For whom is the house kept? Making ICESCR work to counter discrimination against migrants and refugees in access to adequate housing

Abstract

With a critical focus on the interpretative work of the Committee on Economic, Social and Cultural Rights, this article sets out the parameters of applying non-discrimination under article 2(2) of the International Covenant on Economic, Social and Cultural Rights in conjunction with article 11(1) on the right to adequate housing to discrimination suffered by migrants and refugees in access to existing housing. By placing a specific emphasis on understanding states parties’ obligations under the Covenant, the article clarifies the standards applicable for examining discrimination against migrants and refugees in access to housing through a structured account. This account contributes to existing scholarship both on the Covenant in general, and the right to adequate housing in particular. Offering a methodical explanation of how state accountability could be secured to remedy the harms of discrimination against migrants and refugees in access to housing, the article shines light on the normative value and potential of the Covenant in this context.

Introduction

Imagine that a refugee with a temporary residence permit is excluded from queueing for social housing because of a local authority’s rules that limit this possibility only to permanent residents and nationals. ¹ Then imagine that a private housing provider puts an ad on their website, with the note ‘no migrants please’, and the relevant authorities take no action against this conduct. Assuming that the International Covenant on Economic, Social and Cultural Rights ² (ICESCR, the Covenant) is applicable in both instances, how would the examination under the Covenant go?

This paper answers this question. With a focus on the interpretative work of the Committee on Economic, Social and Cultural Rights (CESCR, the Committee), the paper sets out the parameters of applying ICESCR article 2(2) (non-discrimination) in conjunction with article 11(1) (right to adequate housing) to discrimination suffered by migrants and refugees in access to existing housing. Such clarification is needed since, as shown below, it is missing in the current state of the law under the Covenant.

The paper’s focus on access to existing housing means that discrimination in the supply of housing or during occupation, including discriminatory evictions, are not part of it. ³ Throughout the analysis, specific emphasis is placed on understanding states parties’ obligations under the Covenant. ⁴ This is a conscious choice, motivated by 1) ‘the raison d’être of the Covenant which is to establish clear obligations for States parties in respect of the full realization of the rights in question’, ⁵ and 2) the urgency to understand the structure of state accountability for discrimination in access to housing given the stark lack of accountability despite the pervasiveness of this problem. ⁶ This also means that there is a need to point out the limits of such accountability structures amid the workings of migration law. ⁷

Investigating the legal structure of discrimination against migrants and refugees in access to housing has a practical and not only theoretical importance, given that migration is now considered as a structural factor in causing homelessness. ⁸ While contemporary data on the impact of discrimination on migrants’ and refugees’ homelessness and housing exclusion in comparative and comprehensive terms is hard to come by, ⁹ discrimination against these groups in the private and public rental sector and property sales is known to manifest itself in a myriad of overt and covert ways: discriminatory ads, exploiting migrants in need of housing for rent, landlords refusing to rent or sell to individuals with a foreign name or accent or immigrant background, discriminatory social housing allocation practices and so forth. ¹⁰ Discrimination is therefore a major hindrance before migrants' and refugees’ access to adequate housing. ¹¹ This makes it necessary to identify the legal venues by which such practices can be contested and states can be obliged to rectify them under international human rights law.

In what follows, first, the material examined in this analysis, the method by which it was chosen and analysed, and the terminology adopted is briefly explained. Then, the paper’s relationship with and contribution to existing literature is described. The main analysis then begins, which is structured around two main parts. Part I traces, to the extent relevant, the meaning of article 11(1) and article 2(2) and is more descriptive. Part II then turns to exploring the legal issues that arise from the application of article 2(2) in conjunction with article 11(1) and is more analytical.

With that in mind, let us start with setting out the ground rules for the following analysis.

Material, method, and terminology

As for material, the Committee’s outputs examined in this analysis are its general comments and views on merits in individual communications. The general comments adopted by the Committee are not binding; however, they are important sources for interpretation. ¹² The authoritative force of these general comments are stronger where the Committee’s interpretation aligns with the basic rules of treaty interpretation under Vienna Convention on the Law of Treaties (VCLT) articles 31-33, ¹³ and where it is coherent and transparent. ¹⁴ The Committee’s views on individual communications, adopted under the Optional Protocol to the ICESCR (OP-ICESCR), ¹⁵ are not strictly binding on states parties either. ¹⁶ However, I agree that the good faith obligations of states under VCLT article 26 require states parties to comply with the findings of the Committee in individual communications. ¹⁷ Indeed, the OP-ICESCR is an international treaty to which states become a party voluntarily and according to this instrument, the Committee is the body tasked with interpreting and applying the Covenant. An understanding of the Committee’s views as devoid of any legal effect would otherwise deprive the OP-ICESCR of its raison d’être. ¹⁸ Having said that, it is important to approach the Committee’s interpretations in these views critically, to identify and address inconsistencies and/or unclarities where necessary. ¹⁹

I used the United Nations (UN) Treaty Body Database to search for the relevant individual communications. ²⁰ Filtering the search page by ‘Committee’ (CESCR) and ‘Document Type’ (Jurisprudence), it is possible to view all the decisions of discontinuance, admissibility and views on merits issued by the CESCR on this website. Of the views on merits, I chose eight specific communications ²¹ to examine based on three criteria: 1) whether the communication says something notable regarding access to housing and states’ obligations in that specific relation, 2) whether the communication concerns non-discrimination, 3) whether the communication was brought by a migrant or refugee and/or examined/decided as concerning the rights of the latter. The views that meet at least part or, though rare, ideally all these criteria have been studied closely.

The analysis proceeded as follows. I used the interpretations offered by the Committee in its relevant general comments as the starting point, though they surely did not constitute the ending point. I then turned to the identified views on merits in individual communications issued by the Committee to see whether, and if so, how the Committee built on its initial interpretations. I read these views both chronologically and with a view to their relationship with one another in terms of their subject matter, while constantly keeping the general comments on the side. This reading was further aided by the relevant academic commentary. Doing so helped me identify the important blind spots in the doctrine, as well as unclarities, inconsistencies, and matters that merited further thought. I then set out to systematically explain the meaning of non-discrimination under the Covenant following on the ‘stylistic normativity’ of traditional legal scholarship, ²² by offering fresh interpretations where necessary, all with due regard being had to the basic interpretative principles of international law set forth in VCLT articles 31-33.

Lastly, as for terminology, the terms ‘migrant’ and ‘refugee’ are used in a broad sense throughout. Though an undefined term in international law, ‘migrant’ is used here to express persons who temporarily or permanently reside outside of their state of nationality for reasons such as family, study, economic opportunities etc. ²³ As for ‘refugee’, this term is used here to capture not only refugees as defined under the Convention Relating to the Status of Refugees, ²⁴ but anyone who has been granted or has a legal claim to international protection under international, regional, or national law. ²⁵ There is an ongoing debate as to whether ‘migrant’ is or should be used as an umbrella term to encompass refugees as a sub-category thereof as well. ²⁶ In the international arena, such an inclusivist approach is adopted by the International Organisation for Migration, ²⁷ while the UN High Commissioner for Refugees is strictly opposed to this and champions a residualist approach whereby refugees are defined completely distinctly from other groups of migrants on account of their specific protection needs and regimes. ²⁸ Here, neither a wholly inclusivist nor a strictly residualist approach is adopted. While the reference to ‘migrants’ and ‘refugees’ are kept separate throughout to acknowledge the fact that refugees form a distinct group due to their protection needs in international law, I acknowledge the complexity of human mobility which complicates these definitions ²⁹ and the important critique that the migrant/refugee binary is an ‘oversimplification’. ³⁰ The subtleties of this debate cannot be further explored here because of space limitations. At any rate, this is not instrumental to the following analysis, as the latter’s focus is on state obligations on a conceptual level and within the structure of the Covenant.

Relationship with and contribution to existing literature

This analysis builds on and contributes to a growing body of scholarship that examines the potential and shortcomings of international anti-discrimination guarantees of the UN human rights system for migrants and refugees. ³¹ These contributions either focus on differential treatment during entry ³² or upon entry, ³³ without prejudice to the close relationship between these two. This article is focused on discrimination against migrants and refugees while they are in the receiving state. In its approach and material chosen for the analysis, it draws from Briddick. ³⁴ In its aims, however, it differs slightly from Briddick’s work because the main purpose of this analysis is to methodically explain the application of the ICESCR to cases of discrimination against migrants and refugees in access to housing, rather than identifying the kinds of ‘normative overreach’ and ‘normative under-development’ in the CESCR’s work per se. ³⁵ Such an analysis is necessary to help the Committee ‘establish a clear understanding of its approach to questions of discrimination in order to give its pronouncements credibility.’ ³⁶ Surprisingly, a clarification of the legal standards pertaining to the application of article 2(2) within the Covenant’s complicated structure is largely missing in the otherwise rich literature on the ICESCR. ³⁷ The article therefore makes an important contribution to that field in this broader sense too.

Having clarified the main elements of this examination and its contribution, it is now time to commence the analysis with Part I by tracing the meaning of ICESCR article 11(1) and article 2(2) in the current doctrine.

Part I

The personal scope of article 11(1)

Given the lack of a personal scope clause in the Covenant, and the clear wording of article 11(1) which guarantees the right to ‘everyone’, ³⁸ migrants and refugees fall under the personal scope of article 11(1). ³⁹ Article 2(2) buttresses this right with a guarantee of non-discrimination. ⁴⁰

ICESCR article 2(3), which allows developing countries to limit non-nationals’ economic rights under certain conditions, is of no relevance in this context. This clause concerns economic rights, ⁴¹ while the right to housing is a social right. ⁴² Indeed, the object and purpose of the right to housing is not ensuring individuals a means of economic gain through housing towards other ends ⁴³ but ensuring adequate housing for all as an end in itself. ⁴⁴ Therefore, the right to housing cannot be withheld from non-nationals with reference to article 2(3) since it is outside the latter’s material scope. ⁴⁵

The material scope of article 11(1)

Article 11(1) famously provides that states parties ‘recognize the right of everyone to an adequate standard of living for himself and his family, including … housing, and to the continuous improvement of living conditions.’ The interpretation of this provision by the Committee is found in – the outdated ⁴⁶ – General Comment No. 4. ⁴⁷ A detailed overview of the material scope of article 11(1) cannot and need not be offered here. What needs to be clarified is 1) whether access to housing is part of this provision and 2) the kind of obligations that this right gives rise to.

First, according to General Comment No. 4, article 11(1) encompasses not only the right to enjoy an existing home and to be free from forced evictions, but also the right to have access to adequate housing. ⁴⁸ Correspondingly, there is a significant positive dimension of the right. It concerns more than individuals’ freedom from interference and requires states to ensure access to adequate housing, either directly as a housing provider or through other means in the public and private sector. ⁴⁹ This entails, inevitably, that the right has a (re)distributive component, the achievement of which is expected regardless of the economic and social system in each state party so long as it complies with democratic principles. ⁵⁰

Second, the obligation to guarantee the right to access housing (taken on its own, i.e., not in conjunction with the prohibition of discrimination), needs to be read in light of two fundamental provisions of the Covenant which encapsulate the nature of state obligations: article 2(1) and article 4.

Article 4 establishes the conditions which must be observed by states parties when limiting the Covenant rights. It therefore encapsulates the negative obligation to respect. Through its interpretation of article 4 in its eviction jurisprudence, the Committee has clarified that this provision essentially requires a proportionality assessment between the measures taken and the aim sought to be realised. ⁵¹

Article 2(1), encapsulating the positive obligation to realise, ⁵² is an infamously complicated provision. It provides that this obligation is mainly one of taking steps, to the maximum of states parties’ available resources, to progressively achieve the full realisation of the rights recognised in the Covenant by all appropriate means. The Committee has put emphasis on the process of realisation over time, rather than the endpoint of realisation, as encapsulated by the qualifier ‘progressive’. ⁵³ Yet ‘progressive realisation’ remains a painfully unclear type of obligation. ⁵⁴ General Comment No. 4 is opaque about how it ought to be understood in the context of the right to access housing. The growing jurisprudence of the Committee on article 11(1) is also of limited help, as it has concerned evictions rather than access to housing so far. Because of this, states’ positive obligations to guarantee access to housing for all, independently of an eviction context, remain rather unclear. ⁵⁵

On two occasions, though both concerning evictions and access to alternative housing in that context, the Committee has pronounced, in a general fashion, that ‘the lack of affordable, available housing is rooted in growing inequality and housing market speculation’, and that states parties have an obligation to resolve these problems to the maximum of their available resources. ⁵⁶ More specifically, in its views on Hernández Cortés and Rodríguez Bermúdez v Spain, the Committee found a violation of article 11(1) due to Spain’s failure to remedy the authors’ protracted situation of insecure housing. This communication concerned an eviction threat too; however, the Committee explicitly noted states’ positive obligation to realise the right to housing and based its finding on this obligation. ⁵⁷ Another notable exception is El Mourabit Ouazizi and Boudfan v Spain, where the authors’ claim of the unjustifiability of their potential eviction was examined by the Committee exclusively under article 2(1). ⁵⁸ Essentially, the Committee examined whether the authors were facing a risk of homelessness because of the authorities’ lack of action to secure them housing, or, because of the authors’ own conduct; i.e., their lack of diligence in searching for housing and making their housing needs known to the authorities. ⁵⁹ Observing that it was the latter, the Committee found no violation in this communication. It noted that in meeting their obligations under article 2(1), the least that the authorities could do is ‘to have a programme or plan designed to ensure the effective enjoyment of the right’, with ‘deliberate, concrete and targeted measures’ taken ‘towards the fulfilment of the right … within a reasonable time frame’. ⁶⁰ Importantly, the Committee remarked that states may establish ‘administrative procedures to facilitate the protection of the right to housing, including by requiring individuals to take certain administrative steps to inform the authorities of their need for help’, so long as these procedures do not have discriminatory effects or do not place ‘an excessive or unnecessary burden’ on individuals. ⁶¹

Importantly, none of these communications specifically concerned a migrant’s or refugee’s inability to access housing in a state party. The question that arises then is where migrants and refugees stand in relation to all this. Are they considered in any specific manner by the Committee? There is no specific mention of this group in General Comment No. 4, nor is there a general comment dedicated to migrants’ and refugees’ rights under the Covenant. There is the ‘Statement’ issued by the Committee in 2017. ⁶² Yet, in terms of the right to access housing for migrants and refugees, it says very little. ⁶³

Having said that, General Comment No. 4 does note that ‘due priority’ must be given ‘to those social groups living in unfavourable conditions by giving them particular consideration.’ ⁶⁴ It defines members of society who are ‘vulnerable and disadvantaged with regard to housing’ as ‘homeless persons and families, those inadequately housed and without ready access to basic amenities, those living in “illegal” settlements, those subject to forced evictions and low-income groups’. ⁶⁵ Although the Committee does not specifically note migrants and refugees here, it is clear that migrants and refugees can find themselves in one or multiple of these situations in the housing market. ⁶⁶ Alternatively, the requirement can be said to apply to migrants in a vulnerable situation ⁶⁷ and refugees as a vulnerable and disadvantaged group in a broad sense. Yet General Comment No. 4 does not articulate how housing discrimination may occur, let alone be addressed, for specific disadvantaged groups under the Covenant.

It therefore seems that a connection between 1) the fulfilment of the obligation to guarantee the right to access housing, 2) specifically for migrants and refugees, and 3) without discrimination seems lacking in the Committee’s interpretative work on article 11(1). Because of this, the added value of non-discrimination in the Covenant’s complicated guarantee of access to housing for these groups remains unclear. To see whether these shortcomings are addressed by the Committee’s work on article 2(2), the analysis now turns to the CESCR’s interpretation of non-discrimination.

Non-discrimination under article 2(2)

ICESCR article 2(2) provides that states parties ‘undertake to guarantee’ non-discrimination in the enjoyment of the Covenant rights. The Committee defines discrimination comprehensively in its General Comment No. 20 as ‘any distinction, exclusion, restriction or preference or other differential treatment that is directly or indirectly based on the prohibited grounds of discrimination and which has the intention or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, of Covenant rights.’ ⁶⁸ Not every differential treatment amounts to discrimination. So long as a differential treatment is shown to be ‘reasonable and objective’, the meaning of which will be explained in Part II, there is no discrimination. ⁶⁹

Article 2(2) prohibits discrimination on grounds of race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth or other status. ICESCR thus has an open list of grounds, ⁷⁰ and the Committee has adopted a ‘flexible’ approach to interpreting ‘other status’ to capture the dynamic nature of discrimination. ⁷¹ Among the unenumerated grounds which the Committee considers to fall within the scope of article 2(2) are disability, age, nationality, marital and family status, sexual orientation and gender identity, health status, place of residence, and economic and social situation. ⁷² In the context of access to housing, a variety of factors ranging from gender, disability, language, religion, nationality, race, and socioeconomic status can lead to discrimination, alone or most likely in intersection with one another, against migrants and refugees. ⁷³ For instance, in the first example provided at the beginning of this article, namely, the limitation of access to social housing only to nationals and permanent residents, there would likely be discrimination on grounds of nationality or immigration status or both, possibly also in intersection with race, depending on the context. The second example, i.e., a discriminatory housing ad in the private sector, represents a case of racial discrimination.

As for the types of discrimination under the ICESCR, the differentiation between direct and indirect discrimination remains relevant in the Committee’s understanding of discrimination as clarified in General Comment No. 20 ⁷⁴ and the Committee’s jurisprudence, ⁷⁵ even though there is not much guidance on how to make the distinction in conceptual terms. ⁷⁶ Procedurally, one importance of the distinction relates to the burden of proof, as the Committee has found in Trujillo Calero v Ecuador that once a prima facie case of disproportionate impact against an identifiable group is made, it is for the state to show that this does not constitute indirect discrimination on a specific ground. ⁷⁷ Substantively, however, the pertinence of the distinction is less clear. While, in certain jurisdictions, whether discrimination is direct or indirect may have an impact on either the justifiability of differential treatment altogether or on the level of scrutiny, ⁷⁸ the Committee has not adopted such an approach so far. In other words, that discrimination is indirect does not automatically mean that the discriminatory measure will be subjected to less (or more) scrutiny. ⁷⁹

Why indirect discrimination still matters then can be explained by the function it undertakes in identifying more insidious forms of exclusion in the housing context. Indeed, the disproportionate impact of seemingly neutral housing policies on migrants and refugees constitutes a significant form of housing exclusion suffered by the latter. ⁸⁰ Atrey insightfully remarks, in her examination of structural racism, that the assessment of disproportionate impact of such kind risks being absorbed by a utilitarian cost-benefit calculation which, in highly racialised societies, is likely to justify the ‘sacrifice’ of the rights of racialised groups like migrants and refugees. ⁸¹ Hence, distinguishing indirect discrimination is crucial to identify and redress structural and insidious forms of exclusion in the housing sector. Having said that, since the Committee has not yet established how specific grounds impact on the justifiability of indirect differential treatment, except for clarifying that intersectionality calls for strict scrutiny in general, ⁸² it remains to be seen whether the Committee will live up to the challenge of calling out such practices.

Discrimination under the Covenant must be both formally and substantively guaranteed. ⁸³ The Committee notes that this requires going beyond the application of a formal comparator test between similarly situated individuals and ‘paying sufficient attention to groups of individuals which suffer historical and persistent prejudice’. ⁸⁴ It also remarks that direct discrimination includes ‘detrimental acts or omissions on the basis of prohibited grounds where there is no comparable situation’. ⁸⁵ These statements indicate that, in General Comment No. 20, the Committee goes beyond a formalistic understanding of discrimination that is not constrained to a comparator test; although it does not go so far as declaring the test completely irrelevant. In fact, in López Rodríguez v Spain, concerning the right to social security of a prison inmate who lost his non-contributory disability allowance, the comparator test played an important role and the ‘inappropriate’ comparisons drawn by the author were part of the reason why the Committee found no violation of article 2(2) in conjunction with article 9. ⁸⁶ However, in its very recent views on J.T. and others v Finland concerning Sámi people’s right to take part in cultural life of the community, the Committee reiterated its earlier view that eliminating discrimination requires more than simply making comparisons. ⁸⁷ These different views create an instability regarding the application of the comparator test, although the Committee’s latest views can be read to the effect that when the ‘historical and persistent prejudice’ suffered by an identifiable group is clear, such as in the case of indigenous peoples, the Committee will not pay much heed to the necessity of comparison. Future jurisprudence will hopefully shed more light on this issue.

According to the Committee, lack of differential treatment where such treatment is necessary to achieve substantive equality may constitute discrimination as well. ⁸⁸ Also, the Committee takes discrimination in housing by private actors into account, although it does not explain the extent to which states are obliged to regulate these actors’ actions in this context. ⁸⁹

The most novel characteristic of the prohibition of discrimination under the Covenant is that the obligations that it gives rise to are ‘immediate and cross-cutting’. ⁹⁰ This entails that not only negative but also positive obligations to guarantee non-discrimination under the Covenant must be immediately realised. Yet earlier it was clarified that, principally, states’ positive obligations are subject to progressive realisation per article 2(1). This necessitates teasing out what the anti-discrimination obligations of states, which are of immediate nature, consist of and how they apply in conjunction with the principal obligation to ‘progressively realise’ the Covenant rights. Essentially, this means that the relationship, or the distinction, between article 2(1) and article 2(2) of the Covenant needs to be clarified. While this need has been observed in scholarship, there is no analysis that investigates the ramifications of such a clarification, but only expectations for clarification from the CESCR. ⁹¹ The following analysis therefore benefits both the ICESCR literature at large, and the right to access housing in particular.

With this in mind, the analysis turns (in Part II) to the novelties of applying non-discrimination in conjunction with article 11(1) to the benefit of migrants and refugees.

Part II

Applying article 2(2)

Let us revisit the two scenarios introduced in the beginning of the paper. In scenario 1, a refugee with a temporary residence permit is excluded from queueing for social housing because of a local authority’s rules that limit this possibility only to permanent residents and nationals. In scenario 2, a private housing provider puts an ad on their website, with the note ‘no migrants please’, and the relevant authorities take no action against this conduct. What sort of state conduct gives rise to discrimination in these instances? This is the primary question that needs to be answered to apply the Covenant correctly in such instances of discrimination. The first section of this part will explain this. The second and the third sections will then explore the legal standards applicable depending on the characterisation of the discriminatory conduct.

Characterising state conduct correctly

A discriminatory conduct contrary to article 2(2) can be characterised in two ways; either as a failure to realise access to housing without discrimination or as a limitation of the right to access housing. Scenario 1 presented above is an example of discrimination in access to housing against a refugee in the form of a formal limitation. ⁹² Scenario 2, on the other hand, exemplifies a situation where a state fails to realise migrants’ access to housing without discrimination by acting against the conduct of a private housing provider.

Having said that, categorising discrimination as occurring due to either one of these two conducts may not always be so straightforward. Think about a third scenario, where a social housing queue which distributes housing solely based on one’s queue time may leave recently arrived migrants and refugees at a great disadvantage. A policy that is neutral at face value which disproportionately impacts a certain group may be indirectly discriminatory. ⁹³ Should this third scenario be discussed as a limitation by the state, in that it introduces a criterion of access that does not pay sufficient attention to the needs of a specific group? Or should it be discussed as a failure to act by the state to remove differential treatment that arises because of a facially neutral standard?

Since the main cause for concern here seems to be a neglect for the special needs of migrants and refugees, it seems more plausible that this constitutes a failure to remove differential treatment by the state. However, a substantive examination of the discriminator’s reasoning ⁹⁴ as to what factors were considered and how in the performance of the discriminatory conduct may reveal that the aim was in fact excluding migrants and refugees from access to social housing in a seemingly indirect manner, thus preventing them from enjoying this right via state conduct. Then, it would be more apt to characterise this conduct as a (direct) discriminatory limitation. ⁹⁵

Why does the determination of the exact character of a potentially discriminatory conduct matter? Given the nature and structure of state obligations set forth in Part II of the Covenant, it determines which standard of assessment would need to be applied in a complaint. Let us explore these standards further.

Discriminatory conduct as a limitation of the right to access housing

In our first scenario, the state treats refugees differentially from permanent residents and nationals by limiting their access to social housing through an eligibility criterion. In such a case of limitation, the standard of assessment is the following: if the differential treatment is not shown to be objective and reasonable by the state, it amounts to direct discrimination.

According to the Committee, the ‘objective and reasonable justification’ test comprises of establishing 1) the legitimacy of the aim of a measure, 2) its compatibility with the nature of the Covenant rights, 3) that it is solely for the purpose of promoting the general welfare in a democratic society, and 4) that there is a clear relationship of proportionality between the measure and the aim sought. ⁹⁶ This is essentially the standard found in ICESCR article 4, despite the Committee’s failure to refer to this provision explicitly in General Comment No. 20. ⁹⁷ However, the test articulated by the Committee does not mention the requirement of a limitation being ‘determined by law’. ⁹⁸ This should not be read as the requirement of lawfulness being excluded from the assessment of a potentially discriminatory limitation. Indeed, the Committee’s Statement on the duties of states towards refugees and migrants can be read as a confirmation to that effect, as it remarks that differential treatment ‘should be in accordance with the law […]’. ⁹⁹

Regarding legitimate aims, a duality is noticeable in General Comment No. 20. The Committee first says that differential treatment should pursue a legitimate aim. It then states that differential treatment should be solely for the purpose of promoting general welfare in a democratic society. The separate mention of these two criteria may lead to the interpretation that a differential treatment can pursue any legitimate aim, i.e., legitimacy need not be understood as limited solely to the purpose of promoting general welfare in a democratic society. The standard referred to in General Comment No. 20 is therefore confusing as it could be construed as widening, without a clear basis, the scope of the aim that could be cited as legitimate by states. This is slightly problematic because the existence of a single or multiple legitimate aim(s) directly affects the application of the justification test, as shown below. Relatedly, the ordinary meaning of ‘promoting general welfare in a democratic society’ does not have much to do with public order and national security, an aim that can be cited by states in the migration context, but more with increasing the overall living standards. ¹⁰⁰ Importantly, the plain text of article 4 mentions only this aim, i.e., promoting general welfare, as legitimate. ¹⁰¹ Therefore, in accordance with the plain text of the Covenant and the ordinary meaning thereof, the legitimacy standard should be understood as referring solely to the purpose of promoting general welfare in a democratic society when assessing a differential treatment.

As for the proportionality test, the Committee has clarified the meaning of this test in its eviction jurisprudence. Accordingly, the Committee has noted that a limitation must be necessary, i.e., it must be the least restrictive measure available, and the benefits of the limitation in promoting the general welfare must outweigh the impact of the limitation on the enjoyment of the right. ¹⁰²

In light of these clarifications, the key then is ascertaining when differential treatment against migrants and refugees is justifiable and when it amounts to discrimination. Importantly, this must be conducted in light of two constraints inherent to social rights as human rights: 1) the justifiability of differential treatment in the migration context with reference to the need to promote general welfare in a democratic society, which may very well include economic considerations and the need to allocate ‘scarce' resources ¹⁰³ 2) in view of the discretion afforded to states in determining their housing policy.

These constraints, though significant, are not insurmountable. In relation to the first constraint, the specific understanding of the sole legitimate aim as explained above is crucial. ¹⁰⁴ This would require a careful scrutiny of broadly framed migration policy-related aims that might be cited by states, ¹⁰⁵ with a view to understanding whether a specific measure does indeed promote general welfare, and not the welfare of a select majority, in a democratic society. ¹⁰⁶ Even when economic justifications are put forward in the name of promoting general welfare, however, the Committee can and should ‘determine the validity of the differential treatment of aliens by assessing whether it is motivated by sound economic reasons or mere prejudice.’ ¹⁰⁷ Aims such as the need to allocate ‘scarce' resources should be taken very seriously and judged in light of the prevalent socio-economic situation of a respondent state. ¹⁰⁸

In relation to the second constraint, the amount of discretion left to states in designing their housing policy is not endless. However, determining the extent of this discretion is important, as it will shape how the proportionality assessment will take place. The Committee does not have a clear set of principles established in this respect. For example, it is not clear how the specific ground of differential treatment might affect the discretion afforded to a state, and there is no ‘suspect ground’ doctrine developed for now. The only consistent remark made by the Committee, although not in the context of discrimination but generally in relation to article 4, is that ‘the more serious the impact on the author’s rights under the Covenant, the greater the scrutiny that must be given to the grounds invoked for (…) a limitation.’ ¹⁰⁹

Having said that, there are certain factors that may require stricter scrutiny. In Trujillo Calero v Ecuador, concerning the right to social security without discrimination of a woman who was an unpaid domestic worker and a voluntary affiliate to a social security scheme, the Committee noted that

the author is an older person who is in a critical economic situation and has health problems and that the intersection of the alleged gender and age discriminations makes her particularly vulnerable to discrimination in comparison with the general population. This means that particularly special or strict scrutiny is required in considering the question of possible discrimination. ¹¹⁰

In López Rodríguez v Spain, mentioned earlier, the Committee came to the same conclusion because the author was ‘a person with a disability and … also deprived of his liberty, and … therefore at greater risk of discrimination than the population at large.’ ¹¹¹ It appears that the Committee pays close attention to the intersectional nature of the disadvantage suffered as a result of the multiple groups to which a person belongs. ¹¹² This is important for migrants and refugees because it means that their disadvantaged position in the housing market due to their race, ethnicity, nationality, the type of residence permit they hold, likely in intersection with other characteristics such as socioeconomic status, gender etc. ¹¹³ requires strict scrutiny by the Committee in a discriminatory limitation claim. As argued elsewhere, ¹¹⁴ an important communication which represents a missed opportunity for the Committee to clarify exactly this point (in specific relation to the intersection of migration status and socioeconomic status) is Infante Díaz v Spain which concerned the exclusion of an undocumented migrant from access to social housing as an adequate housing alternative against the background of an eviction threat. ¹¹⁵

Lastly, another factor that would require very strict scrutiny by the Committee would be differential treatment via an interference with the enjoyment of the minimum core of the right to housing. ¹¹⁶ The minimum core of the right to housing is a controversial matter which I cannot discuss here. What needs to be clarified, however, is that article 2(2) is a command that guides the operationalisation of the Covenant rights. As such, it does not have a core itself. This understanding is supported by a systematic interpretation of the Covenant given that article 2(2) is in Part II of the Covenant, rather than Part III which sets out the substantive rights. Non-discrimination obligations therefore also pertain to the minimum core of the right to housing and complement it. ¹¹⁷ Accordingly, any differential treatment that limits migrants’ and refugees’ enjoyment of the minimum core of the right to housing must be subjected to the strictest scrutiny possible.

Discriminatory conduct as a failure to realise access to housing without discrimination

In our second scenario, where a private housing provider puts an ad on their website, with the note ‘no migrants please’, and the relevant authorities take no action against this conduct, the discriminatory conduct is in the state’s failure to act, to realise migrants’ right to access housing without discrimination. In such cases, ascertaining the relevant standard of assessment depends on resolving another legal issue: the clarification of the relationship/distinction between article 2(1) and article 2(2).

Consistently with the understanding of non-discrimination obligations as immediate, the application of article 2(2) excludes article 2(1) for the purposes of ascertaining and assessing states’ obligations to realise access to housing without discrimination. This means that under the ICESCR, article 2(2) is not a mere temporal exception to article 2(1), but a separate source of positive obligations in the specific context of discrimination. Accordingly, article 2(2) must be distinguished from article 2(1). ¹¹⁸ Despite the long-standing understanding of non-discrimination obligations as immediate, the Committee has not clarified the distinction between these provisions explicitly in this manner so far. Nor has the ICESCR literature delved much into the repercussions of such a separation. Yet for reasons explored further below, such clarification matters. However, before that, a note on the meaning of ‘immediacy’ of the obligations is due.

There is broad consensus that the elimination of de jure discrimination can be immediately achieved since states need to just abolish discriminatory laws and regulations and abstain from adopting them. When it comes to substantive or de facto discrimination, too, anti-discrimination obligations are immediate. ¹¹⁹ However, this is questioned in the literature by asking whether the elimination of this type of discrimination can be immediately achieved, given that it will have economic implications and require long-term investment and planning. ¹²⁰ Although this makes sense, that de facto discrimination may not be immediately ‘prevented, diminished and eliminated’ does not prevent the conclusion that state obligations in this sphere are of an immediate nature. Indeed, the results may not be immediately achievable; nevertheless, states cannot postpone taking action to prevent, diminish and eliminate discrimination as it occurs in different contexts in daily life. ¹²¹ General Comment No. 20 makes this point clear too, with its emphasis on the immediacy of the obligation to ‘adopt the necessary measures to prevent, diminish and eliminate’ the causes of de facto discrimination, rather than to immediately achieve the outcome of having eradicated de facto discrimination. ¹²² Such an understanding does not deny the fact that these efforts depend on resources either. Resource-dependency determines whether the lack of action to remove differential treatment is justifiable in a specific instance or not; not the nature of an obligation as such.

Following this clarification, I argue that the legal consequences of clarifying that article 2(2) gives rise to separate positive obligations independently of article 2(1) are threefold:

1. The standard of reasonableness continues to shape the assessment of whether a state has discharged its positive obligation to realise the right to access housing without discrimination in light of OP-ICESCR article 8(4).

2. Arguments concerning resource constraints remain relevant in the reasonableness assessment; however, they would need to be subjected to further scrutiny.

3. Retrogressive measures are excluded from the vocabulary of non-discrimination under the Covenant.

Each sub-section below will further expand on these points. ¹²³

The continuing relevance of reasonableness

First, reasonableness continues to be the standard of assessment in individual communications for determining whether a state has discharged its positive obligation to realise the right to access housing without discrimination in light of OP-ICESCR article 8(4). ¹²⁴ This is consistent with the wording of the provision, which states that ‘when examining communications …, the Committee shall consider the reasonableness of the steps taken … in accordance with part II of the Covenant.’ The reference here is not only to article 2(1), but to Part II of the Covenant to which article 2(2) belongs. ¹²⁵ Therefore, if a migrant or refugee argues that a state has failed to prevent discrimination against them in their access to housing, the Committee would need to assess the reasonableness of the steps taken by the state to that effect. Importantly, since article 2(1) would be out of the equation in such a claim of discrimination, any state attempt to play the progressiveness card would be invalid in the determination of reasonableness.

In a ‘Statement’ from 2007, the Committee established certain criteria to give substance to the reasonableness assessment in general. ¹²⁶ Two of these, i.e., the importance of the time-frame within which steps are taken and the need to consider the situation of disadvantaged and marginalised groups, are particularly pertinent for migrants and refugees. ¹²⁷ Additionally, the Committee requires evaluating whether a measure is discriminatory or not for when assessing the measure’s reasonableness. Discrimination, therefore, has been made internal to the assessment of reasonableness. An application of this understanding, though implicitly, can be traced in El Mourabit Ouazizi and Boudfan, noted above. Although this communication did not concern article 2(2), here, the Committee did not find that the procedures established by Spain in accordance with article 2(1) for the authors to access social housing or another housing alternative discriminated against them. ¹²⁸ Regrettably, the Committee did not explicitly refer to the reasonableness test here. However, its reasoning signals that it does not find it reasonable to expect a state to give someone housing when an individual hasn’t done enough to find housing themselves, as long as the administrative procedures that one must clear to that end are not discriminatory. ¹²⁹

The internalisation of non-discrimination to the reasonableness test partly makes sense; however, it raises the following question: if a measure’s reasonableness is to be assessed by considering, inter alia, whether it is discriminatory or not, how should the reasonableness of the steps taken to prevent discrimination in the first place be assessed?

The only article 2(2) communication examined in conjunction with article 11(1) by the Committee, i.e., Walters v Belgium concerning a Belgian national, ¹³⁰ gives certain indications in this respect. ¹³¹ This communication concerned the indirectly discriminatory consequences of a rigid legal framework of eviction on the socioeconomically disadvantaged elderly population. The main issue was the state’s failure to protect Mr Walters against his eviction which allegedly gave rise to indirect discrimination because of the relevant tenancy legislation in force. The Committee did not clarify the nature of the state’s conduct at issue here. ¹³² It has noted, however, that states should ensure that housing-related costs are proportionate to income levels, and that special measures to protect the elderly population should be adopted. ¹³³ It remarked that

(…) the possibility of a disproportionate impact of such a policy on the right to adequate housing of certain groups in vulnerable situations entails a double obligation for any State party that chooses such a regulatory framework. Firstly, the State party must establish a mechanism to monitor the impact of the application of the legal framework on the most vulnerable and marginalized groups […]. Second, the policy must include mechanisms and flexibility to ensure that the application of the legal framework does not have a disproportionate impact in certain cases. ¹³⁴

To what extent it is reasonable to expect a state to monitor the indirectly discriminatory effects of its legal framework on vulnerable populations and/or which mechanisms to prevent such disproportionate impact can be considered reasonable is unclear from these statements. However, it was enough for the Committee that no domestic authority took the disproportionate impact at issue into consideration, ¹³⁵ indicating a proceduralisation of the reasonableness assessment. The Committee also considered the existence of alternative ‘reasonable’ measures to reduce the impact of the legislative framework on the author, such as mediation, which were unexplored by Belgium. ¹³⁶ These findings, coupled with the factual background of the communication and the inadequacy of the alternative housing options offered to the author, led to a violation. ¹³⁷ From this, one can tentatively infer that in a more obvious, direct discrimination claim of the kind given in our example, it might be more straightforward (depending on the specific facts) for the Committee to find that the state did not do what it reasonably could to prevent discrimination against migrants.

Lastly, the Committee seems to have conflated the proportionality test and the reasonableness test in its work. ¹³⁸ The rationale behind this is not explained, and the Committee seems unclear on how and when exactly reasonableness should be applied differently from or together with proportionality. Indeed, when discussing the justifiability of an eviction for example, the Committee commonly refers to proportionality and reasonableness together. ¹³⁹ However, an eviction can either be a limitation of article 11(1) via direct state interference (i.e., state authorities evicting an individual) which would necessitate a proportionality analysis (article 4), or it could constitute a failure to realise the right via state omission (i.e., to prevent a forced eviction by a third party) which would necessitate a reasonableness analysis (article 2(1)). The Committee is not clear on this important distinction, and commonly refers to both article 2(1) and article 4 en bloc. ¹⁴⁰ Although some do not question this stance, ¹⁴¹ this is problematic because reasonableness is the standard set for assessing the steps taken to realise the Covenant rights in OP-ICESCR article 8(4), not for the justifiability of limitations. ¹⁴² It is not at all a given that these two standards will or should be always applied together.

The contention here is not that proportionality and reasonableness have nothing in common in terms of substance. Drawing from the proportionality analysis to give further content to the reasonableness standard is not problematic as such, or is perhaps even unavoidable. ¹⁴³ For instance, Young has suggested that proportionality (‘as principle, but not as structured test’) manifests itself through reasonableness in the Covenant. ¹⁴⁴ However, that this is done without any apparent justification by the Committee is problematic because it undermines the structure of the Covenant and creates conceptual confusion. This could be avoided with more sophisticated reasoning.

Assessment of resource constraints in the reasonableness analysis

The second consequence of understanding article 2(2) as encompassing a separate set of positive obligations than article 2(1) is that arguments concerning resource constraints remain relevant in the reasonableness assessment; however, the immediacy of article 2(2) obligations would necessitate their further scrutiny than what may be sufficient under article 2(1). ¹⁴⁵ This concurs with the Committee’s understanding in General Comment No. 20 that

A failure to remove differential treatment on the basis of lack of available resources is not an objective and reasonable justification unless every effort has been made to use all resources at the state party’s disposition in an effort to address and eliminate the discrimination, as a matter of priority. ¹⁴⁶

This makes the added value of non-discrimination under the ICESCR very clear. While a claim of lack of access to housing brought by a migrant or refugee solely under article 11(1) is susceptible to a potentially laxer scrutiny (due to article 2(1)), the trigger of article 2(2) in conjunction therewith immediately narrows states’ room for manoeuvre when arguing for lack of resources. Relatedly, the starting point of the assessment shifts significantly: when it comes to progressive realisation, the starting point is that rights are realisable if resources are there, while in non-discrimination which requires immediate realisation, the starting point is that differential treatment must be removed unless the lack of resources is to an extent that makes such a removal impossible despite every effort. The non-discrimination clause therefore provides migrants and refugees with an important venue in terms of advancing claims against states and demanding immediate action to address their housing needs.

It must be noted, however, that the viability of such immediate action by a state would differ between the two types of positive obligations to realise, namely, the obligation to protect against the conduct of third parties and the obligation to fulfil. This is because of the different degrees to which the two types of duties demand states’ resources. This means that, in the assessment of resource constraints, it would be necessary to pay attention to whether a measure contested as discriminatory is 1) one concerning the conduct of a third party against which the state remains passive, just like the scenario used in this paper, or 2) one concerning a state’s failure to fulfil the right to access housing by adjusting, for example, the distributive criteria used for the provision of housing (such as waiting times) to prevent discrimination. Achieving the latter is a more resource-intensive effort than the former. So, depending on the socio-economic context, a state’s failure to fulfil the right in such a manner may be justifiable on account of lack of resources even under strict scrutiny. ¹⁴⁷

No room for retrogressive measures

The third and last consequence of understanding article 2(2) as encompassing a separate set of positive obligations than article 2(1) is the exclusion of retrogressive measures from the vocabulary of non-discrimination under the Covenant. Retrogressive measures can be described as measures which constitute rollbacks in the advancement of the realisation of the Covenant rights. ¹⁴⁸ As such, they relate to article 2(1) rather than article 4. ¹⁴⁹ Then, arguing that non-discrimination obligations arise under article 2(2) independently of article 2(1) means that discrimination cannot be characterised as a retrogressive measure in violation of article 2(1). Instead, discriminatory treatment perpetrated by or attributable to the state constitutes an interference with the enjoyment of Covenant rights in violation of article 4. To be clear, a retrogressive measure can be discriminatory; however, discriminatory treatment in the enjoyment of socioeconomic rights cannot be characterised as retrogression. This means that it is not possible to justify discrimination against migrants and refugees in access to housing as a necessary retrogressive measure, i.e., as a necessary setback in the realisation of the right, even in times of economic crisis. Rather, even then, differential treatment which hinders access to housing should be subjected to a proportionality assessment in accordance with article 4.

The clarification of this point is important for migrants and refugees for purposes of legal certainty. Indeed, when exactly retrogressive measures constitute a breach of states’ obligations under the Covenant remains relatively unclear. ¹⁵⁰ Although the Committee’s main stance on retrogressive measures is that they ‘would require the most careful consideration and would need to be fully justified by reference to the totality of the rights provided for in the Covenant and in the context of the full use of the maximum available resources’, ¹⁵¹ the standard by which the justifiability of retrogressive measures is to be assessed is uncertain. ¹⁵² Moreover, the Committee’s professed firm stance on retrogressive measures has not really been put into practice, which further adds to the uncertainty surrounding the concept and its legal repercussions. The only statement made by the Committee on this matter in the context of austerity measures was rather weak as it did not declare such measures to be very difficult to justify. On the contrary, it was rather deferential toward states parties. ¹⁵³ Considering the openness of migration policy context to arguments such as the need to roll back the protection of the right to access housing for migrants and refugees to curb inward movements, ¹⁵⁴ if such practices occurred in the form of discrimination, there would have been an instability as to how to assess their justifiability were they to be characterised as retrogressive measures. However, clarifying that this is not retrogression but a violation of the right to non-discrimination precludes such instability, as it is then clear how such measures ought to be scrutinised, i.e., with a proportionality analysis which, depending on the context, the conduct, the type of discrimination at issue, the relevant grounds etc., would ideally be carried out via strict scrutiny. ¹⁵⁵

Conclusion

This analysis has shown two things. First, the Committee’s work so far on the right to access housing and non-discrimination is relatively underdeveloped and at times disorderly, particularly in conceptual terms, in considering and addressing the many ways in which migrants and refugees can be discriminated against in access to housing. However, second, there is a way out of this conceptual unclarity, through a methodical interpretation of the Covenant and a consistent and dynamic application of the important substantive standards that it provides. Why the tools available to that end are not or may not be put to work by the Committee is another question that more clearly relates to the institutional and structural constraints within which the Committee functions. That question is to be answered in another work. For now, at least, this paper makes the normative value and potential of the Covenant in this context clear.