy: A Moral and Legal Analysis 10–11 (1980) (noting that the term “discrimination” is “morally neutral”).

  129. [129]. “Discrimination” is thus essential to the hiring process. See Sugarman, supra note 51, at 395 (noting that in assessing applicants “employers need to have some ways (formal and/or informal) to winnow the prospective employees in order to make a final selection”). When the applicant pool is large enough, employers may adopt generalized hiring policies to efficiently screen applicants, even if those rules ultimately exclude capable potential employees. Id. at 396. Some of these crude screens may be related to health. See, e.g., id. at 397 (explaining that “the employer may find it efficient to exclude everyone who appears, based upon some rule-of-thumb, to present a substantial risk of significantly higher than average costs of health insurance, sick leave or worker’s compensation claims”).

  130. [130]. See Fullinwider, supra note 128, at 11.

  131. [131]. See id. (“For some, it may be enough that a practice is called discriminatory for them to judge it wrong.”); Sugarman, supra note 51, at 415 (emphasis added) (describing the “most important” aspect of establishing the need for a lifestyle discrimination statute “is to show that this sort of discrimination is unfair for the same sorts of underlying reasons that make unfair discrimination on the basis of race, sex, etc.”).

  132. [132]. See Fullinwider, supra note 128, at 11 (explaining that while “[t]he dictionary sense of ‘discrimination’ is neutral[,] . . . the current political use of the term is frequently non-neutral, perjorative [sic]”).

  133. [133]. See Elizabeth Pendo, Working Sick: Lessons of Chronic Illness for Health Care Reform, 9 Yale J. Health Pol’y L. & Ethics 453 (2009) (discussing differences in employer-provided health care for chronically ill employees); see also Mary Crossley, Discrimination Against the Unhealthy in Health Insurance, 54 U. Kan. L. Rev. 73 (2005) (arguing that risk-shifting in the context of health insurance disproportionately burdens unhealthy individuals); Jessica L. Roberts, “Healthism”: A Critique of the Antidiscrimination Approach to Health Insurance and Health-Care Reform, 2012 U. Ill. L. Rev. 1159.

  134. [134]. Jessica L. Roberts, Preempting Discrimination: Lessons from the Genetic Information Nondiscrimination Act, 63 Vand. L. Rev. 439 (2010) (describing the reasons behind the passing of the Genetic Information Nondiscrimination Act).

  135. [135]. Roberts, supra note 133 (describing discrimination on the basis of health-status in insurance as a justification for the Affordable Care Act).

  136. [136]. Id. at 1171 (defining “healthism” as “discrimination on the basis of health status”).

  137. [137]. Elizabeth Mendes, Americans Don’t Want Biases in Hiring Smokers, the Overweight, Gallup (July 22, 2011),

  138. [138]. Lee, supra note 1 (internal quotation marks omitted).

  139. [139]. Id. (quoting Dotty Griffith) (internal quotation marks omitted).

  140. [140]. Privacy Line, supra note 4.

  141. [141]. Id.

  142. [142]. Id.

  143. [143]. See Lecker, supra note 78, at 52 (“The 95% of those who believed their employers should not discriminate against smoking cited two reasons. The first was that it is not illegal to be a smoker . . . . The second reason given was that smoking is a personal right and it was discriminatory otherwise.”).

  144. [144]. Privacy Line, supra note 4.

  145. [145]. Id.; see also Jacobson, supra note 4 (proposing that employers might also someday ban hiring overweight applicants). Dotty Griffith of the Texas ACLU asked a similar question with respect to Methodist’s hiring ban: “At what point do you give up your rights and autonomy? Will they not employ those who ride motorcycles and drink alcohol?” Lee, supra note 1 (internal quotation marks omitted).

  146. [146]. And they are probably right. Texas currently has no state lifestyle discrimination statute. For a discussion of state lifestyle discrimination statutes, see infra Part III.A.

  147. [147]. Pradia, supra note 16 (emphasis added) (internal quotation marks omitted) (quoting Alex Hernandez).

  148. [148]. Sugarman, supra note 51, at 433 (explaining that “citizens support antidiscrimination principles primarily because they believe such rules are just”).

  149. [149]. 42 U.S.C. § 12112(a) (2006 & Supp. V 2011).

  150. [150]. 29 U.S.C. § 794(a) (2006) (“No otherwise qualified individual with a disability . . . shall, solely by reason of . . . disability, be excluded from the participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance . . . .”).

  151. [151]. 42 U.S.C. § 12102(1) (defining disability in the ADA); see also 29 U.S.C. § 705(20)(a) (2006 & Supp. V 2011) (defining disability in the Rehabilitation Act).

  152. [152]. See Jessica L. Roberts, Health Law as Disability Rights Law, 97 Minn. L. Rev. 1963, 1977 (2013) (“The concepts of ‘health’ and ‘disability’ have a complicated, long-standing relationship.”).

  153. [153]. See Bradley A. Areheart, Disability Trouble, 29 Yale L. & Pol’y Rev. 347, 362­–70 (2011); see also Ramona L. Paetzold, Why Incorporate Disability Studies into Teaching Discrimination Law?, 27 J. Legal Stud. Educ. 61, 74–76 (2010).

  154. [154]. For a discussion of the health disparities faced by people with disabilities, see Nat’l Council on Disability, The Current State of Health Care for People with Disabilities 12-14 (2009).

  155. [155]. See 29 C.F.R. § 1630.2(j)(3)(iii) (2013) (certain impairments “substantially limit . . . major life activities,” including cancer which “substantially limits normal cell growth”).

  156. [156]. Notably, the American Medical Association (“AMA”) recently classified obesity as a disease. While that decision carries no legal authority, it does represent a willingness to treat obesity, in and of itself, as a medical condition. See Andrew Pollack, AMA Recognizes Obesity as a Disease, N.Y. Times (June 18, 2013),

  157. [157]. Cook v. R.I. Dep’t of Mental Health, Retardation, & Hosps., 10 F.3d 17 (1st Cir. 1993).

  158. [158]. Id. at 20–21.

  159. [159]. Id.

  160. [160]. Id. at 25.

  161. [161]. Id. at 24.

  162. [162]. Id. at 23. (“Although the jury did not return a special finding as to whether plaintiff actually had a cognizable impairment, or was merely regarded by MHRH as having one, the district court, without objection, charged in the alternative; hence, plaintiff is entitled to prevail on this appeal so long as the evidence supports recovery under one of these theories. In this instance, we believe the record comfortably justifies either finding.”).

  163. [163]. In Sutton v. United Air Lines, Inc., 527 U.S. 471, 482, 489 (1999), the Court held that determinations of who qualifies as a person with a disability should consider mitigating measures when assessing the extent an impairment limits functioning under the first prong, and that a finding of disability under the third prong required a mistaken belief on the part of the employer that either the claimant had a substantially limiting impairment when she in fact did not or that a non-disabling impairment is substantially limiting. The Court further retracted the ADA’s definition of disability in Toyota Motor Manufacturing, Kentucky, Inc. v. Williams, 534 U.S. 184, 200-01 (2002), holding that to qualify as individuals with disabilities, Title I claimants must demonstrate they are severely restricted from performing activities that are central to most people’s daily lives, not merely a task, or class of tasks, that are required for the claimant’s job.

  164. [164]. Abigail Kozel, Note, Large and in Charge of Their Employment Discrimination Destiny: Whether Obese Americans Now Qualify as Disabled Under the Americans with Disability Act Amendments Act of 2008, 31 Hamline J. Pub. L. & Pol’y 273, 290 (2009) (“Following a similar timeline as general ADA claims, obesity-as-a-disability claims under the ADA began with broad coverage, but quickly limited any hope for obese plaintiffs.”).

  165. [165]. See Smaw v. Va. Dep’t of State Police, 862 F. Supp. 1469, 1474 (E.D. Va. 1994).

  166. [166]. Id. at 1472–73.

  167. [167]. Andrews v. Ohio, 104 F.3d 803, 810 (6th Cir. 1997).

  168. [168]. See EEOC v. Watkins Motor Lines, Inc., 463 F.3d 436, 440 (6th Cir. 2006).

  169. [169]. Id. at 443.

  170. [170]. ADA Amendments Act of 2008, Pub. L. No. 110-325, 122 Stat. 3553 (codified as amended at 42 U.S.C. 12101 (Supp. V 2011)).

  171. [171]. Id. § 2(a)(5).

  172. [172]. Id. § 2(b)(3).

  173. [173]. See Kozel, supra note 164, at 327 (stating that “with the passage of the ADAAA, certain claims of obesity-as-a-disability will likely be successful in the future”). But see Lescoe v. Pa. Dep’t of Corr.-SCI Fracksville, 464 F. App’x 50, 53–54 (3d Cir. 2012) (finding the plaintiff failed to establish his morbid obesity under both the first and third prongs of the definition of disability); Sibilla v. Follett Corp., No. CV 10-1457(AKT), 2012 WL 1077655, at *9 (E.D.N.Y. Mar. 30, 2012) (finding that the claimant’s obesity was not a disability under the regarded as prong because claimant failed to establish that her employer perceived her to have an impairment).

  174. [174]. See Maltby & Dushman, supra note 51, at 651–52 (discussing the difficulties of using the ADA to prohibit bans on hiring smokers).

  175. [175]. See Brashear v. Simms, 138 F. Supp. 2d 693, 694–95 (D. Md. 2001) (finding that “common sense compels the conclusion that smoking, whether denominated as ‘nicotine addiction’ or not, is not a ‘disability’ within the meaning of the ADA”).

  176. [176]. For example, courts have held that alcoholism may be an impairment, although not a per se disability. See, e.g., Burch v. Coca-Cola Co., 119 F.3d 305, 317­–18 (5th Cir. 1997).

  177. [177]. See Michael Moore, ADA Amendments May Open the Door for Nicotine Addiction Claims, Pa. Lab. & Emp. Blog (Oct. 29, 2008),
    articles/discrimination-harassment/ada-amendments-may-open-the-door-for-nicotine-addiction-claims/ (explaining that pursuant to the amended ADA “employers may well be required to reasonably accommodate nicotine-addicted employees much as they would need to do so with other addictions”).

  178. [178]. Again, the ADA most assuredly covers impairments that result from nicotine use and addiction. The question, however, remains whether nicotine addiction qua nicotine addiction may independently qualify for protection.

  179. [179]. Courts have however answered this question in the negative. See, e.g., Brashear, 138 F. Supp. 2d at 695.

  180. [180]. See, e.g., Chadwick, supra note 77, at 125 (“Although smoking may lead to eventual disability, it is not, in and of itself, a disability because it does not limit major life activities as required under the ADA.”); Valleau, supra note 42, at 475–78 (explaining why nicotine addiction does not qualify as a disability under the pre-Amendments Act ADA); Daniel M. Warner, “We Do Not Hire Smokers”: May Employers Discriminate Against Smokers?, 7 Emp. Responsibilities & Rts. J. 129, 132–37 (1994) (asserting that pursuant to the pre-Amendments Act ADA smoking does not qualify as a disability under both the first and third prongs of the definition); Mark W. Pugsley, Note, Nonsmoking Hiring Policies: Examining the Status of Smokers Under Title I of the Americans with Disabilities Act of 1990, 43 Duke L.J. 1089, 1090 (1994) (arguing that the pre-Amendments Act ADA does not cover nicotine addiction and, as a result, the ADA does not prohibit hiring bans based on tobacco use).

  181. [181]. See Jon Hyman, Smoking as a Disability Redux, Ohio Employer’s L. Blog (Nov. 16, 2009), (“[T]he ‘regarded as’ prong of the new ADA is sufficiently broad to possibly encompass actions taken against employees pursuant to employer anti-smoking policies.”).

  182. [182]. See Mitchell H. Rubinstein, Is Smoking a Disability Under the ADA as Amended??, Adjunct L. Prof Blog (Nov. 23, 2009), (“I would be surprised if smoking per se (as opposed to smoke related illnesses) are protected under the ADA.”). But see Hyman, supra note 181 (“Employees can claim that anti-smoking policies violate the ADA. Addiction is a protected disability. Diseases related to or caused by smoking (cancers, lung diseases, asthma, and other respiratory conditions, for example) are also protected disabilities. Employees will claim that an adverse action taken pursuant to an anti-smoking policy is being taken because the employer regards the employee as disabled. Adverse actions taken against employees because of smoking should now be viewed as high risk, at least until courts begin weighing in on this controversial issue.”).

  183. [183]. The law defines genetic information as “(i) [an] individual’s genetic tests, (ii) the genetic tests of family members of such individual, and (iii) the manifestation of a disease or disorder in family members of such individual.” Genetic Information Nondiscrimination Act of 2008 § 201(4)(A), 42 U.S.C. § 2000ff(4)(A) (Supp. V 2011).

  184. [184]. Pursuant to Title II of GINA, employers may not use genetic information “to limit, segregate, or classify the employees of the employer in any way that would deprive or tend to deprive any employee of employment opportunities or otherwise adversely affect the status of the employee as an employee.” Id. § 2000ff-1(a)(2).

  185. [185]. See D. Wendy Greene & Jessica L. Roberts, Who Are We in the Workplace?: How Title VII and GINA Combat DNA-Based Race Discrimination (forthcoming) (on file with authors).

  186. [186]. For possible explanations for why Congress passed GINA, see Roberts, supra note 134.

  187. [187]. See Roberts, supra note 133 (outlining the traditional risk assessment strategies associated with insurance).

  188. [188]. See Kristin M. Madison et al., The Law, Policy, and Ethics of Employers’ Use of Financial Incentives to Improve Health, 39 J.L. Med. & Ethics 450, 451 (2011); Mello & Rosenthal, supra note 61, at 192. As noted, Baylor has begun charging its employees who smoke an insurance surcharge, see supra text accompanying note 4, and Scotts forced employees who did not participate in the wellness program, or who failed to meet their fitness goals, to pay heightened premiums, see supra text accompanying note 35. For additional examples, see Henderson, supra note 27, at 1546–52, which describes numerous employer wellness programs consisting of both positive and negative incentive structures.

  189. [189]. But see Jennifer S. Bard, When Public Health and Genetic Privacy Collide: Positive and Normative Theories Explaining How ACA’s Expansion of Corporate Wellness Programs Conflicts with GINA’s Privacy Rules, 39 J.L. Med. & Ethics 469, 474–75 (2011) (explaining that wellness programs may simply shift costs from employers to the public and may not have a measurable impact on participant health).

  190. [190]. Mello & Rosenthal, supra note 61, at 192; Harald Schmidt et al., Carrots, Sticks, and Health Care Reform—Problems with Wellness Incentives, 362 New Eng. J. Med. e3(1), e3(1) (2010) (differentiating between “carrots” and “sticks” in employer wellness programs and demonstrating how the two categories blend when actually implemented).

  191. [191]. Mello & Rosenthal, supra note 61, at 194.

  192. [192]. Id. at 194–95 (internal quotation marks omitted).

  193. [193]. Howard K. Koh & Kathleen G. Sebelius, Promoting Prevention Through the Affordable Care Act, 363 New Eng. J. Med. 1296, 1296 (2010) (describing the ACA’s “vibrant emphasis on disease prevention”); see also Bard, supra note 189, at 470–71 (outlining the ACA’s emphasis on prevention).

  194. [194]. Koh & Sebelius, supra note 193, at 1297 (“Since tobacco dependence and obesity represent substantial health threats, the Act addresses these specific challenges in a number of ways.”).

  195. [195]. See 29 U.S.C. § 1182 (a)(1)(A)–(H) (2006) (amending ERISA).

    Subject to paragraph (2), a group health plan, and a health insurance issuer offering group health insurance coverage in connection with a group health plan, may not establish rules for eligibility (including continued eligibility) of any individual to enroll under the terms of the plan based on any of the following health status-related factors in relation to the individual or a dependent of the individual: (A) Health status. (B) Medical condition (including both physical and mental illnesses). (C) Claims experience. (D) Receipt of health care. (E) Medical history. (F) Genetic information. (G) Evidence of insurability (including conditions arising out of acts of domestic violence). (H) Disability.

    Id.; see also 42 U.S.C. § 300gg-1(a)-(b) (2006 & Supp. V 2011) (amending the Public Health Service Act (“PHSA”)).


  196. [196]. Mello & Rosenthal, supra note 61, at 193-94 (explaining that while HIPAA prohibits insurers from discriminating against individuals within groups, insurers may still make group decisions based on health-related factors, such as charging a higher risk group an increased group premium or choosing not to cover certain health conditions).

  197. [197]. Madison et al., supra note 188, at 462 (explaining that while HIPAA’s nondiscrimination provisions preclude differentiation on the basis of certain health-related factors “[t]hey carve out an exception, however, for wellness programs designed to promote health or prevent disease”); Schmidt et al., supra note 190, at e3(1) (noting that while HIPAA prevents group health insurers from discriminating against individual group members when setting premiums it “does not prevent insurers from offering reimbursements through wellness programs” (internal quotation marks omitted)); see also Madison et al., supra note 188, at 462 (explaining that “[t]he ACA provisions are patterned after the HIPAA regulations”).

  198. [198]. Koh and Sebelius, supra note 193, at 1297 (noting that the law provides grants for small businesses to institute wellness programs and requires federal support for existing health and wellness programs); Madison et al., supra note 188, at 451 (explaining that the ACA “both reflects and promotes growing interest in employer incentive programs”).

  199. [199]. Participation incentives include reimbursements for gym memberships or smoking cessation programs, rewards for diagnostic testing (regardless of result), or waiving co-payments for seeking certain kinds of medical services, like prenatal care.

  200. [200]. Attainment incentives include rewards for losing a specified amount of weight or achieving a particular BMI, lowering cholesterol a specified amount, or quitting smoking.

  201. [201]. Mello & Rosenthal, supra note 61, at 193 (“HIPAA makes it easy for health plans to reward members for participating in health-promotion programs but difficult to reward them for achieving a particular health standard.”).

  202. [202]. See 26 C.F.R. § 54.9802-1(f)(1) (2013); 29 C.F.R. § 2590.702(f)(1) (2012); 45 C.F.R. § 146.121(f)(1) (2011). The ACA effectively codifies HIPAA’s wellness program regulations. See Patient Protection and Affordable Care Act, Pub. L. No. 111-148, 124 Stat. 119, § 1201 (2010) (codified as amended at 42 U.S.C. § 300gg (Supp. V 2011)) (amending PHSA section 2705(a)(1)–(9) and creating section 2705).

  203. [203]. 29 C.F.R. § 2590.702(f).

  204. [204]. However, the ACA raised the amount of eligible rewards from 20% of the total cost of coverage under HIPAA to 30% of the total cost of coverage. See 26 C.F.R. § 54.9802-1(f)(2); 45 C.F.R. § 146.121(f)(2).

  205. [205]. 78 Fed. Reg. 33,158 (June 3, 2013) (to be codified at 26 C.F.R. pt. 54; 29 C.F.R. pt. 2590; 45 C.F.R. pts. 146-47).

  206. [206]. Madison et al., supra note 188, at 462 (“Health plans are permitted to distinguish between those who are in poor health or who fail to engage in healthy behaviors, and those who are healthier, in order to develop programs aimed at improving health.”).

  207. [207]. See 45 C.F.R. § 146.121(f)(3).

  208. [208]. Mello & Rosenthal, supra note 61, at 194 (“It is unclear whether obesity would be considered a medical condition that makes it unreasonably difficult to achieve a weight goal, but a reasonable inference from other examples given in the rules is that morbid obesity could well be considered as such.”).

  209. [209]. See Roberts, supra note 133, for a critique of the antidiscrimination framing of the Affordable Care Act.

  210. [210]. Patient Protection and Affordable Care Act, Pub. L. No. 111-148, 124 Stat. 119, § 1201 (2010) (codified as amended at 42 U.S.C. § 300gg (Supp. V 2011)) (“A group health plan and a health insurance issuer offering group or individual health insurance coverage may not impose any preexisting condition exclusion with respect to such plan or coverage.”).

  211. [211]. Id. (amending PHSA section 2701(a)).

  212. [212]. Id.

  213. [213]. See Sugarman, supra note 51, at 414 (stating that legal protection does not turn on how an individual became disabled—even if the disability was the result of an irresponsible choice—and explaining that in the case of lifestyle discrimination an individual has an ongoing choice, whereas with disability she cannot go back).

  214. [214]. The trait–conduct distinction, while useful, may break down in certain contexts. See Jackson, supra note 20, at 167 (noting that “many activities can be characterized as either an ‘activity’ or a ‘status’”). For example, anorexia, a psychiatric diagnosis and hence an impairment that may form the basis of a disability, could consequently be considered a trait. However, thinness could likewise be understood as rooted in conduct because the individual must behave in a particular way to maintain her weight.

  215. [215]. See Sugarman, supra note 51, at 414–15 (explaining the status–choice divide).

  216. [216]. See, e.g., Robert M. Howie & Laurence A. Shapero, Lifestyle Discrimination Statutes: A Dangerous Erosion of At-Will Employment, a Passing Fad, or Both?, 31 Emp. Rel. L.J. 21, 21 (2005) (explaining that existing federal antidiscrimination statutes “reflect judgments by our federal government that it is not in our best interest as a society to tolerate employment discrimination against individuals because of . . . characteristics such as race or national origin over which they have no control”); Sugarman, supra note 51, at 414 (noting that “race, sex, age and ancestry/national origin are generally understood to be matters of status, that is, a characteristic that one cannot avoid”).

  217. [217]. Of course, the most glaring exception to this generalized statement is the widely recognized antidiscrimination protection of religion. Sugarman, supra note 51, at 414–15 (explaining that “[m]any people would say that one can, at any time, abandon one’s religious belief, or adopt a new one” but noting that religion also has status elements imposed by parents, the use of labels, and ancestry).

  218. [218]. Rogers v. Am. Airlines Inc., 527 F. Supp. 229, 232 (S.D.N.Y. 1981).

  219. [219]. Id. at 233.

  220. [220]. For example, 80% of survey respondents indicated they believed that significant overweight/obesity is the result of personal lifestyle choices. Newport, supra note 94.

  221. [221]. Major, supra note 82.

  222. [222]. Insurance can be understood as a risk pooling agreement to indemnify against future loss. See Dictionary of Health Insurance and Managed Care 138 (David Edward Marcinko & Hope Rachel Hetico eds., 2006) (defining “health insurance” as “[c]overage that provides for the payments of benefits as a result of sickness or injury”); see also James Surowiecki, Fifth Wheel, New Yorker (Jan. 4, 2010),
    ta_talk_surowiecki (describing risk-based pricing as “the raison d’être of a health-insurance company”).

  223. [223]. See Dictionary of Health Insurance and Managed Care, supra note 222, at 12–13 (defining “adverse selection”).

  224. [224]. Id. at 188 (defining “moral hazard”).

  225. [225]. See Mark V. Pauly, The Economics of Moral Hazard: Comment, 58 Am. Econ. Rev. 531 (1968).

  226. [226]. See Howie & Shapero, supra note 216, at 31–32.

  227. [227]. The number of states with lifestyle discrimination statutes varies depending upon how broadly that term is defined. See Maltby, supra note 19, at 1647 n.43 (stating that thirty states and the District of Columbia have statutes preventing termination for legal activities).

  228. [228]. The word “lifestyle” is frequently associated with voluntariness, such as the notion of “lifestyle choices.”

  229. [229]. See Henderson, supra note 27, at 1591–92 (identifying two types of state lifestyle discrimination statutes: “(1) those protecting use of lawful products or engaging in legal activities, (2) those protecting tobacco or alcohol [use]”).

  230. [230]. See Cal. Lab. Code §§ 96(k), 98.6 (West 2011); Colo. Rev. Stat. § 24-34-402.5 (2012); 820 Ill. Comp. Stat. Ann. 55/5 (West 2012); Minn. Stat. § 181.938 (2012); Mont. Code Ann. § 39-2-313 (2011); Nev. Rev. Stat. § 613.333 (2011); N.Y. Lab. Law § 201-d (McKinney 2009); N.C. Gen. Stat. § 95-28.2 (2011); N.D. Cent. Code § 14-02.4-03 (2009 & Supp. 2012); Tenn. Code Ann. § 50-1-304 (West 2013); Wis. Stat. § 111.321 (2012).

  231. [231]. Conn. Gen. Stat. § 31-40s (2013); D.C. Code § 7-1703.03 (2008); Ind. Code Ann. § 22-5-4-1 (West 2005 & Supp. 2012); Ky. Rev. Stat. Ann. § 344.040 (West 2011 & Supp. 2012); La. Rev. Stat. Ann. § 23:966 (2010); Me. Rev. Stat. Ann. tit. 26, § 597 (2007); N.H. Rev. Stat. Ann. § 275:37-a (2010); N.J. Stat. Ann. § 34:6B­–1 (West 2011); N.M. Stat. Ann. § 50-11-3 (2007); Okla. Stat. Ann. tit. 40, § 500 (West 2013); R.I. Gen. Laws § 23-20.10-14 (2008); S.C. Code Ann. § 41-1-85 (2012); S.D. Codified Laws § 60-4-11 (2009); Va. Code. Ann. § 2.2-2902 (2011); W. Va. Code Ann. § 21-3-19 (LexisNexis 2008); Wyo. Stat. Ann. § 27-9-105 (2013).

  232. [232]. Minn. Stat. § 181.938(2).

  233. [233]. Mont. Code Ann. § 39-2-313(1).

  234. [234]. See Ramshaw, supra note 13 (noting Michigan’s law, as well similar ordinances in six U.S. cities).

  235. [235]. See, e.g., Conn. Gen. Stat. § 31-40s(a) (exempting “any nonprofit organization or corporation whose primary purpose is to discourage use of tobacco products by the general public”); 820 Ill. Comp. Stat. Ann. 55/5(b) (exempting “any employer that is a non-profit organization that, as one of its primary purposes or objectives, discourages the use of one or more lawful products by the general public”); Mo. Rev. Stat. § 290.145 (2012) (exempting “not-for-profit organizations whose principal business is health care promotion”); Mont. Code Ann. § 39-2-313(3)(c) (exempting “a nonprofit organization that, as one of its primary purposes or objectives, discourages the use of one or more lawful products by the general public”).

  236. [236]. See, e.g., Cal. Lab. Code § 96(k) (West 2011) (allowing “[c]laims for loss of wages as the result of demotion, suspension, or discharge from employment”); La. Rev. Stat. Ann. §23:966(A) (protecting “an individual, during the course of employment”). Colorado’s, South Dakota’s, and Tennessee’s statutes only protect current employees from termination. See Colo. Rev. Stat. § 24-34-402.5(1) (2012) (“It shall be a discriminatory or unfair employment practice for an employer to terminate the employment of any employee due to that employee’s engaging in any lawful activity off the premises of the employer during nonworking hours . . . .”); S.D. Codified Laws § 60-4-11 (“It is a discriminatory or unfair employment practice for an employer to terminate the employment of an employee due to that employee’s engaging in any use of tobacco products off the premises of the employer during nonworking hours . . . .”); Tenn. Code Ann. § 50-1-304 (West 2013) (prohibiting “[d]ischarge for refusal to participate in or remain silent about illegal activities, or for legal use of agricultural product”).

  237. [237]. See, e.g., Colo. Rev. Stat. § 24-34-402.5(1)(a); D.C. Code § 7-1703.03(a) (2008); Minn. Stat. § 181.938(3)(a)(1); Mont. Code Ann. § 39-2-313(3)(a)(ii); N.M. Stat. Ann. §50-11-3(B)(2) (2007); N.C. Gen. Stat. § 95-28.2(c)(1) (2011); S.D. Codified Laws § 60-4-11(1); Wis. Stat. § 111.35(2)(c) (2012); Wyo. Stat. Ann. § 27-9-105(iv) (2013).

  238. [238]. Mo. Rev. Stat. § 290.145.

  239. [239]. See Cal. Lab. Code § 96(k).

  240. [240]. See Grinzi v. San Diego Hospice Corp., 14 Cal. Rptr. 3d 893, 896, 901 (Cal. Ct. App. 2004); Barbee v. Household Auto. Fin. Corp., 6 Cal. Rptr. 3d 406, 412 (Cal. Ct. App. 2003).

  241. [241]. Samuel D. Warren & Louis D. Brandeis, The Right to Privacy, 4 Harv. L. Rev. 193, 193 (1890).

  242. [242]. See Sissela Bok, Secrets: On the Ethics of Concealment and Revelation 20 (1983) (explaining the importance of secret-keeping to communal living).

  243. [243]. See John Christman, Autonomy in Moral and Political Philosophy, Stan. Encyclopedia Phil. (Aug. 11, 2009), (defining “individual autonomy”).

  244. [244]. Americans have a strong desire to be treated as individuals, which translates into a strong desire for autonomy and individualized treatment. See Sugarman, supra note 51, at 406.

  245. [245]. See Christman, supra note 243 (“Autonomy is the aspect of persons that undue paternalism offends against. . . . Autonomy is the ability to so decide, so paternalism involves a lack of respect for autonomy.”).

  246. [246]. Numerous areas of law support the rights to privacy and autonomy, including tort, constitutional, and criminal law. See Sugarman, supra note 51, at 402–04 (but also noting that “the privacy claim of employees about their off-work behavior is not exactly analogous to the privacy rights just discussed”). But see Rives, supra note 31, at 556 (identifying two torts related to the invasion of employee privacy: “(1) unreasonable intrusion upon the seclusion of another and (2) unreasonable publicity given to another’s private life”).

  247. [247]. Torts scholar William Prosser famously expounded on the meaning of privacy in tort law in his article Privacy. William L. Prosser, Privacy, 48 Calif. L. Rev. 383, 389 (1960) (expounding on the meaning of privacy in tort law).

  248. [248]. See Daniel J. Solove, A Taxonomy of Privacy, 154 U. Pa. L. Rev. 477, 524–25, 527 (2006) (exploring the tort of breach of confidentiality as an issue of privacy).

  249. [249]. See id. at 495–97, 526–27 (exploring the Fourth Amendment notion of “a reasonable expectation of privacy”).

  250. [250]. See id. at 557–58 (describing procreative autonomy as an issue of privacy).

  251. [251]. See id. at 561–62 (describing decisions related to consensual adult sex as issues of privacy).

  252. [252]. See Al Gini, Work, Identity and Self: How We Are Formed by the Work We Do, 17 J. Bus. Ethics 707, 707–08 (1998) (“As adults there is nothing that more preoccupies our lives [than work]. . . . It is in work that we become persons.”).

  253. [253]. See Sugarman, supra note 51, at 405 (“[A]t any one time [a person’s] most valued asset is, in effect, [her] job. If, however, employers attach conditions to jobs that restrict personal autonomy, [employee] privacy is as restricted as it would be were the conditions attached by government.”). But see Henderson, supra note 27 (arguing that corporations make better nannies than governments); Sonne, supra 19, at 137 (“Despite analogies to, and lessons drawn from, state power and its limits, the voluntary nature of the private work relationship must be distinguished from the coercive arm of the state when discussing these issues.”).

  254. [254]. See Sugarman, supra note 51, at 405 (“[An employee is] put to the unwelcome choice of surrendering [her] job or surrendering part of [her] identity.”).

  255. [255]. See Henderson, supra note 27, at 1530 (explaining but not adopting the position that “[s]moking and drinking after hours . . . may raise costs, in terms of risk and health insurance outlays, but there is an intuition that the restrictions on private behavior are different in some way—more out of bounds”). These criticisms translate into strong normative claims about the rightness and wrongness of such employment practices. See, e.g., Maltby, supra note 19, at 1639 (describing Henry Ford’s regulation of his employees’ off-duty conduct as “wrong” and stating that “[t]he fact that Henry Ford signed people’s paychecks did not give him the right to control their private lives”).

  256. [256]. See, e.g., Maltby, supra note 19, at 1643 (arguing that “[e]mployment decisions should be based on how well you do your job, not on your private life”). Some courts have likewise taken this position, striking down conduct-based employment policies as too tenuously related to on-the-job performance to be justifiable. See, e.g., Borse v. Piece Goods Shop, Inc., 963 F.2d 611, 622 (3d Cir. 1992) (striking a policy for public policy reasons because “there are areas of an employee’s life in which his employer has no legitimate interest” (quoting Geary v. U.S. Steel Corp., 319 A.2d 174, 180 (Pa. 1974))). Interestingly, Virginia law includes an exception to the public policy exception for workplace safety.

                    However, as noted in Part I, health-related off-duty conduct does impact an individual’s ability to do her job, such as the lower rates of productivity and higher absenteeism among smokers and obese people. Thus, employers may enjoy a very real interest in their employees’ off-duty conduct.
  257. [257]. See, e.g., Rives, supra note 31, at 554 (framing the issue as “an employee’s right to engage in lawful recreational activities as well as an employee’s right to lawfully use lawful products during nonworking hours”).

  258. [258]. Berman & Crane, supra note 42, at 1652 (“Some [tobacco control advocates] have argued that such policies constitute unethical discrimination that tobacco control advocates should not countenance.”).

  259. [259]. Maltby, supra note 19, at 1641; see also Maltby & Dushman, supra note 51, at 646 (describing the health risks associated with any number of activities, including consuming “[c]affeine, alcohol, red meat, [and] sugar,” as well as hobbies such as skiing and riding motorcycles).

  260. [260]. See Sugarman, supra note 51, at 407 (explaining that a blood specimen taken to screen for nicotine could then be screened for other things like HIV). Although GINA prohibits employers from requesting or requiring genetic information from their employees, the concern remains that employers could engage in this type of intrusive conduct clandestinely.

  261. [261]. See id. at 415 (explaining that “[l]ifestyle discrimination is more about individual autonomy” than status-based discrimination).

  262. [262]. For an in-depth exploration of the role of immutability in employment discrimination law in recent years, see generally Sharona Hoffman, The Importance of Immutability in Employment Discrimination Law, 52 Wm. & Mary L. Rev. 1483 (2011).

  263. [263]. See Anthony P. Goldstone & Philip L. Beales, Genetic Obesity Syndromes, 36 Obesity & Metabolism 37 (2008).

  264. [264]. See Lindsay F. Wiley, Stigma and the Emerging Law of Obesity Control 24–25 (unpublished manuscript) (on file with author) (describing the current environment as ‘obesigenic’).

  265. [265]. Almost 70% of smokers would like to quit. See Ann Malarcher et al., Quitting Smoking Among Adults—United States 2001–2010, 60 Morbidity & Mortality Wkly. Rep. 1513, 1513 (2011), available at

  266. [266]. See Schmidt, Voigt & Emanuel, supra note 58, at 1370 (stating that “smoking is addictive and therefore not completely voluntary” and “[i]t is therefore wrong to treat smoking as something fully under an individual’s control”); see also Sugarman, supra note 51, at 435 (calling this line of argument the “sympathy for the addict” position).

  267. [267]. See, e.g., Wendy K. Mariner, The Affordable Care Act and Health Promotion: The Role of Insurance in Defining Responsibility for Health Risks and Costs, 50 Duq. L. Rev. 271, 301 (2012); Schmidt et al., supra note 190, at e3(1) (citing the failure of diets to achieve long-term weight loss even though dieters want and try to lose weight). Permitting employers to consider health-related factors when implementing wellness programs reintroduces health-based rating into group insurance, thereby undermining the purpose of the nondiscrimination provisions. Hence, Mariner calls for Congress to abandon the wellness program exception. See generally Mariner, supra.

  268. [268]. See Henderson, supra note 27, at 1524–26 (describing internalities).

  269. [269]. Many people who engage in unhealthy behaviors are members of historically disadvantaged groups and encounter barriers to health. For a discussion of structural and institutional barriers to health, see infra Part III.B.1.c.

  270. [270]. See Maltby, supra note 19, at 1645 (“Very few of us are proud of our bad habits. Surveys repeatedly show that most smokers want to quit. Millions of us make New Year’s resolutions to eat less, go to the gym more often, and cut down on our drinking.”).

  271. [271]. See Brian King et al., Vital Signs: Current Cigarette Smoking Among Adults Aged ≥ 18 Years—United States, 2005–2010, 60 Morbidity & Mortality Wkly. Rep. 1207, 1208 (2011), available at (“Among racial/ethnic populations, non-Hispanic American Indians/Alaska Natives had the highest prevalence (31.4%), followed by non-Hispanic whites (21.0%) and non-Hispanic blacks (20.6%).”).

  272. [272]. See Jennifer C. Pierotti, Comment, The “Bottom Line”: A Smokescreen for the Reality that Anti-Tobacco Employment Practices Are Hazardous to Minority Health and Equality, 26 J. Contemp. Health L. & Pol’y 441, 445 (2010) (arguing that employment bans on tobacco use “will disproportionately disadvantage minority groups that have higher rates of smoking and higher incidence of smoking-related disease, in addition to groups that have less support in cessation measures”).

  273. [273]. See Adult Obesity Facts, Ctrs. for Disease Control & Prevention, (last updated Aug. 16, 2013) (“Non-Hispanic blacks have the highest age-adjusted rates of obesity (49.5%) compared with Mexican Americans (40.4%), all Hispanics (39.1%) and non-Hispanic whites (34.3%).” (citing Katherine M. Flegal et al., Prevalence of Obesity and Trends in the Distribution of Body Mass Index Among US Adults, 1999–2010, 307 J. Am. Med. Ass’n 491 (2012)).

  274. [274]. Section (k) codifies disparate impact doctrine and offers this defense. 42 U.S.C. §2000e-2(k) (2006).

  275. [275]. See generally Nat’l Council on Disability, supra note 154 (discussing health disparities between people with disabilities).

  276. [276]. Id.

  277. [277]. See infra note 280.

  278. [278]. For example, one author notes that insurance surcharges and other employer wellness initiatives may unfairly disadvantage lower-income workers: “Direct and indirect increases would disproportionately hurt lower-paid workers, who are generally less healthy than their higher-paid counterparts and thus in greater need of health care, less likely to meet the targets, and least likely to be able to afford higher costs.” Schmidt et al., supra note 190, at e3(3).

  279. [279]. See Madison et al., supra note 188, at 455–58 (describing discrimination and ethical concerns associated with wellness programs).

  280. [280]. Schmidt et al., supra note 190, at e3(2); see also Bard, supra note 189, at 472–74 (outlining ways in which employer-provided wellness programs might be discriminatory).

  281. [281]. Privacy Line, supra note 4 (stating that employers “deserve great latitude in hiring, which makes legislation problematic”).

  282. [282]. The judge in one seminal opinion on the subject opines:

    Trade is free; so is employment. The law leaves employer and employe[e] to make their own contracts; and these, when made, it will enforce; beyond this it does not go. Either the employer or the employe[e] may terminate the relation at will, and the law will not interfere, except for [the] contract broken.

    Rives, supra note 31, at 555 (quoting Payne v. W. & Atl. R.R., 81 Tenn. 507, 519–20 (1884)); see also Howie & Shapero, supra note 216, at 21; Jackson, supra note 20, at 149–50; Lawrence, supra note 32, at 512–14; Sonne, supra note 19, at 154–57; Sugarman, supra note 51, at 424.

  283. [283]. Henderson, supra note 27, at 1591 (asserting that “most states follow the common law employment-at-will regime”).

  284. [284]. Sugarman, supra note 51, at 426.

  285. [285]. Mont. Code Ann. § 39-2-903 (2011) (“‘Good cause’ means reasonable job-related grounds for dismissal based on a failure to satisfactorily perform job duties, disruption of the employer’s operation, or other legitimate business reason. The legal use of a lawful product by an individual off the employer’s premises during nonworking hours is not a legitimate business reason, unless the employer acts within the provisions of 39-2-313(3) or (4).”).

  286. [286]. Unif. Law Comm’rs’ Model Emp’t Termination Act § 3, cmt. a. (1991); see Theodore J. St. Antoine, Labor and Employment Law in Two Transitional Decades, 42 Brandeis L.J. 495, 510–11 (2004).

  287. [287]. See Berman & Crane, supra note 42, at 1660; Howie & Shapero, supra note 216, at 21; Maltby, supra note 19, at 1642; Sonne, supra note 19, at 157–60 (describing antidiscrimination laws and wrongful discharge as exceptions to at-will employment). Additional exceptions to at-will employment include unionized workers, government employees, and high earners with specific contracts. See Sugarman, supra note 51, at 424.

  288. [288]. See Rives, supra note 31, at 555.

  289. [289]. See id. (noting that there are “four public policy exceptions that appear to be the most common: refusing to commit illegal acts, exercising statutory rights, whistleblowing, and performing civic duties”).

  290. [290]. See supra Part I.B.1.

  291. [291]. Henderson, supra note 27, at 1561–62.

  292. [292]. See Samuel R. Bagenstos, The Supreme Court, the Americans with Disabilities Act, and Rational Discrimination, 55 Ala. L. Rev. 923, 925 (2004) (identifying Supreme Court decisions as advancing “the premise that antidiscrimination law—even in its prohibition of intentional discrimination—should prohibit nothing more than irrational discrimination”).

  293. [293]. The federal statutes that prohibit such discrimination are, of course, Title VII, the ADA, the Rehabilitation Act, and GINA.

  294. [294]. See Sugarman, supra note 51, at 432 (noting that “[a] strong civil rights approach to lifestyle discrimination in employment would adopt the general position that employers, rather than employees, must internalize costs that are associated with employee privacy”).

  295. [295]. Henderson, supra note 27, at 1552.

  296. [296]. Sugarman, supra note 51, at 428.

  297. [297]. Id. (“The core idea is that workers who do not like employer intrusions into their private lives should choose to work for an employer who does not seek to control the off-duty behavior that the employee engages in.”).

  298. [298]. Henderson, supra note 27, at 1561 (explaining that a no-smoking policy “may cause the firm’s wages to rise, as the firm may have to pay more to attract workers who will be subject to the policy (and are uniquely valuable to the firm) or who put some value on employers not having these rules,” and that “[t]hese latter employees may be concerned about their externality-causing behaviors being captured by future rules or believe that the nannyism signals something else unattractive about the firm—say, a low tolerance for shirking”).

  299. [299]. Henderson, supra note 27, at 1557 (“So if a firm implements an antismoking policy that raises labor costs without offsetting increases in productivity or reductions in healthcare expenses, and thus puts the firm in worse cost position vis-à-vis competitors, this can be expected to reduce the firm’s financial performance and therefore managers’ compensation.”).

  300. [300]. In fact, I have made a similar point in the context of health insurance. See Roberts, supra note 133, at 1197 (“Characterizing decisions made by health insurers based on health status as ‘discrimination’ may dilute the power of this language when used in its more traditional contexts.”).

  301. [301]. GINA is the primary exception to this general rule. See Roberts, supra note 134, at 441.

  302. [302]. See Sugarman, supra note 51, at 428 (noting “the problem of lifestyle discrimination is not seen as a terribly serious one”); see also Howie & Shapero, supra note 216, at 22 (calling lifestyle discrimination statutes “unnecessary intrusions of the at-will employment relationship without a significant corresponding [public] policy benefit”).

  303. [303]. See, e.g., Lawrence, supra note 32, at 516 (asserting that protecting smokers “elevates cigarette smoking to a civil right”); see also Henderson, supra note 27, at 1610 (“It is equally absurd . . . that legislatures and courts protect something like ‘smokers’ rights’ from the discipline of the market.”).

  304. [304]. Courts have likewise been reluctant to recognize unhealthy conduct as a constitutionally protected right. See, e.g., Grusendorf v. Okla. City, 816 F.2d 539, 543 (10th Cir. 1987) (upholding an Oklahoma City Fire Department policy prohibiting trainees from smoking because smoking is not a fundamental right). In assessing a challenge to the smoking ban, the Tenth Circuit explained: “We need look no further for a legitimate purpose and rational connection than the Surgeon General’s warning on the side of every box of cigarettes sold in this country that cigarette smoking is hazardous to health.” Id.

  305. [305]. See Maltby, supra note 19, at 1647 n.45 (noting Regina Carlson’s argument that lifestyle discrimination protection “would elevate drug addiction to civil rights status, along with race and sex” (quoting Matthew Reilly, Florio Urged to Provide Smokers Bias Protection, Star-Ledger, Jan. 4, 1991) (internal quotation marks omitted)).

  306. [306]. See Berman & Crane, supra note 42, at 1663 (“To elevate the nation’s leading cause of preventable death to the status of a protected civil right is illogical, undermines health education messages, and trivializes the concept of civil rights. Employment-discrimination laws should focus on protecting employees from invidious discrimination based on immutable characteristics or the exercise of constitutionally protected rights. They should not be used as tools to block employers from promoting healthy lifestyle choices.”).

  307. [307]. Blumenthal & Huang, supra note 121.

  308. [308]. Berman & Crane, supra note 42, at 1652 (“[T]obacco-free workforce policies have the potential to dramatically influence general smoking prevalence. This is a case where business interests appear to converge with public health interests.”). But see Schmidt, Voigt & Emanuel, supra note 58, at 1371 (“By cherry-picking ‘low-risk’ employees and denying employment to smokers, employers neglect [their obligation to contribute to public health], risk hurting vulnerable groups, and behave unethically.”).

  309. [309]. Even libertarians, who value autonomy and freedom of choice, recognize the social benefit to limited amounts of paternalistic regulation. See, e.g., Cass R. Sunstein & Richard H. Thaler, Libertarian Paternalism Is Not an Oxymoron, 70 U. Chi. L. Rev. 1159, 1159 (2003) (“The idea of libertarian paternalism might seem to be an oxymoron, but it is both possible and desirable for private and public institutions to influence behavior while also respecting freedom of choice.”).

  310. [310]. For example, Jim Lowe lost 137 pounds in conjunction with Scotts’ initiative. Conlin, supra note 32, at 64–65.

  311. [311]. Id. at 64 (“Gym rats earn special pins they display on ID badge lanyards; these have become a coveted status object. Competition for trips to Hawaii, free massages and facials, and other cash and prizes is fierce. One group of employees started having lunch together every day to keep each other from peeling out of the parking lot for a smoke. Doughnuts have disappeared.”).

  312. [312]. See supra Part III.A.

  313. [313]. Lending consistency to diverse state protections for genetic information was one of the reasons cited in favor of passing GINA. See Gaia Bernstein, The Paradoxes of Technological Diffusion: Genetic Discrimination and Internet Privacy, 39 Conn. L. Rev.241, 288 (2006) (“[T]he failure of the current patchwork of state and federal laws in affecting individuals’ public fears points to the need for a comprehensive federal statute.”).

  314. [314]. Thanks to Dan Schwarcz for this insight.

  315. [315]. See Schmidt, Voigt & Emanuel, supra note 58, at 1370 (stating that nicotine bans “disproportionately and unfairly affect groups that are already burdened by high unemployment rates, poor job prospects, and job insecurity”).

  316. [316]. Historically, in the individual health-insurance market, people at lower levels of health may find themselves uninsured either because of preexisting exclusions or because premiums are too expensive to afford, forcing them to pay out-of-pocket. Likewise, in the group market, people in poorer health still pay more out-of-pocket for health care because of cost-sharing mechanisms, such as deductibles, co-insurance, and co-pays. See generally Crossley, supra note 133, at 74.

  317. [317]. See Sarah B. Schindler, Of Backyard Chickens and Front Yard Gardens: The Conflict Between Local Governments and Locavores, 87 Tul. L. Rev. 231 (2012); see also Tara Parker-Pope, Money is Tight, and Junk Food Beckons, N.Y. Times (Nov. 3, 2008),
    04/health/nutrition/04well.html?_r=2&. But see Mark Bittman, Op-Ed., Is Junk Food Really Cheaper?, N.Y. Times (Sept. 24, 2011),

  318. [318]. See, e.g., Latetia V. Moore et al., Availability of Recreational Resources in Minority and Low Socioeconomic Status Areas, 34 Am. J. Preventive Med. 16, 18 & fig.2 (2008); Lisa M. Powell et al., The Relationship Between Community Physical Activity Settings and Race, Ethnicity and Socioeconomic Status, 1 Evidence-Based Preventive Med. 135, 141 tbl.4, 142 (2004); see also Paul Choitz et al., Urban Fitness Centers: Removing Barriers to Promote Exercise in Underserved Communities, 21 J. Health Care for Poor & Underserved 221, 222–27 (2010) (discussing the impact of opening two fitness centers in a poor community).

  319. [319]. As noted previously, the effectiveness of employer wellness programs has been repeatedly called into question. See, e.g., Horwitz, Kelly & DiNardo, supra note 58; Mariner, supra note 267, at 282–304.

  320. [320]. However, post-ACA, employer-provided health insurance could become less important. See Monahan & Schwarcz, supra note 70 (exploring how the ACA may incentivize employers to exclude high-risk employees from employer-sponsored insurance).

  321. [321]. Americans pay two-and-a-half times more per capita on health care than other developed countries. Jason Kane, Health Costs: How the U.S. Compares with Other Countries, PBS Newshour (Oct. 22, 2012),

  322. [322]. See Catherine Hoffman & Julia Paradise, Health Insurance and Access to Health Care in the United States, 1136 Annals N.Y. Acad. Sci. 149, 149 (2008) (asserting that given the high cost of American health care “health insurance has become nearly essential”).

  323. [323]. See id. at 152 (explaining that “[m]ore than 90% of the uninsured cite cost as the main barrier to getting care (as do more than half the insured)”); see also id. at 153 (stating that “uninsured adults are significantly more likely to delay or forgo care and to have unmet needs than their insured counterparts”).

  324. [324]. Id. at 155.

  325. [325]. See Seth J. Chandler, The Architecture of Contemporary Healthcare Reform and Effective Marginal Tax Rates, 29 Miss. C. L. Rev. 335, 371 (2010) (describing the effect of the low income subsidy on marginal tax rates); see also Daniel N. Shaviro, Emp’t Policies Inst., Effective Marginal Tax Rates on Low-Income Households 3–4 (1999), available at http:// (describing the effect of high marginal tax rates on the near-poor).

  326. [326]. See supra Part III.B.1.c (discussing the health of historically disadvantaged populations).

  327. [327]. See Nat’l Insts. of Health, U.S. Dep’t of Health & Human Servs., NIH Health Disparities Strategic Plan and Budget: Fiscal Years 2009–2013, at 12–14 (2013), available at
    Plan%20and%20Budget%202009-2013.pdf (describing the health disparities faced by racial and ethnic minorities).

  328. [328]. Hoffman & Paradise, supra note 322, at 149.

  329. [329]. Nat’l Council on Disability, supra note 154, at 1 (“People with disabilities frequently lack either health insurance or coverage for necessary services, such as specialty care, long-term services, prescription medications, durable medical equipment, and assistive technologies.”); id. at 11–12 (“The health care system in the United States is complex, highly fragmented, and sometimes overly restrictive in terms of program eligibility. This leaves some people with disabilities with no health care coverage and others with cost-sharing obligations and limits on benefits that prevent them from obtaining health-preserving prescription medications, medical equipment, specialty care, dental and vision care, long-term care, and care coordination.”).

  330. [330]. For an in-depth account of the health disparities facing people with disabilities prior to health reform, see Elizabeth Pendo, Reducing Disparities Through Health Care Reform: Disability and Accessible Medical Equipment, 2010 Utah L. Rev. 1057, 1059–65; Roberts, supra note 152. See generally Nat’l Council on Disability, supra note 154.

  331. [331]. Nat’l Council on Disability, supra note 154, at 9–10, 23; see also id. at 34–35.

  332. [332]. Hoffman & Paradise, supra note 322, at 149.

  333. [333]. Id.

  334. [334]. Id.

  335. [335]. Id.

  336. [336]. See, e.g., Wiley, supra note 264, at 60 (arguing for a destigmatization strategy in obesity regulation and a revival of the interest in antidiscrimination protections).

  337. [337]. Jessica Jackson notes that a law that protects employees but not applicants will just encourage employers to screen at the hiring stage. Jackson, supra note 20, at 168–69.

  338. [338]. Legislation in several states allows insurance adjustments. However, scholars writing in this area have been critical of certain kinds of insurance surcharges based on employee wellness. See, e.g., Berman & Crane, supra note 42, at 1671–73 (critiquing insurance surcharges); Maltby, supra note 19, at 1644 (“To be fair, surcharges should apply to all health-related off-duty behavior.”).

                    Additionally, it is worth noting that while from a practical perspective surcharges and reductions operate in more or less the same manner, sticks are often perceived as communicating negative, stigmatizing attitudes in a way that carrots are not, regardless of the absolute monetary outcomes of the policies themselves.
  339. [339]. Incentives for Nondiscriminatory Wellness Programs in Group Health Plans, 78 Fed. Reg. 33,158, 33,158 (June 3, 2013) (to be codified at 45 C.F.R. pts. 146–147).

  340. [340]. Id. at 33,167.

  341. [341]. See Schmidt et al., supra note 190, at e3(3) (“Attainment incentives that primarily benefit the well-off and healthy should be phased out, and the focus should shift to participation-incentive schemes tailored to the abilities and needs of lower-paid employees.”). For a discussion of those structural barriers, see supra Part III.B.1.c.

  342. [342]. For example, Title I of the ADA protects only “qualified individual[s].” See 42 U.S.C. §12112 (2006).

  343. [343]. See id. § 12112(d)(3); Gary S. Marx, Disability Law Compliance Manual § 3:36 (2012).

  344. [344]. Interestingly, one New England Journal of Medicine article argued that hiring bans on smokers actually runs counter to the mission of health providers. See Schmidt, Voigt & Emanuel, supra note 58, at 1369–70 (calling tobacco bans adopted by health-care providers “paradoxical,” “callous,” and “contradictory” and asserting that “[j]ust as they should treat people regardless of their degree of responsibility for their own ill health, they should not discriminate against qualified job candidates on the basis of health-related behavior”).

  345. [345]. See, e.g., Al Lewis & Vik Khanna, Is It Time to Re-Examine Workplace Wellness ‘Get Well Quick’ Schemes?, Health Aff. Blog (Jan. 16, 2013),
    01/16/is-it-time-to-re-examine-workplace-wellness-get-well-quick-schemes (“We suggest several strategies . . . to re-allocate wellness dollars from ‘get well quick’ schemes to the much more challenging, but ultimately more rewarding, task of truly creating a culture of wellness, a workplace that can attract and retain healthier, presumably more productive, people than competitors do.”).

  346. [346]. See Joseph Brownstein, Is Whole Foods’ Get Healthy Plan Fair?, ABC News (Jan. 28, 2010),
    story?id=9680047; Edgar Sandoval & Kathleen Lucadamo, Whole Foods to Give Greater Employee Discounts to Workers with Lower BMI, Cholesterol, Daily News (Jan. 27, 2010, 1:40 AM), www.nydailynews.

  347. [347]. Sandoval & Lucadamo, supra note 346 (quoting Peggy Howell, a spokeswoman for the National Association to Advance Fat Acceptance).

  348. [348]. Brownstein, supra note 346.


Assistant Professor of Law, University of Houston Law Center. 

Thank you to Josh Blackman, Zack Bray, Jessica Clarke, Mary Coombs, Ben Edwards, Dave Fagundes, Peter Huang, Elizabeth Weeks Leonard, Dan Schwarcz, Ron Turner, and faculties of the Temple School of Law, the Florida State University College of Law, the Thurgood Marshall School of Law, the University of Houston Law Center, and the University of Miami School of Law for commenting on earlier versions. I presented a portion of this project at the American Association of Law Schools 2013 Annual Meeting during the Law, Medicine, and Health Care Section Panel. My appreciation also goes to Louis Holzer and TJ Tesch for research assistance, Emily Lawson for library support, and Elaine Gildea for administrative aid. Finally, many thanks to my editors at the Iowa Law Review, especially Courtney Burks.