The Harvard Law School Human Rights Program is exploring in a comparative and cross-disciplinary manner the concept of indirect discrimination on the basis of religion through a workshop in April 2020, hosted in cooperation with the Harvard Law School Project on Disability, the Columbia Law School Human Rights Institute, and the Harvard Human Rights Journal. The Journal’s website will be publishing a series of short essays reflecting views of the participants leading up to the workshop to spark thoughtful discussion.
What is the Right against Discriminatory Impact Based on Religion?
Gerald L. Neuman
The pursuit of equality in international human rights law includes both prohibitions of intentional discrimination and prohibitions of practices with discriminatory impact. The latter category, often designated as “indirect discrimination,” raises numerous questions that have not been fully explored. One important subset of those questions relates to indirect discrimination on the basis of religion.
Indirect discrimination norms do not prohibit actions with differential impact in absolute terms, but rather require that sufficient justifications are provided for those actions. Some of the questions relate to the kind or degree of justification that must be provided. Others relate to the showings of differential impact that must be made to prompt the inquiry into justification.
Take for example the old practice of Sunday closing laws in countries where the majority of the population is Christian. These laws facilitated a day of worship or leisure for many shopkeepers, but they posed an incidental economic disadvantage to others who observed a different day of the week as a day of rest.[] Assuming that they serve a valid secular purpose, how should their indirect impact be evaluated?
Or take the newer laws in some European states that regulate appearance in public with the face fully covered, which have been controversial within and among human rights tribunals.[] For some Muslim women, these laws expose them to criminal or civil sanctions as a result of their religious practice. If they are viewed as government action with discriminatory impact, how should these laws be analyzed? Does it matter that only a small minority of Muslim women believe in fully covering the face? Does it matter whether other religions in the particular country also believe in fully covering the face, or how many people have secular reasons for fully covering the face? Should the analysis consider the impact on members of a particular religion, or on holders of a specific religious belief? What empirical evidence should be submitted to support a claim of indirect discrimination, and what kind of justification can the state offer that would outweigh a demonstration of differential impact?
Moreover, the law against face covering could be analyzed instead as an interference with the right to religious freedom rather than as a denial of equality. Various human rights treaties guarantee the right to manifest religion or belief in practice. Viewed in that light, should the analysis be the same, or is the impact of the prohibition on other believers and nonbelievers irrelevant, and is the level of justification required to outweigh the religious freedom claim the same or different?
Some claims of indirect discrimination involve isolated situations and do not reflect a broader pattern of systematic disadvantage for the members of the relevant religion or holders of the relevant belief. Some claims of indirect discrimination involve disproportionate consequences that flow from prior direct discrimination, or even from intentional but permissible differentiation.[] Should the characterization of the causes affect the analysis of how the differential impact could be justified?
These questions, among others, can be posed in the abstract or in relation to particular legal systems. Under the U.S. Constitution, the Equal Protection Clause does not include a prohibition of actions with discriminatory impact based on religion (see Washington v. Davis (1976)), but certain statutory provisions applying to certain fields of activity, such as employment and housing, do. Currently, the Free Exercise Clause of the First Amendment does not regulate unintended government interference with religious practices by neutral laws, but statutes like the federal Religious Freedom Restoration Act do, and may provide stronger protection than the statutory antidiscrimination norms would with regard to the same kind of interference.
European human rights law includes prohibitions of both direct (intentional) and indirect governmental discrimination on the basis of religion or belief, within the scope of European Union law and Article 14 of the European Convention on Human Rights.[] An EU directive regulates private discrimination (direct and indirect) on grounds of religion or belief in employment, but so far the EU has resisted efforts to enact a broadly phrased directive that would regulate, inter alia, private discrimination on several grounds, including religion or belief, in education, housing and access to other goods and services. Directives can be more detailed than generally worded treaty norms tend to be, and can list exceptions to an antidiscrimination rule rather than leave the situations they cover to case-by-case adjudication.
At the global level, Article 26 of the International Covenant on Civil and Political Rights (ICCPR) guarantees equal protection of the law and requires states to protect everyone against discrimination “on any ground such as race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth or other status.” The Human Rights Committee, the treaty body that monitors compliance with the ICCPR, interprets this obligation as applying to both direct and indirect discrimination by both public and private actors. The Human Rights Committee is in the habit, when it reviews states’ reports on their compliance, of recommending the enactment of comprehensive antidiscrimination legislation that covers all the types of discrimination addressed by Article 26.[] The Committee recognizes that practices with differential effect based on a covered ground are not absolutely prohibited, but rather they must be reasonable and objective and serve a legitimate purpose;[] in this regard, reasonableness includes an inquiry into proportionality.[] The European Court of Human Rights similarly understands the concept of indirect discrimination as involving the absence of “a ‘reasonable relationship of proportionality’ between the means employed and the aim sought to be realized.”[] The Inter-American Court of Human Rights has asserted that there is a jus cogens rule of international law that prohibits all forms of discrimination, both direct and indirect, by all actors, public and private.[]
One might then ask, what sort of comprehensive anti-discrimination legislation should states enact, and how should it apply to indirect discrimination on grounds of religion? Should all states adopt a generally phrased prohibition on public and private practices that have disproportionate effect on a religious group or the holders of any religious belief, and that lack a legitimate and proportionate justification? Should application of this general standard be left to case-by-case adjudication without further legislative guidance? Or should more specificity be given with regard to proportionality and the normative weight that attaches to various factors in its evaluation? Does the “weight” attached to indirect discrimination differ from the “weight” attached to direct discrimination, and if so how? Should the evaluation be conducted in the same way in all fields of public and private activity, or should indirect discrimination be more strictly regulated in certain contexts? Should the prohibition of indirect discrimination be truly comprehensive and exceptionless, or is there room for statutory carve-outs to accommodate traditions or the rights of others?
With these questions in mind, one might further ask how much variation in the answers from state to state is appropriate. May or should states concentrate on different fields for regulation, or enact different exceptions? May or should they adapt their legislation to particular patterns of systematic disadvantage in their societies? Or do universal rights require uniform legislation?
Evidently legal systems vary in how they treat these issues, and my own impression (subject to persuasion otherwise) is that some variation is appropriate. Leaving all issues of indirect discrimination open for case-by-case adjudication is unfair to both complainants and defendants and does not provide an effective means of implementation for the norm. Societies should not strive to eliminate every conceivable disadvantage that results from any religious belief, and the contexts in which disadvantage is most urgent may depend on local conditions.
If that perspective is correct – which remains to be discussed – then at the global level, human rights bodies should not insist that all states pursue the same model for regulating indirect discrimination, but should monitor the suitability of each state’s legislation to the problems that it faces. In reviewing individual cases that have already been before national courts, global human rights bodies should not assume that each case should be decided as the global body would have decided it in the first instance, but should examine the reasoning that produced the prior decision. At the regional level, the European Court of Human Rights would be justified in affording a certain margin of appreciation – perhaps narrower than what the Court currently provides, but still not reduced to zero.
These are some opening thoughts to whet the reader’s appetite for the series of essays that follows. They should make clear that there is a lot discuss.
 See Braunfeld v. Brown, 366 U.S. 599 (1961) (finding no violation of the Equal Protection, Free Exercise, or Establishment Clauses).
 A majority of the European Court of Human Rights upheld the well-known French law in S.A.S. v France (ECtHR 2014) (Grand Chamber), over a dissent echoed more recently in the concurring opinion in Dakir v. Belgium (ECtHR 2017). The Human Rights Committee, in contrast, found that the same French law violated both freedom of religion and the prohibition of discrimination in Yaker v. France (Human Rights Committee 2018), while leaving ambiguous whether the law involved direct or indirect discrimination.
 To illustrate this last point with a nonreligious example, in Taddeucci and McCall v Italy, (ECHR 2016), the European Court held that although Italy’s nonrecognition of same-sex marriage did not violate Article 14 (in conjunction with Article 8) at the relevant time, its refusal to treat a same-sex couple as a family for immigration purposes because they were not married did amount to indirect discrimination on the basis of sexual orientation in violation of Article 14 (in conjunction with Article 8).
 There is also a broader prohibition of discrimination under Protocol No. 12 to the European Convention, but fewer than half of the states in the Council of Europe have ratified that Protocol.
 See, e.g., Concluding observations on the fifth periodic report of the Netherlands, UN Doc. CCPR/C/NLD/CO/5 (2019), para. 14 (recommending that the state’s legislation “Provides full and effective protection against discrimination on all the prohibited grounds under the Covenant in all spheres, including the private sphere, and prohibits direct, indirect and multiple discrimination”); Concluding observations on the first periodic report of Pakistan, UN Doc. CCPR/C/PAK/CO/1 (2017), para. 12.
 General Comment No. 18: Non-discrimination (1989), para. 13.
 See Ory v France, communication no. 1960/2010 (2014) (separate opinion of Mr. Fabián Omar Salvioli).
 E.g., D.H. v Cyprus (ECHR 2007) (Grand Chamber), para. 196; Di Trizio v Switzerland (ECHR 2016), para. 91.
 See, e.g., Case of the Mapiripán Massacre v. Colombia, (No. 134) (Inter-Am. Ct. H.R. 2005), para. 178; Juridical Condition and Rights of the Undocumented Migrants, Advisory Opinion OC-18/03, (Inter-Am. Ct. H.R. 2003).