Reasonable Accommodation11. I would like to thank Professor Daniel Sarmento for bringing to my attention the importance of the concept of reasonable accommodation. I dedicate this article to my friends Leonardo, from Criciúma, and Diana and Daniel, from Rio, people who have surpassed countless barriers and who are fighting for inclusion.

2. Brief notes about the UN Convention on the Rights of Persons with Disabilities

The UN/CRPD was approved in December 2006 and ratified by Brazil in March 2007 in New York. In 2008, it was incorporated into the Brazilian Constitution in the form of a constitutional amendment, under the terms of Article 5, paragraph 3 of the Constitution (Legislative Decree No. 186, BRASIL, 2008). As the first international convention to be approved in the new millennium, the UN/CRPD concerns itself with a portion of the population that has been, and still is, the subject of discrimination and oppression: people with disabilities. The cornerstones of the Convention are inclusion, parity of participation, full enjoyment of rights and dignity for people with disabilities, with additional attention given to the juxtaposition of other factors of exclusion and discrimination, such as combinations of gender, age, childhood, poverty and disability.

The UN/CRPD is noteworthy on many levels. It unifies in one international document a range of human rights recognized for a “creditor group” (FIGUEIREDO, 2010), people with disabilities, and reaffirms the “universality, indivisibility and interdependence” of human rights. In this respect, the UN/CRPD establishes a series of legal concepts, some of which are fairly innovative in some legal systems, such as the notions of (a) disability; (b) discrimination; (c) reasonable accommodation.

On disability, the UN/CRPD does not espouse a merely biomedical approach (i.e. one that “holds that there is a causal and dependent relationship between the bodily impairments and the social disadvantages experienced by persons with disabilities” in which “a body with impairments should be the subject of medical knowledge intervention” (DINIZ; BARBOSA; SANTOS, 2009, p. 66-68), but instead combines it with a more comprehensive and even combative approach than the exclusively biomedical one: the social model of disability. Within this social model, disability “has come to be understood as the experience of inequality shared by people with different types of [bodily or mental] impairments,” the cause of which is not their bodies but rather barriers, obstacles and social oppressions. Disability, therefore, has moved beyond the purely biomedical field, grounded in measurement, evidence, treatment and cure, and into a moral and cultural understanding that countless obstacles are found outside our bodies, in the material and moral environment that surrounds us (DINIZ; BARBOSA; SANTOS,2009, p. 70). Nevertheless, the social model incorporated by the UN/CRPD does not completely discard elements of the biomedical model, particularly in regard to the fundamental rights of promotion, protection and access to health.

On discrimination, the UN/CRPD alters conventional thinking in two ways. First, the UN/CRPD expands the concept, defining it also in virtue of not offering reasonable accommodation. Therefore, in addition to the traditional forms of unequal and discriminatory treatment, the denial of reasonable accommodation that does not impose an undue burden constitutes discrimination against people with disabilities. This illustrates the singularity of the concept of reasonable accommodation, since, by modifying the content of discrimination, it “demonstrates the recognition of environmental barriers as a preventable cause of inequalities experienced by people with disabilities” (DINIZ; BARBOSA; SANTOS, 2009, p. 70; EMENS, 2008, p. 877). Second, the UN/CRPD expressly broadens the meaning of discrimination, including direct and indirect forms, such as adverse impact discrimination.

If there was ever any doubt about adverse impact discrimination figuring in Brazilian constitutionalism, it has now dissipated. The doctrine of adverse or disproportionate impact discrimination was developed by the U.S. Supreme Court in the case of Griggs v. Duke Power Co. (apud SARMENTO, 2006, p. 125; MARTEL, 2007), decided in 1971. This model of discrimination differs from direct discrimination, in which the legal norm or administrative practice is intentionally and on its face discriminatory. Adverse impact discrimination, in contrast, occurs when public or private measures that are neither discriminatory in their origin nor imbued with discriminatory intent, end up causing manifest harm, normally upon their application, to certain minority groups whose physical or mental characteristics or lifestyles distinguish them from the general group of people for which the policy was designed. Based on this doctrine, courts can invalidate or make exceptions to laws, administrative acts or even collective agreements and business policies, creating a barrier against indirect harm caused to minorities (SARMENTO, 2006, p. 125; MARTEL, 2007).

This concept of discrimination raises an important question: what is reasonable accommodation? The text of the UN/CRPD offers a definition:

Reasonable accommodation: means necessary and appropriate modification and adjustments not imposing a disproportionate or undue burden, where needed in a particular case, to ensure to persons with disabilities the enjoyment or exercise on an equal basis with others of all human rights and fundamental freedoms.
(NAÇÕES UNIDAS, 2006a).

Seemingly, reasonable accommodation is not a problematic or complex concept. It means using all available mechanisms to adjust practices, materials, environments, general rules, etc. to the differences between people in order to assure them equal opportunities. However, the term “reasonable” does create ambiguity since it is not clear what constitutes reasonable mechanisms. The existing ones? An ordinary one? The best possible ones? And what are we to understand by “undue burden”? Does this expression refer only to the cost? The subsequent sections are intended to provide an explanation of the concept in other legal systems, in order to develop an interpretive proposal for the case of Brazilian law.

3. Reasonable accommodation and undue burden – an overview of foreign and international law

The concept of reasonable accommodation originated in the United States following the approval of the Equal Employment Opportunity Act of 1972 (USA, 1972) that aimed to combat discrimination in the labor market. The term was originally used in reference to religious discrimination, requiring employers to demonstrate that they would be unable to reasonably accommodate the religious practices of their employees without undue hardship. Since then, the expressions reasonable accommodation and undue hardship have figured prominently in U.S. law. However, it was with the approval of the Americans with Disabilities Act-ADA in 1990 that they took hold and became the subject of more intense discussions, both in doctrine and in case law (USA, 1990).

Canada may also be considered one of the birthplaces of reasonable accommodation. In the 1970s, while the U.S. was forging the concept, Canada had a conservative discrimination case law, primarily because it did not recognize doctrines such as adverse impact discrimination. Nevertheless, Canada ended up importing from the U.S. both the concept of adverse impact and of reasonable accommodation. Moreover, the development of the concept of reasonable accommodation flourished more in the Canadian system than in the U.S., converting Canada’s legal rulings into a reference on the subject.

Over the past decade, a number of European countries and the EU have also incorporated the concepts of reasonable accommodation and undue burden into their legal systems. In more recent years, after the approval of the UN/CRPD, they have been adopted in the field of global human rights protection. The long history of these two concepts and their diffusion across various different countries and regional and global human rights protection systems would appear to indicate their suitability, borrowed from other legal systems after careful consideration and based on clear and stable legal structures. Nevertheless, Waddington demonstrates that the concepts were forged in a highly questionable manner in the U.S., meaning that the selected reference has proven to be complex and confusing. Moreover, European countries have not responded to the challenge of developing these concepts in identical ways, which has made their application even more challenging (WADDINGTON, 2008, p. 318-319).

To better understand the subject, an overview shall be presented of the concepts in the U.S., Canada, and some European countries as well as in the European Union. This survey will make it possible to draw some conclusions about the concepts and also identify the more controversial points, thereby allowing the Brazilian interpretation to avoid the same highs and lows, and rights and wrongs encountered elsewhere. An understanding of the concepts in other legal systems is the first step towards a fertile legal dialogue in times of “transconstitutionalism” (NEVES, 2009, p. 265) and intense exchanges between national, regional and international human rights protection systems. It is worth noting that Brazil has an advantage over most other States. Here, the concepts enjoy constitutional status, making superfluous many of the discussions that took place elsewhere on legislative interpretation, on correlation between international or community law documents and the internal legal system, on the constitutionality of the laws and on the differences between the administrative and legislative spheres.

3.1 The United States of America – on how not to read the convention

The narrative on reasonable accommodation and undue burden in United States law reads like a clear example of “judicial backlash” (KRIEGER, 2003, p. 340; MALHOTRA, 2007, p.9). Introduced legislatively in a socially and legally eventful period marked by an active civil rights movement, reasonable accommodation was the subject of timid interpretation by the judiciary, less assertive than expected, particularly considering two decisions on matters of religion delivered by the Supreme Court prior to the adoption of the concept. In both cases, there is a recurrent allusion to the issue of reasonable accommodation, although obviously without being explicitly mentioned:

(a) in the Yoder case, the Supreme Court decided in favor of families belonging to the religious Amish community who refused to send their children to high school, in breach of state laws mandating compulsory schooling until age 16. Despite claims by the state of the need for universal education of children and adolescents, the Court accommodated the interests at stake, making an exception to the law to allow the community to educate their adolescents (USA,Wisconsin v. Yoder, 1972);

(b) in Sherbert v. Verner the Court decided that the denial of unemployment benefits unduly infringed on the exercise of the religious convictions of an Adventist woman who refused job offers that required her to work on Saturday, a day of rest and worship when members of her denomination are forbidden to work. In doing so, it accommodated the interests of a religious group, by making an exception to the general legal norm (USA, Sherbert v. Verner, 1963; MARTEL, 2007, p. 33).

Note that these two rulings, in spite of their being made prior to legislative adoption of the concept of reasonable accommodation, were imbued with its spirit. People belonging to minority religious groups who were adversely impacted by impartial laws or administrative acts had their rights asserted and exceptions were made to reconcile lifestyles and profound beliefs that differ from the mainstream, thereby avoiding discrimination and forced assimilation and integration by the state. It would be plausible to suppose, therefore, that the Supreme Court would follow the same path when interpreting reasonable accommodation and undue hardship after the legislation came into force.

This did not happen. The reasons are difficult to identify and though this is not the place for speculation, it is suspected that the Court was possibly apprehensive about the implications that reasonable accommodation involving religious matters in private labor relations might have on the separation of church and state. In this regard, the Court adopted remarkably restrictive positions, especially concerning: (a) the verification of the costs and burdens to be borne by the employers, which was carried out through application of the de minimis rule, i.e., any cost beyond the minimum constituted an undue burden; (b) the definition of the principle of equality, which was limited and formal; (c) the restrictive interpretation of parties with the duty to accommodate; (d) the distorted interpretation of legal texts, concerning both their wording and their legislative history (USA, Trans Word Airlines, Inc. v. Hardison, 1977; USA, Ansonia Board of Education v. Philbrook,1986).

One striking aspect of the Supreme Court’s interpretation of legal texts is the fact that it separated the term “reasonable,” the adjective qualifying accommodation, from undue hardship. An examination of undue hardship is only made if employers are unable to provide anyaccommodation deemed reasonable. This being the case, any proposed accommodation, even if not accepted by the claimant – given its insufficiency, inefficiency or because it results in a loss of wages or employment benefits or even because it stigmatizes – is generically reasonable, and the accommodating party is not required to incur any additional costs to obtain a more adequate accommodation for those involved. And once considered reasonable, no proof of undue hardship is even required. Furthermore, an accommodation may be deemed not reasonable in and of itself, regardless of costs, as explained below.

Following the approval of the ADA (USA, 1990), a renewal of jurisprudence on reasonable accommodation might have been expected now that it was applicable to people with disabilities. Nevertheless, the courts understood that, since legislators had chosen terms previously defined in rulings on religious matters, the intention was to maintain the existing meaning. The restrictive interpretation persisted. Even worse, a new reading, grounded in elements of the law and economic schools of thought, constrained the scope of reasonable accommodation even further.

The heavily criticized decision was given by Chief Judge Posner, in a court of appeals, in theVande Zande case. Mrs. Lori Vande Zande became paralyzed from the waist down. Confined to a wheelchair, she occasionally suffered from pressure ulcers that required her to take time off work. The first aspect analyzed in the case was whether these ulcers come within the legal concept of disability. Although they do not themselves constitute a disability, they were caused by the disability. Consequently, they were considered part of the disability, giving rise to a duty of reasonable accommodation to the point of undue hardship. With regard to accommodations, Vande Zande made two requests of her employer, the state of Wisconsin: (a) to lower a sink in the kitchenette so she could wash her coffee cup and get a glass of water, that is, use the services of the kitchenette like the other employees.³ She was only able to reach the sink in the bathroom, where she performed the activities she would prefer to do in the kitchenette. As far as she was concerned, while using the bathroom did not interfere directly with her work, it caused her moral damages and a degree of stigmatization in the workplace; (b) restoration of 16.5 hours of sick leave she took as a result of her employer’s refusal to let her work full time at home for eight weeks (USA, Vande Zande v. Wisconsin Dep’t of Admin., 1995).

When interpreting the ADA, Posner understood that the legal norm provides a dual and broad definition of disability, including (a) a physical or mental impairment that substantially limits one or more of the major life activities of an individual; (b) the state of being regarded as weak, fragile or defective to the outside observer. To define reasonable accommodation and undue hardship, Posner drew on previous cases. For workplace accommodation, he understood that “the employer must be willing to consider making changes in its ordinary work rules, facilities, terms and conditions in order to enable a disabled individual to work” (USA, Vande Zande v. Wisconsin Dep’t of Admin., 1995).

According to Posner, the difficulty rested in defining the term “reasonable.” For the claimant, reasonable meant effective accommodation for the individual. Her case argued that costs should not be considered in the definition of “reasonable,” but rather in the examination of undue hardship. Posner disagreed. He suggested that “reasonable” be interpreted as an adjective that weakens the term “accommodation,” that is, in the sense in which the word is applied in civil law (e.g., reasonable effort, reasonable care), indicating an ordinary attempt, and not the maximum possible or the maximum desirable (USA, Vande Zande v. Wisconsin Dep’t of Admin., 1995).

The expression “undue hardship,” which had been interpreted by the trial court judge as that which unduly burdens an employer according to their financial condition, was defined by Posner as a combination of two elements: (a) the benefits to the disabled person; (b) the resources of the employer. In Posner’s opinion, the examination of costs comes into play for a second time, when the person requesting the accommodation must demonstrate that it is: (a) effective and (b) proportional to costs. Then the party with the duty to accommodate may present two defenses: (a) excessive costs in relation to the benefits of the accommodation and (b) the impossibility of the expense given its financial condition (USA, Vande Zande v. Wisconsin Dep’t of Admin., 1995).

Concerning Mrs. Vande Zande’s request for the kitchenette sink to be lowered, Posner concluded that the accommodation was unnecessary since she already had use of another equally effective sink in the bathroom. Therefore, she already had a reasonable accommodation:

Her argument rather is that forcing her to use the bathroom sink for activities (such as washing out her coffee cup) for which the other employees could use the kitchenette sink stigmatized her as different and inferior; she seeks an award of compensatory damages for the resulting emotional distress. We may assume without having to decide that emotional as well as physical barriers to the integration of disabled persons into the workforce are relevant in determining the reasonableness of an accommodation. But we do not think an employer has a duty to expend even modest amounts of money to bring about an absolute identity in working conditions between disabled and nondisabled workers. The creation of such a duty would be the inevitable consequence of deeming a failure to achieve identical conditions “stigmatizing.” That is merely an epithet. We conclude that access to a particular sink, when access to an equivalent sink, conveniently located, is provided, is not a legal duty of an employer. The duty of reasonable accommodation is satisfied when the employer does what is necessary to enable the disabled worker to work in reasonable comfort.
(USA, Vande Zande v. Wisconsin Dep’t of Admin., 1995).

Since the employer was the state of Wisconsin, which could hardly claim undue hardship in virtue of one hundred and fifty dollars, the relationship established between the concepts of reasonableness and undue hardship leads one to believe that even modest amounts of money can be denied if the accommodation is not reasonable (SUNSTEIN, 2007, p. 8).

Concerning the possibility that the claimant work at home if necessary, Posner refused, based on a series of preconceptions such as the difficulty of supervising the employee, the effect on teamwork, the impact on work colleagues, the reduction in the quality of the work performance. Furthermore, he questioned whether granting permission to work at home would not cause a reduction in the duty to accommodate, since employers could adopt this option to avoid having to make changes to the workplace, thereby segregating people with disabilities (USA, Vande Zande v. Wisconsin Dep’t of Admin., 1995).

The ruling caused controversy on a number of points: (a) the lack of an effective cost-benefit analysis (SUNSTEIN, 2007); (b) the limited interpretation of the benefits, especially given moral damages and stigmatic harms (SUNSTEIN, 2007; EMENS, 2008; MINOW, 2008); (c) the restrictive reading of the underlying principles of the ADA (SUNSTEIN, 2007, 2008).

Cost-benefit analysis has some important virtues and vices. The former include: (a) exposure of the fact that a refusal to accommodate may stem from mere habit or prejudice; (b) demonstration of the potential benefits to the disabled person and the potential costs to the employer; (c) identification of assumptions or preconceptions that are not based on reality. The latter include (a) possible incorporation of unreliable assumptions, instead of basing analysis on reality; (b) possible oversight of a crucial aspect of discrimination and the costs of daily humiliation, inequality and stigmatization (SUNSTEIN, 2007). The ruling, as Sunstein asserts, features the vices of cost-benefit analysis, since it minimizes the significant costs of stigmatic harm to and daily humiliation of people with disabilities (SUNSTEIN, 2007, p. 2). Furthermore, the decision rests on assumptions about working from home. However, a serious cost-benefit analysis of reasonable accommodation includes more than just economic factors – necessarily proven or simulated according to credible calculations. It must also include the costs of stigma and the benefits of inclusion, not only to the person requesting the accommodation, but also to third parties (SUNSTEIN, 2007, p. 2-4; EMENS, 2008).

This issue of benefits inspired the extensive study of Emens (2008). Emens dismisses the hypothesis that benefits are only felt by the individual who requests the accommodation, demonstrating that they can extend to the accommodating party and also, directly or indirectly, to second and third parties. The way to address the benefits should be less atomistic and more extensive, interactive, relational and collective, which modifies and complicates empirical analyses of costs and benefits (EMENS, 2008, p. 843; MINOW, 2008). In the Vande Zande decision, Posner emphasized the benefits to the employee and the costs to the employer, overlooking the benefits to other disabled employees, to employees more generally, to third parties outside the workplace and, ultimately, to the employer itself. This is a static conception of accommodation, as it “understands accommodation as a special thing done for one or a few individuals, for a subset of the population, to make it possible for those different individuals to participate in, for example, the workplace.” What is proposed is a dynamic model of accommodation that “understands accommodation as a process of interrogating the existing baseline, by focusing on the part of the population that was neglected in the creation of that baseline, [and] to make changes to that baseline that may affect everyone” (EMENS, 2008, p. 894). This being the case, there are a series of benefits and costs to be assessed, many of which are not economic in nature.

Finally, there are signs that the objective of the ADA was distorted in Vande Zande. In the case, it implicit from the interpretation that the purpose of the ADA was to save public funds, by removing people from the scope of welfare and social security (USA, Vande Zande v. Wisconsin Dep’t of Admin., 1995; USA, Borkowski v. Valley Central School District, 2002a; EMENS, 2008, p. 870). Sunstein emphatically denies that this is the purpose of the ADA. The underlying ideal of the ADA is the inclusion of people with disabilities, previously considered inferior and of a lower caste. It represents, therefore, the struggle for profound equality, for universal human rights, for justice (NUSSBAUM, 2007) and plurality. A non-assimilationist inclusion, beyond simple integration, is that which reconstructs and reorders the standards of architecting environments and creates rules for the workplace, schools and universities; that is, the intention is to make us realize that we establish the physical and regulatory structures that surround us through an archetype of normality that creates barriers for thousands of people. In short, it is an anti-caste proposal (SUNSTEIN, 2001, p. 155-182; 2008, p. 21) to create obstacles for oppression, not for people.

On the application of reasonable accommodation in the context of the ADA, there has been at least one more important case heard by the Supreme Court; it upheld the restrictive definition of parties with the duty to accommodate by excluding unions from its spectrum and by refusing the possibility of accommodation if it affects collective bargaining agreements or seniority systems. Furthermore, the Court clearly rejected the interpretation of the word “reasonable” as an indication of the effectiveness of the accommodation, taking a similar approach to the one formulated by Posner, i.e., that while the accommodation should be effective, the word “reasonable” is a modifier that weakens the noun. To make matters worse, the Court did not consider the magnitude of the employer when judging whether the costs constituted an undue hardship (USA, U.S. Airways, Inc. v. Barnett, 2002b).⁴

3.2 Canada – afine mosaic

In Canada, the development of the concept of reasonable accommodation began after the introduction of the U.S. doctrine of adverse impact. Once adverse impact was recognized, the duty would follow to accommodate reasonably, unless an undue hardship was demonstrated by the party owing the duty. The case law deals primarily with accommodation for religious minorities. It first appeared in the O’Malley case (CANADA, Ont. Human Rights Comm. v. Simpsons-Sears, 1985b), in which there were already signs that the thrust of the court rulings would be different than in the U.S.⁵

First of all, the Canadian Court made it clear that the term “reasonable,” when associated with “accommodation,” should be regarded as dependent on proof of undue hardship, i.e., accommodation is not reasonable when it imposes undue hardship. Therefore, while in the U.S. accommodation may be considered unreasonable in itself, even when hardship is minimal, in Canada accommodation fails to be reasonable if and only if there is proof that it will cause undue hardship on the party being asked to accommodate (MALHOTRA, 2007, p. 12).

Secondly, the Canadian Court established six factors to be considered when determining undue hardship in the workplace: (a) financial costs; (b) impact on collective bargaining agreements; (c) problems of employeemorale; (d) interchangeability of workforce and facilities; (e) size of the employer; (f) safety. In real situations, an analysis is made of the weight to be conferred to each of these factors (CANADA, Central Alberta Dairy Pool v. Alberta, 1990).

Thirdly, the Canadian Court did not restrict the range of parties with the duty to accommodate. The U.S. Supreme Court resisted the inclusion of parties other than those expressly mentioned in the legal norms, especially private parties. In Canada, participants in cases of discrimination, such as unions and condominium associations, were deemed to be parties with the duty to accommodate or similar duties even if circumstances were simply unfortunate and unintentional (CANADA, Central Okanagan School District No. 23 v. Renaud, 1992; CANADA, SyndicatNorthcrestv. Amselem, 2004).

By broadly interpreting the types of parties with the duty to accommodate, the Court took a truly interesting step in the procedure for arriving at reasonable accommodation. It considered this procedure to be an opportunity for dialogue, in which all the parties involved should participate. In this regard, the party requesting the accommodation has the duty to explain their limitations and needs and, if possible, to identify alternatives. The subject of the request, meanwhile, has the duty to offer reasonable proposals that, if genuinely reasonable, the claimant has the duty to facilitate and contribute to their implementation. According to the Court, “discrimination in the workplace is everybody’s business” (CANADA, Central Okanagan School District No. 23 v. Renaud, 1992). Therefore, fourthly, it can be noted that the pursuit of reasonable accommodation refers to a process of dialogue that is multilateral, participatory and inclusive.

By charging professional associations with the duty to accommodate, the Court makes it clear that this type of organization can also cause discrimination, whether directly or through adverse impact. Although the theory is not formulated in the words of the Court, note the presence of the three-dimensional model structured by Nancy Fraser, which claims that in demands for justice and inclusion there is a need to combine recognition, redistribution and representation (FRASER, 1996). Indeed, unions are professional associations that are directly involved with the issue of redistribution in capitalist societies. To suppose that unions cannot give rise to demands for recognition is to assume that they represent entirely homogenous groups whose demands for inclusion, self-respect and dignity are exactly the same, or at the very least that they exhibit a high level of parity of participation. The casuistry of both the U.S. and Canada indicate that unions are not always this impartial and representative. Architectures of social exclusion, whether intended or not, can be reproduced on a micro-scale in redistribution organizations. Membership in a professional association alone does not mean that workers from minority groups have their demands for recognition taken into consideration or that they effectively enjoy parity of participation.

Fifthly, the Supreme Court of Canada emphatically refused to adopt the U.S. constitutional standards of interpreting undue hardship based on the de minimis test for two reasons: (a) thede minimis test is not compatible with the concept of undue hardship formulated by the Court; (b) the legal term is not hardship alone, but hardship qualified as undue. The de minimis test, however, leads to the recognition of only the hardship and there will always be some hardship; it is the cost of protecting the fundamental rights and shaping an inclusive society (CANADA, Central Okanagan School District No. 23 v. Renaud, 1992; CANADA, Commission ScolaireRégionale de Chambly v. Bergevin, 1994). What there cannot be is an excessive hardship:

The Hardison de minimis test virtually removes the duty to accommodate and seems particularly inappropriate in the Canadian context. More than mere negligible effort is required to satisfy the duty to accommodate. The use of the term “undue” infers that some hardship is acceptable; it is only “undue” hardship that satisfies this test. The extent to which the discriminator must go to accommodate is limited by the words “reasonable” and “short of undue hardship”. These are not independent criteria but are alternate ways of expressing the same concept. What constitutes reasonable measures is a question of fact and will vary with the circumstances of the case.
(CANADA, Central Okanagan School District No. 23 v. Renaud, 1992, emphasis added).

In more recent decisions, the Supreme Court has tried to refine its interpretation of reasonable accommodation, mainly by correlating the reasonableness of an accommodation with the widely acknowledged proportionality test. There have been two stages of comparison between reasonable accommodation and proportionality. First, the Court connected the steps of the test of reasonable accommodation. In the Court’s interpretation, an examination of whether there are less harmful means of implementing a measure (minimal impairment) permits an analysis of whether a reasonable accommodation is available for the situation. If an accommodation is identified and undue hardship is not proven, this will be the minimal impairment for those adversely affected by measures that, when viewed generally and abstractly, are fair and proportional. Obviously, the public administration and the legislator are not required to foresee each and every possible adverse impact of a normative act that is not intentionally discriminatory and that in general does not infringe on the rights of the majority of people. Nevertheless, if the judiciary is confronted with an adverse impact, it is authorized, when examining proportionality, to find a means that minimizes infringement on the rights of those adversely affected by the measure, thereby recognizing a duty of reasonable accommodation to the point of undue hardship:

This correspondence of the concept of reasonable accommodation with the proportionality analysis is not without precedent (…).
In my view, this correspondence between the legal principles is logical.  In relation to discrimination, the courts have held that there is a duty to make reasonable accommodation for individuals who are adversely affected by a policy or rule that is neutral on its face, and that this duty extends only to the point at which it causes undue hardship to the party who must perform it. Although it is not necessary to review all the cases on the subject, the analogy with the duty of reasonable accommodation seems to me to be helpful to explain the burden resulting from the minimal impairment test with respect to a particular individual, as in the case at bar. In my view, Professor José Woehrling correctly explained the relationship between the duty to accommodate or adapt and the Oakes analysis in the following passage: ‘Anyone seeking to disregard the duty to accommodate must show that it is necessary, in order to achieve a legitimate and important legislative objective, to apply the standard in its entirety, without the exceptions sought by the claimant.  More specifically, in the context of section 1 of the Canadian Charter, it is necessary, in applying the test fromR. v. Oakes, to show, in succession, that applying the standard in its entirety constitutes a rational means of achieving the legislative objective, that no other means are available that would be less intrusive in relation to the rights in question (minimal impairment test), and that there is proportionality between the measure’s salutary and limiting effects. At a conceptual level, the minimal impairment test, which is central to the section 1 analysis, corresponds in large part with the undue hardship defence against the duty of reasonable accommodation in the context of human rights legislation.
(CANADA, Multani v. Commission scolaire Marguerite-Bourgeoys, 2006, emphasis added).

The correlation between reasonable accommodation and proportionality became nebulous in a more recent ruling, since they were deemed to be distinct constitutional tests. The reasonable accommodation standard should be used when analyzing the application of a law or when examining administrative acts and practices, covering both public and private entities. The proportionality test, meanwhile, should be applied to less individualized contexts on the constitutionality of laws and normative acts, involving the relationship of the legislator with the subjects of the law. Furthermore, one of the stages of the proportionality test – minimal impairment – was dissociated from reasonable accommodation:

Minimal impairment and reasonable accommodation are conceptually distinct. Reasonable accommodation (…) envisions a dynamic process whereby the parties – most commonly an employer and employee – adjust the terms of their relationship in conformity with the requirements of human rights legislation, up to the point at which accommodation would mean undue hardship for the accommodating party (…).

A very different kind of relationship exists between a legislature and the people subject to its laws. By their very nature, laws of general application are not tailored to the unique needs of individual claimants. (…). A law’s constitutionality under section 1 of the Charter is determined, not by whether it is responsive to the unique needs of every individual claimant, but rather by whether its infringement of Charterrights is directed at an important objective and is proportionate in its overall impact. While the law’s impact on the individual claimants is undoubtedly a significant factor for the court to consider in determining whether the infringement is justified, the court’s ultimate perspective is societal. The question the court must answer is whether the Charter infringement is justifiable in a free and democratic society, not whether a more advantageous arrangement for a particular claimant could be envisioned. 

Similarly, ‘undue hardship’, a pivotal concept in reasonable accommodation, is not easily applicable to a legislature enacting laws (…).

In summary, where the validity of a law of general application is at stake, reasonable accommodation is not an appropriate substitute for a proper section 1 analysis based on the methodology of Oakes.
(CANADA, Alberta v. Hutterian Brethren of Wilson Colony, 2009).

Note that the majority of the Court drew a significant distinction between proportionality and reasonable accommodation. The former would be used to control the constitutionality of general and abstract laws and normative acts confined to the field of public law and the latter to public and private administrative practices, in which there is a greater individualism. As a result, demands for reasonable accommodation are practically impossible when it comes to laws or normative acts that are more general and abstract. Therefore, the majority reading removes the legislator from the scope of the duty to accommodate (CANADA, Alberta v. Hutterian Brethren of Wilson Colony, 2009).

Notwithstanding this last decision and its potential consequences, the set of rulings by the Supreme Court of Canada on reasonable accommodation for religious minorities is very rich, both in conceptual definition and in their comprehensiveness and scope. The interpretation of ideas associated with protection and promotion of equality and dignity in order to build a society based on plural and intercultural foundations, a genuine inclusive mosaic, seems to have prospered much more in Canada than in other Western countries. The open nature of inclusive constitutional construction prompted the expansion of the concept of reasonable accommodation into the field of disability on far more interesting and fruitful terms than the U.S. model, which “dramatically contrasts” with that of Canada. (MALHOTRA, 2007, p. 14).

On the matter of disability, there are numerous Canadian decisions on reasonable accommodation, most in administrative arbitration systems. According to scholars, the interpretation has been as sophisticated as the reading on religious accommodation: “Disability rights jurisprudence has applied the Supreme Court of Canada’s decision in Dairy Pool in the context of providing reasonable accommodations to workers with disabilities in increasingly complex and sophisticated ways” (MALHOTRA, 2007, p. 15-16).

Finally, it should be clarified that the Court has unanimously reviewed the distinctions betweendirect discrimination and adverse impact discrimination. Prior to this, the Court separated the two, applying different standards and remedies for each one. The Court understood that a new approach was necessary for cases of discrimination and related accommodation, adopting a new standard for examining discrimination and reasonable accommodation in order to simplify the interpretations of the Canadian legislation and to clarify the scope of reasonable accommodation. Nonetheless, it made note of the importance of having the previous standards in place, primarily the set of rulings that recognized and developed the notions of adverse impact and reasonable accommodation (CANADA, British Columbia (Public Service Employee Relations Commission) v. BCGSEU, 1999).

3.3 European States and the European Union

Reasonable accommodation was not part of the legal systems of European countries until 2000, when the Council of the European Union adopted the Employment Equality Directive (Directive 2000/78/CE, UNIÃO EUROPEIA, 2000; WADDINGTON, 2008, p. 317). After this adoption, the concept spread across the national systems and was strengthened with the authorization and signing, by the European Union, of the UN/CRPD in 2007 and 2009, respectively (NEVES, 2010, p. 5).

To explain how the concepts were understood in the national and community systems, this article shall draw on the comprehensive research of Waddington (2008), written with the purpose to consider “how the Member States of the European Union have responded to the challenge” to apply and interpret the dual concepts of reasonable accommodation and disproportionate or undue burden. Waddington found that countries have employed different terminologies for accommodation (e.g., adaptation, adjustments, steps, appropriate measures) without any significant legal consequences. However, the term “reasonable” has been the subject of some very different interpretations. In a few States, this was the only limiting modifier used; in others, the term “reasonable” comes hand-in-hand with the expression of disproportionate or undue burden, or some other analogous expression. The question then is: what meanings are conferred on “reasonable”? How do they interact with the notion of disproportionate burden? (WADDINGTON, 2008, p. 323-326).

The author concludes that there are three approaches to understanding the word reasonable and its interaction with undue burden: (a) an accommodation will be reasonable if it does not impose excessive difficulties or costs on the party with the duty to accommodate. This reading is usually followed by the stricter test of disproportionate burden. Therefore, an accommodation may be deemed unreasonable without ever applying the burden test. Or, on very rare occasions, and it is difficult to conceive of such a situation, an accommodation could be deemed reasonable and also impose an undue burden. (b) An accommodation will be reasonable if it is effective, i.e., if it allows the claimant to carry out the activities that gave rise to the request. In defense, the accommodating party may claim an undue or disproportionate burden; (c) an accommodation will be reasonable if it is effective for the right holder while not imposing excessive inconveniences or costs on the accommodating party (WADDINGTON, 2008, p. 339). Given the diversity of interpretations, Waddington (2008, p. 339-340) concludes that it is up to the European Court of Justice (ECJ), by standardizing the readings and applications, to determine the interpretation.

So far, there have been three key rulings in the field, two from the ECJ and one from the European Court of Human Rights (ECHR). In its decision on the Chacón Navas case, the ECJ did not extend the protection from dismissal for disabled workers to employees who become sick, thereby distinguishing sickness from disability (TRIBUNAL DE JUSTIÇA EUROPEU, Sonia Chacón Navas v Eurest Colectividades SA, 2006). In the Coleman case, meanwhile, the ECJ recognized that the prohibition on discrimination and harassment does not only apply to people with disabilities, but also to those who are primary caretakers of people with disabilities. In the abstract, the ECJ upheld the thesis that a person without disabilities who is responsible for the care of a child with disabilities can suffer harassment in the workplace to the extent that they can request reasonable accommodation. It also established the standards for the burden of proof on the matter of discrimination and accommodation (TRIBUNAL DE JUSTIÇA EUROPEU, S. Coleman v. Attridge Law and Steve Law, 2008).

In an emblematic case, trialed in 2009, dealing with the incapacity of an individual to assume mandatory military service, the ECHR for the first time, as summarized by Kanter (2009): (a) alluded to reasonable accommodation; (b) recognized a violation of the right to not be discriminated against in virtue of disability: (c) expressly mentioned the UN/CRPD (TRIBUNAL EUROPEU DE DIREITOS HUMANOS, Affaire Glor v. Suisse, 2009). Despite having drawn on the concept of reasonable accommodation, the ECHR did not expressly mention the term in the ruling, nor did it specify the forms and readings it could assume in the case. Another important aspect of the ruling, though the limitations of this article do not permit a lengthy discussion here, is the dividing line between sickness and disability.

4. Reasonable accommodation: hermeneutic proposals from an inclusive constitutional perspective

The UN/CRPD is formally part of the Brazilian Constitution. The first conclusion to draw from this is that all past and future legal norms on disability need to be written and interpreted in accordance with the Constitution. Its interpretation should be based on inclusive constitutional hermeneutics. This means treating the Constitution as a touchstone for breaking with repressive, oppressive, segregative and assimilationist socio-political and economic environments. It is a normative framework designed to break down structures of direct and de facto power, as well as the machinery of symbolic power, in the sense that Bourdieu gives the term (BOURDIEU, 2010, p. 8); a constitutional document modeled on dignity, liberty, equality, solidarity, justice, participation and plurality. Therefore, the interpretive guidelines that are suggested are based on the inclusion and the equal and dignified participation of all human beings in a wide variety of settings.

It should be noted that the UN/CRPD was approved as a mechanism for the protection and promotion of the rights of people with disabilities, within both the biomedical and the social conception of disability. It is a document with inclusive and emancipatory purposes intended to minimize the barriers that contribute to or shape the asymmetry between humans. On this point, the article agrees with Sunstein. The document is underpinned by an anti-caste principle. A side effect of protecting the rights of people with disabilities can even result in less spending and increased saving within some government departments, although it is doubtful that the network of protection and promotion for people with disabilities resulting from the UN/CRPD is intended primarily to reduce public spending.

Now that the initial considerations have been made, the next step is to propose how to interpret the dual concepts of reasonable accommodation and undue burden in Brazilian law. To begin with, although the UN/CRPD is part of the constitutional block, it was written in international terms for States Parties to adopt “legislative, administrative and other measures” to protect rights. In this regard, one might think that the UN/CRPD would acquire a programmatic character, attributing to States Parties commitments ad futurum, without having to be self-executing. Despite being written in a future-oriented tone and the presence of judicial resistance to the self-executing nature of IHRT clauses (STF, ADI 1480-DF/MC,BRASIL, 1996), it is believed to serve as current guidance to the Brazilian Supreme Court (STF) (STF, HC 87585/TO, BRASIL, 2008). Regarding the understanding of the STF on IHRTs, (STRAPAZZON, 2009, p. 379-399), the wording of Article 5, paragraph 2 c/c Article 5, item XXV, both from the Brazilian Constitution (BRASIL, 1988), confers self-executing status to a significant portion of the clauses of IHRTs, excluding only the more sensitive points; the application of these sensitive points requires complex public policies with particularities outside the scope of the judiciary, whether by virtue of institutional limitations and the separation of state functions or by virtue of the impossibility of effective implementation through the judicial system. Reasonable accommodation, as international casuistry shows, normally takes place to make exceptions to a general law or to the general rules of private entities, accommodating the needs of a person given the unique obstacles that a body or mind confronts. Therefore, it is a space in which the court may be called upon to act without significant problems, from a lack of provisions in the legislative or administrative fields, to allow exceptions to the general norms of these state functions.

The question then is: how should one read the binomial reasonable accommodation and undue burden in the Brazilian Constitution? Three points shall be addressed: (a) the holders of the right to accommodation; (b) the terms “reasonable accommodation” and “undue burden”; (c) the parties with the duty to accommodate.

The first point mentions the holders of the right to reasonable accommodation. It asserts, from the outset, that there is a fundamental right to reasonable accommodation. This can be inferred not only from the terms of the UN/CRPD, but also from the concept of discrimination. Therefore, if there is no undue burden, the absence of reasonable accommodation – if one exists – constitutes discrimination. There is a right to not be discriminated against, correlated with a duty to not discriminate. By consequence, there is a fundamental right to reasonable accommodation, provided one exists, up to the point of undue burden.

According to the UN/CRPD, the holders are people with disabilities, from the perspective of a social model combined with biomedical components. Therefore, all persons with disabilities are entitled to reasonable accommodation, provided they have the skills, qualifications, licenses, etc. that are necessary for the job, task or activity for which they are claiming accommodation.⁶ For example, blind people cannot – yet – demand reasonable accommodation to work as bus drivers, since they do not have the required skills and licenses. But they could demand reasonable accommodation to work as a public prosecutor, provided they pass the exam and have the proper university qualifications. On this matter, there is one important caution. Disability sometimes rules out some activities entirely, since no reasonable accommodation is available to offer alternatives. However, there are countless examples in which it is ingrained but unfounded assumptions that lead people to believe that a person with a given disability cannot perform certain activities. It should not be forgotten that one of the purposes of reasonable accommodation is precisely to open up to review the practices, assumptions and methods that have molded our environment.

Attention needs to be paid to the possibilities of under-inclusion and over-inclusion (on the topic, (STRUCHINER, 2005, p. 147 and following). In the rulings mentioned in the previous sections, it has been noted that a debate has emerged on the boundary between sickness and disability, and on the possibility of primary care takers being entitled to reasonable accommodation. This issue should be explored on a case-by-case basis, to avoid both under-inclusion and a potential over-inclusion (e.g., by equating to disability the very temporary needs resulting from a sickness), which could distort the concept and make it unrealizable. In the meantime, cases of inclusion of individuals not considered persons with disabilities under the biomedical model, but deemed as such by the new concept of the UN/CRPD, will be unusual to begin with, since many infra-constitutional norms amount to under-inclusive rules from the perspective of the UN/CRPD.

Concerning the components of the term “reasonable accommodation,” it is understood that accommodation refers to all the modifications, adjustments, adaptations and even flexibilities to be carried out in the material and normative environment in which it is claimed, through the employment of a wide variety of mechanisms, from techniques, to technologies to a review of procedures and even exceptions in working hours and the workplace as well as in the performance of tasks, academic activities, etc.

The problem lies in the meaning of “reasonable.” Humberto Ávila has researched STF case law and found three senses of the term: (a) reasonableness as fairness: a guideline that requires that general and abstract norms be judged in light of peculiarities of the case, establishing conformity between the norm and the specifics of the case. It would suggest private justice. (b) reasonableness as congruence: a guideline that requires conformity between the legal norms and the external conditions of application, requiring a real cause to justify adoption of the measure; (c) reasonableness as equivalence: a guideline that requires a relation of equivalence between two measures, the adopted measure and the criterion that gives its dimension (e.g., the punishment and the act committed) (ÁVILA, 2009, p. 156 and following). Reasonableness is also applied as that which is ordinary or routine.

In a review of foreign cases, it appears that in the U.S. the word “reasonable” limits accommodation to what is regarded as common sense to demand of someone. In Canada, meanwhile, the interpretation was quite different, and “reasonable” took on a broad connotation as all possible efforts to accommodate up to the point of undue burden. In the European context, there were three meanings, in particular one that considered the term as what is effective and for the individual or group in question. Which conception, then, is best suited to Brazilian constitutionalism? From the outset, “reasonable” should not be read as what is ordinary since this undermines the root of the purpose of accommodation, which is to offer alternatives, usually through an exception to what is common. It would be incongruous; it would be to interpret a fundamental right in the most restrictive way possible, in conflict with the recommended technique. Among the senses identified by Ávila (2009), it is believed that reasonableness as fairness is what most resembles reasonable accommodation, since it requires the molding of the general and abstract normto the specifics of the case. In this first sense, the term “reasonable” shall only limit accommodation if the accommodation is not at all efficient.

Furthermore, when judging what is “reasonable,” the sense of reasonableness as congruencecould also prove useful, since it would determine whether exceptions can be made to a measurein accordance with the causes that underpin it (e.g.,when the disability and the accommodation in questionreduce,below the minimum level, the degree of safetyintended by the measure). “Reasonable” would then be another limitation of accommodation, i.e., there would be general norms that would not permit some exceptions, because they would result in significantly diminishing their purpose (e.g., health, safety, equality). This assessment would resemble one of the stages of the proportionality test, of necessity, which analyzes whether the chosen means have the least impact on fundamental rights, according to the proposed ends. Nevertheless, the result would not necessarily be the customary invalidation of the entire legal act (e.g., declaration of unconstitutionality or illegality…), but the proposal of intermediary solutions, such as can occur in the so-called intermediary decisions (on the topic, CERRI, 2001, p. 84 and following.).

Although reasonableness as congruence is one of the senses of “reasonable,” it is proposed that this determination be made against undue burden, since “reasonable” and “undue burden” form a dual concept to be considered together. The defense for non-accommodation lies in the undue burden it will cause, insofar as the analysis becomes easier and clearer in the assessment of the burden. It is recommended that the term “reasonable” be interpreted as what is effective to adapt so as to make the material and normative environment meet the needs of the person with disabilities with the minimum possible segregation and stigma, with attention to the specifics that make it permissible to loosen or make exceptions to generally applicable norms and practices.Effectiveness is not restricted only to political aspects; in contrast, it is extendable to less tangible aspects, such as avoiding stigma, humiliation and embarrassment. For example, in Vande Zande, reasonableness was considered only from the practical angle, ignoring the sense of inferiority the employee felt by using the bathroom sink for her eating activities. To reflect on this point, imagine the opposite situation: if Vande Zande had to use the collective kitchen sink to brush her teeth or to wash her hands after using the toilet. How would her colleagues have felt?

To achieve reasonable accommodation, the dialogue procedure is imperative. It is incumbent on the person with disabilities or the representative entity to identify the best alternatives, since they have the knowledge and experience regarding the barriers to be overcome and the most effective ways of doing so; it is the application of the motto “nothing about us, without us.” It is important for all those involved – often second parties, such as student or work colleagues – to participate, at least for clarification purposes, except in situations that require confidentiality (e.g., mental disorders). Accommodating parties need to be open to dialogue and provide proof of undue burden.

Undue burden is the defense that allows the accommodating party to excuse itself and it shall be defined on a case-by-case basis, bearing in mind the valuable suggestion of the Supreme Court of Canada that while there will always be some burden, it may not be undue. For this reason, a method such as the de minimis test should be discarded, since it only considers the extent of the burden, ignoring the textuality. There are a range of factors to consider in determining whether a burden is undue and they can be split into two groups. The first is the purpose of the general measure for which an exception is being sought through accommodation. If the purpose of the measure is significantly hindered or undermined, the burden will be considered undue. It will not be sufficient to demonstrate that the measure was implemented in good faith, impartially or equally. The defense will only be complete if it can be shown that the accommodation obstructs the intended purpose.⁷

The second is a detailed comparison of costs and benefits. It has been mentioned that costs and benefits are not only economic and financial in nature (although these obviously enter into the calculation, together with the size of the accommodating party). On the benefits, it is worth recalling that accommodation is not intended to benefit just one individual; its raison d’êtreis far vaster, consisting of a dynamic model of accommodation (supra, EMENS, 2008, p. 894). It includes direct and indirect benefits, taking into consideration first and second parties, and also third parties. On the costs, attention should be paid to those that are mitigated by compensations or gains to the accommodating party, which could include incentives, exemptions and state immunities, or even gains from marketing social responsibility. The costs of stigmatization and humiliation are weighed up, as are the costs to second parties, when applicable (e.g., work colleagues).

Finally, it needs to be determined who should accommodate. For this, it is necessary to reiterate the differences in the concept of discrimination according to the UN/CRPD: (a) it includes, in addition to discrimination in its traditional forms, adverse impact discrimination; (b) the denial of reasonable accommodation constitutes discrimination, up to the point of undue burden. It is indisputable that the State, in all its ramifications, has a duty to accommodate. Private parties are included as well when there is a connection with fundamental rights. On this point, there are difficulties. It is believed that all private entities that perform functions via concession, license, etc. have this duty, as do private sector organizations providing services that are public in nature. For example, telephone companies, public transport operators, private schools and universities have this in common. There is no doubt that organizations that rely on public money, even indirectly, are included (e.g., social services, foundations, NGOs). In the labor market, the Brazilian Constitution (BRASIL, 1988) states that employers have a duty to uphold fundamental rights, making it possible and justifiable to include the terms of the UN/CRPD, paying particular attention, however, to the capacity of each employer with regard to undue burden. Union organizations also have the duty, primarily when the discrimination results from their agreements and conventions. Moreover, guidelines apply linking the private sector to fundamental rights (on the topic, SARMENTO, 2005).

Ideally, the legislative process is the way to define the concept, but it is believed that until new legislation on accommodation is forthcoming, it is up to the courts to rule, on a case-by-case basis, on who deserves accommodation, who has the duty to accommodate, what constitutes reasonable accommodation and what counts as an undue burden, when dialogue between the parties involved is unsuccessful. When examining these cases, it is important to try to promote this dialogue in a judicial setting. If this is not possible, the decision needs to observe the terms of the UN/CRPD, i.e, interpretation in accordance with the Constitution of the existing legal norms for the protection of the rights of people with disabilities.

5. Conclusions

In order to outline some proposals on how to interpret the dual concepts of reasonable accommodation and undue burden in Brazilian constitutionalism, the main aspects of the topic in other legal systems have been revisited, with an emphasis on those that gave rise to these concepts. This review contributed to a constitutional dialogue, permitting identification of the strengths and weaknesses of the constitutional and international reading.

The following conclusions can be drawn:

(a)  reasonable accommodation is a concept that modifies the legal content of discrimination, which is now present if reasonable accommodation is not provided to the point of undue burden;

(b)  people with disabilities, from the perspective of a social model combined with biomedical elements, are holders of a fundamental right to reasonable accommodation to the point of undue burden in a wide variety of environments. Other parties can become holders of the fundamental right to reasonable accommodation, to the point of undue burden, and their inclusion should be analyzed on a case-by-case basis;

(c)  accommodation consists of modifications, adjustments, adaptations and even flexibilities in the material and normative environment in which it is claimed, through the employment of a wide variety of mechanisms;

(d)  “reasonable” is the effective accommodation for the individual or group, and the idea of effectiveness includes the prevention and elimination of segregation, humiliation and stigma;

(e)  reasonable accommodation must be the product of a process of dialogue between the parties involved;

(f)   the defense of reasonable accommodation is that it will cause an undue burden. In essence, the burden will be undue when: (g.1) adopting an accommodation excessively undermines the purpose of the general measure, posing risks to safety, health and well-being etc.; (g.2) in the balance of costs and benefits, the accommodation proves to be too expensive. Note that the cost-benefit analysis is not limited to financial aspects, nor does it only consider the two parties directly involved.