Spheres of sanctuary: introduction to special issue

ABSTRACT

At a time when restrictive immigration policies are high on the political agenda of many states, resistance to such policies is emerging from below. One form of resistance is the provision of sanctuary by civil society and subnational administrations. Sanctuary can be understood as the creation of a safe space for immigrants where they are beyond reach for immigration law enforcement. Research on sanctuary has been proliferating in the last few years. Nevertheless, a common definition and analytical framework are still missing. This article aims to fill this gap. In it, we provide a definition of sanctuary which identifies four key features that all sanctuary initiatives have in common, and that distinguish them from other forms of social and political action. Further, we distinguish between three spheres in which sanctuary can be enacted – the territorial, social and discursive spheres – and develop a typology that identifies the spaces, modes of enactment, and types of actors associated with each sphere. We show how our typology can be used to address both empirical and normative questions that require comparing different types of sanctuary. Our aim is to provide new impulses to the growing research agenda on the contentious politics and practices of sanctuary.

KEYWORDS:

• Migration
  
• sovereignty
• territorial sanctuary
• social sanctuary
• discursive sanctuary

My refuge

My fortress

My hope

You who save me

From the grip of evil

You who protect me

From fear and danger

You who give me refuge

In Your holy house

Excerpt from the poem ‘My refuge‘ by Hayarpi Tamrazyan (2018)¹

1. Introduction

In October 2018, 21-year old Hayarpi Tamrazyan took refuge inside the Bethel Church, a small protestant chapel in the Hague, Netherlands. Together with her parents and two siblings, she lived inside the church for months in order to avoid imminent deportation. The family fled from Armenia, where the father had received death threats because of his political activism. The Tarmazyans had been living in the Netherlands for nine years as asylum seekers, and had applied for permission to stay on the basis of a rule called ‘child pardon’, but their asylum and child pardon applications were ultimately denied (Grüll 2019). Taking advantage of a Dutch law inherited from medieval times that bars police from entering houses of worship during religious services, the Bethel Church organised 24-hours non-stop services to protect the family (Pérez-Peña 2018). According to the media, 1000 pastors from around the Netherlands and from other European countries were mobilised to assist in holding uninterrupted services (Kingsley 2019). After 96 days, the continuous service ended because the country’s political parties agreed to a compromise that allowed for certain family asylum applications to be reassessed, permitting the Tamrazyan family and hundreds of others to stay in the country (Boffey 2019).

This case of church asylum is extraordinary, but it is not unique. Just as the Tamrazyan family in the Netherlands, there are hundreds of persons taking sanctuary in multiple churches in Germany, which are organised into a national movement called Asyl in der Kirche. According to the website of the movement, as of 02 November 2022 there were 508 persons taking sanctuary in a German church (see: www.kirchenasyl.de). This movement sees itself as part of a broader international coalition uniting churches in the United States, Canada and Europe, civil society organisations, and city administrations that all strive to provide ‘sanctuary’ to irregular migrants and rejected asylum seekers (Lippert and Rehaag 2012).

Sanctuary can be broadly understood as the creation of a safe space for immigrants where they are protected from immigration law enforcement. The immediate goal of providing sanctuary is to block or prevent deportation, detention, or physical harm to the immigrant. Differently from mere acts of humanitarianism, however, sanctuary practices and policies generally also pursue a larger political goal. They are at once an act of solidarity with immigrants and an act of protest against national immigration policies. The provision of sanctuary thus involves making political claims, engaging in collective action, and non-cooperation with law enforcement authorities. In some cases, it may also involve acts of civil disobedience. Understanding the precise nature and character of this form of contentious politics is what we set out to do in this article, as well as discussing some of the normative justifications for sanctuary from a political science perspective.

This special issue contributes in two ways to the literature on sanctuary and on migration studies more broadly. First, it makes an empirical contribution by analysing the phenomenon in Europe. For a long time, the academic literature on sanctuary focused almost exclusively on Canada and the United States (for an overview, see: O’Brien 2020). Recently, there has been a surge of academic interest in the topic of sanctuary also in Europe, mainly driven by the fact that several European cities declared themselves to be cities of sanctuary or refuge, and some have organised themselves into transnational networks of urban solidarity (Lacroix and Spencer 2022; Oomen 2019). A number of recent research projects address this new trend towards the rise in sanctuary cities and the spread of specific sanctuary policies in Europe (Kaufmann and Strebel 2021; Hermansson et al. 2022; Bazurli 2019; Timmerman et al. 2020). Nevertheless, publications bringing together different European case studies are still rare.

Second, and most importantly, the special issue makes an analytical contribution to the literature by conceptualising sanctuary as a political phenomenon that takes place in various (territorial, social and discursive) spheres, is enacted by multiple authors, has differential effects on migrants, and articulates diverse types of contestation. So far, most studies of sanctuary focus on only one type of sanctuary – sanctuary cities, church sanctuary, or, to a lesser extent, sanctuary networks enacted by civil society. While all authors acknowledge that sanctuary is a broader phenomenon, very few publications bring together and compare multiple manifestations of sanctuary. This special issue helps to fill this gap. Each of the empirical articles in this volume addresses a different sphere of sanctuary, and some contributions address the interaction between different spheres, as well as theorising the normative implications of this broader understanding of sanctuary.

This introductory article is structured as follows. In section 2 we discuss the key features that sanctuary initiatives have in common and that distinguish them from other forms of social and political action. Our definition is large enough to encompass a wide variety of manifestations of sanctuary practices at different times and in different places. Section 3 presents a typology that enables us to differentiate between various spheres in which sanctuary can be enacted. We build the typology around three categories: ‘space’ (where does the provision of sanctuary take place?); ‘enactment’ (how is the sanctuary enacted?) and ‘provider’ (who is providing the sanctuary?). The typology is meant as a tool for comparing different cases. It also helps to assess the potentials and limitations of different enactments of sanctuary with regards to their main aim of providing immigrants with a safe space where they are shielded from immigration law enforcement. Following that, in section 4, we outline different ways in which this typology can be made fruitful for a comparison between the spheres. Finally, in section 5, we conclude the introduction by providing an overview of the articles included in the special issue.

2. Multiple manifestations of sanctuary

The modern practice of granting ‘sanctuary’ to irregular migrants and rejected asylum seekers draws on an ancient religious and legal tradition whereby fugitives and criminals of all kinds could take refuge in a church or temple and, while on the premises of this sacred place, were shielded from punishment for their crimes. According to the legal historian Karl Shoemaker, sanctuary already existed as a tradition in ancient Egypt and Greece. It was a pagan and pre-Christian tradition that was taken up by Christian churches and eventually made its way into Roman law, being first codified in the 4th century A.D. (Shoemaker 2011, 10). Later, sanctuary became a part of cannon law, was repeated in a host of medieval royal legislative commands, and constituted a central feature of medieval life in Europe. From the sixteenth century onwards, however, sanctuary started to be associated with impunity, and was seen as detrimental to emerging notions of the rule of law due to its infringement upon the province of the sovereign. Thus, the famous Enlightenment thinker Cesare Beccaria writes in the eighteenth century that ‘within the borders of a country there should be no place independent of its laws’, and that ‘to multiply such places of asylum is to create so many small sovereignties  …  where laws have no say’ (quoted in Shoemaker 2011, 18). Church asylum was thus considered to seriously interfere with the administration of justice.

Since the 1980s, church sanctuary has re-emerged as a transnational movement in North America (United States and Canada) and Europe. In the United States (US), the ‘New Sanctuary Movement’ was formally launched in 2007 (Yukich 2012; Caminero-Santangelo 2012). The qualifier ‘new’ in the name makes reference to another sanctuary movement that unfolded in the 1980s as a response to civil wars raging in Central America (Perla and Coutin 2012; Cunningham 1995). The current movement has been growing and presently counts 40 regional sanctuary coalitions throughout the US (www.sanctuarynotdeportation.org). In Europe, the official launching event of the New Sanctuary Movement took place in Berlin in 2010 with an international conference, the adoption of a Charta, and the formal proclamation of a European Network of Sanctuary Churches (Just 2012). There are records of active sanctuary alliances in Germany, Belgium, Norway, Switzerland, France, the United Kingdom (UK), and the Netherlands (MacFarlane and Mitchell 2019). The movement is strongest in Germany where the European movement started and where it has acquired a semi-official status (Mourão Permoser 2022a, 2022b). In both Europe and the US church asylum has come under intensive criticism by authorities. Churches are accused of putting parochial morality above legality and misusing religious prerogatives to unduly interfere with state matters, essentially undermining thereby the liberal-secular political order.

Next to churches, the main locus of sanctuary are cities and other sub-national-level jurisdictions. Sanctuary cities, counties, provinces, regions or states adopt measures aimed at providing a safe environment for irregular immigrants to participate in society, make use of institutions, and receive services without fear of discrimination or deportation. The main instrument they use to achieve this goal are so-called ‘firewall policies’, that is, policies adopted by subnational administrations that instruct local public service providers not to cooperate with immigration authorities (Carens 2013, 130–147; Bauböck and Permoser 2023). According to a recent study, there are 400 subnational jurisdictions in the US that refuse to some extent cooperation with federal immigration authorities (Avila et al. 2018). In Europe, several cities have declared themselves to be cities of sanctuary or cities of refuge.² Examples of such cities include Barcelona, Mailand, Zurich and Sheffield, among others (Bauder 2017; Squire and Darling 2013; Bazurli and de Graauw 2023; Wyn Edwards and Wisthaler 2023; Humphris 2023; Lacroix 2023). Unlike in North America this is not always coupled with the adoption of concrete firewall policies. Firewall policies are justified in practical terms and ‘as a necessary moral and ethical response’ (Lasch et al. 2018, 1754) to inhumane immigration policies. Critics in turn reject these policies as illegal and unethical, arguing that they contradict federal legislation and unduly protect undeserving immigrants. Similar to churches, sanctuary cities that implement effective firewalls can be regarded as defying an aspect of national state sovereignty. Although sanctuary policies are themselves enacted by local governments, they also contest the national government’s monopoly in immigration law. As we discuss in section 3, even purely discursive sanctuary proclamations by subnational governments articulate such contestation although on their own they do little to effectively protect undocumented migrants. It must also be noted that the degree of constitutionalised autonomy of subnational governments is relevant for their capacity to go beyond discoursive sanctuary. This is easier in federal states or in countries with strong municipal autonomy, such as Switzerland.

More recently, the notion of sanctuary has also been applied to other sites where protection against violence, arrest and deportation is granted to unauthorised migrants and asylum seekers, regardless of whether these initiatives use the term sanctuary in their self-description or not. Mann and Mourão Permoser (2022), for example, apply the label of sanctuary to refer to NGO ships engaging in maritime rescue operations in the Mediterranean. These vessels rescue migrants and offer them a place of safety where they are temporarily shielded from border control agents. Mann and Mourão Permoser argue that

rescue ships function in a similar way as sanctuaries in that they constitute themselves as spaces where migration categories do not apply and where, for a limited period of time, every person is entitled to stay under a condition of temporary amnesty. (Mann and Mourão Permoser 2022, 18)

Also focusing on civil society rather than local governments, Cuison Villazor and Gulasekaram (2019) speak of the existence of ‘sanctuary networks’ composed of a multiplicity of actors acting together to protect asylum seekers and migrants. They mention sanctuary campuses and school districts, sanctuary workplaces, as well as rapid response networks created by citizens to warn immigrants of law enforcement raids. They use the term sanctuary broadly to refer to ‘a range of policies and programs adopted by public and private entities or organizations that decline or limit voluntary participation in federal immigration enforcement practices or seek to shield noncitizens from federal enforcement efforts’ (Cuison Villazor and Gulasekaram 2019, 1209). In this volume, Boudou (2023) applies the notion of sanctuary to citizens helping migrants cross the border in the Alpine region between France and Italy, and Elsrud, Lundberg, and Söderman (2023) use it to describe the actions of civil society networks providing hiding places and various other forms of assistance to migrants in Sweden.

Given this expansive use of the term, the question of definition has become central for academic analysis. Whereas the increased use of the term sanctuary could point to an organic growth and diversification of the sanctuary movement itself, it is not always clear whether this is indeed the case. Rather than being a result of actual trends in the real world that can be studied empirically, the increased use of the term sanctuary could also be an artefact generated by an academic literature that has started to apply the concept of sanctuary indiscriminately to an ever-wider range of phenomena.³ Such ‘conceptual overstreching’ bears the risk of diluting the analytical value of the term. On the other hand, a too narrow definition of sanctuary carries the risk of being excessively rigid and tied to past experiences, while being unable to include current and future variations of the same phenomenon. In order to avoid these double risks, our first aim in this introductory article will be to provide some conceptual clarification.

3. Towards a general definition

Despite their differences, all modern sanctuary initiatives are united by four key features. The first (I) is that they all serve to protect individuals whom political authorities try to seize for purposes of punishment or removal from the territory by blocking or temporarily suspending the effective enforcement of a political sovereign’s will or laws. Sanctuary movements aim therefore to bring about a partial or temporary suspension of political authority whose operations are inhibited ‘here’ (e.g. in a church or city) and ‘now’ (for the time being but not necessarily forever). For this reason, sanctuaries have been defined as ‘a spatial practice of setting a specific territory, location, or building outside the bounds of sovereign authority and legal redress, albeit in often temporary ways’ (Darling and Bauder 2019, 8; Bagelman 2016).

The spatial element of protection is a defining feature of sanctuary that is at the same time also a limitation. Sanctuary, in order to be effective, must be more than a mere act of humanitarianism or resistance by an individual or NGO. It must be associated with an actual physical space of safety for its beneficiaries, a space either de jure or de facto beyond the reach of law enforcement authorities, a place where one can take refuge and seek protection from state power. As we will discuss in section 3, this spatiality is even present where declaring sanctuary is primarily a symbolic gesture of claiming a right to offer protection in the absence of powers and resources to effectively provide it.

Because this physical space is a space of protection from state power, it is normally located beneath the level of the nation-state. Here, two objections should be addressed. First, one might argue that sanctuary movements and networks are not necessarily sub-national because they can also stretch across national borders. Indeed, advocacy for sanctuary or for specific policies and practices associated with sanctuary is also articulated in national and international arenas. However, the physical space where sanctuary is granted will generally not be that of the whole state territory, but will be found somewhere within or at the margins of a sovereign state’s territory.

Second, one might argue that, despite the fact that sanctuary is always articulated in opposition to nation-state policies (the sanctuary must be a sanctuary from something), it also depends on a certain degree of leniency and toleration by the state. Such toleration is often due to a lack of state capacities for detention and deportation but – in liberal democratic states – can also arise from legal and political obstacles, such as constitutional duties to respect the autonomy of churches or cities, or a fear of local mobilizations of civil society in support of immigrants.⁴ A tension between contestation and conformity is at the heart of any form of sanctuary, and also characterised its medieval or ancient forms. This tension is pointed out and analysed in all contributions in this special issue, with authors differing about the degree to which they see sanctuary as enacting a radical form of resistance or a relatively conformist political action.

With regards to the spacial dimension of sanctuary, it is not by chance that the most common locations for the enactment of sanctuary have historically been the church, the city and the border. These are places where nation-state sovereignty is limited or contested. The church has a second sovereign (God), the city is a co-sovereign, and the border is often either de jure or de facto a ‘no man’s land’. The paradigmatic space where a sovereign is absent are international waters, where international maritime law – and in particular the obligation to rescue that is addressed to individuals rather than to states – has developed precisely out of the inability of states to effectively exercise authority and control over the entire extension of the sea (Mann 2016). To a lesser extent, this applies also to territorial borderlands between two countries, particularly where these are difficult to control by virtue of geographical or climatic peculiarities, as in the mountains of the Alps or in the desert between Mexico and the United States.

Since they are defined (at least in part) in opposition to national-level policies and practices, sanctuaries cannot become permanent and all-encompassing national institutions. They are spaces of exception that are limited in space and often also in time. They are limited in space because they provide protection within a part of the state territory without, however, overturning the overall immigration regime. They are also often limited in time because the inherent instability of such arrangements conflicts with national laws in democratic states committed to the rule of law. Moreover, in states with effective means of enforcement, the emergence and persistence of sanctuaries will depend on a certain level of toleration by national state authorities, as explained above. For this reason, sanctuary arrangements are unlikely to be indefinite, although sanctuary provided by subnational governments acting within their constitutional powers have arguably a greater capacity to be sustained over time than sanctuary provided in churches and in other physical spaces managed by civil society organisations. From this discussion, we are now able to derive the second and third key features of contemporary sanctuaries: (II) sanctuaries are limited both spatially and temporally and (III) they are practiced at the sub-national level, as a form of resistance or opposition to national-level laws, but ultimately also depend upon the state’s toleration of the practice.

Related to this, the fourth (IV) key feature of modern sanctuaries relates to the relationship between humanitarianism and politics. Della Porta and Steinhilper (2021) argue that modern solidarity movements with migrants often blurr the boundaries between humanitarian action and contentious politics. This is also true of sanctuary. Every instance of granting sanctuary serves to help one person or a group of persons who are in need of protection. At the heart of sanctuary is therefore always a humanitarian impulse to provide emergency assistance. However, in contrast with their historical predecessors, modern sanctuary movements focused on migration are also political movements that protest against a migration and asylum regime that is perceived to be unjust, illegitimate or at least deficient. Across different types of sanctuary movements or arrangements, the mix between humanitarian and political motivations is likely to vary, and this will influence their forms of action. And in every instance of sanctuary there might be other sets of motivations at work as well, of course. Nevertheless, we consider it to be a characteristic feature of sanctuary that all will combine humanitarian and political motivations to some degree.

To sum up, modern sanctuary movements are characterised by four core features. Sanctuary practices create safe spaces for irregular migrants (I) by blocking or suspending state action against them, (II) are limited in space and time, (III) are enacted at the sub-national level in opposition to national policies, and (IV) are motivated by a combination of humanitarian and political reasons. These features provide us with operational criteria that can be useful in defining whether a particular practice can be analytically considered to be an instance of sanctuary or not. They help to answer the question ‘how would you recognize a sanctuary if you see one?’. This is useful in a context in which there is no fixed terminology emerging from the field, with activists in different countries using different terms to refer to essentially the same type of practice. Importantly, our criteria are based on a discussion of the means by which protection is achieved, the level of governance involved, and the motivations of the activists. They do not restrict the provision of sanctuary to any specific location or set of actors. This is important because the provision of sanctuary can take many froms and be enacted in many different places, from churches and cities to hospitals, university campuses or even boats. By not restricting sanctuary to any one location or set of actors, our definition is able to encompass all these different manifestations.

4. Three spheres of sanctuary

However, not all manifestations of sanctuary are the same. Which criteria can we use in order to distinguish various types of sanctuary? Do various types of sanctuary differ in their effectiveness and do they need distinct normative justifications? In order to be able to analyse the potentials and limitations of sanctuary and assess them normatively, we need to categorise them and develop a typology. We turn to this task now.

We propose a typology that identifies three analytically distinct spheres in which sanctuary can be enacted: the territorial, social and discursive spheres. Our wider conception of sanctuary allows for a more ecumenical approach with regard to the territorial scales and material scopes of sanctuary practices. While most analyses have focused on sanctuary cities, the papers included in our special issue demonstrate that sanctuary policies exist also at the level of autonomous provinces, such as for example in Wales (Wyn Edwards and Wisthaler 2023). The articles in this special issue also demonstrate that effective social practices can emerge in border regions that are not politically bounded territories, such as in the Alpine region between France and Italy (Boudou 2023), or in the form of loose networks created by a multiplicity of civil society activists, such as in Sweden (Elsrud, Lundberg, and Söderman 2023). With regard to material scope, sanctuary practices may range from isolated acts of resistance against deportation to institutional arrangements and local level laws.

In order to capture these distinctions, we developed our typology around the following questions:

• Space: where does the provision of sanctuary take place?

• Enactment: how is the sanctuary enacted?

• Provider: who is providing the sanctuary?

A key feature of our typology is the interaction between the categories we have called ‘enactment’ and ‘space’. These categories refer to the means by which sanctuary is enacted and the locus of enactment, respectively. As mentioned in the previous section, sanctuary is enacted when a specific territory or location is set up as lying outside the bounds of sovereign authority with regards to the enforcement of immigration controls, even if only temporarily. But how is this done: Through public policies, laws and regulations implemented by governmental authorities in a particular jurisdiction? Through social practices and intra-organizational codes of conduct implemented by civil society organisations and social movements in a particular physical space? Or through discursive practices and symbolic measures adopted by any of these actors in the public sphere of communication?⁵

When there is a general correlation between spaces and modes of enactment, we have a specific type of sanctuary whose characteristic features – in terms of space, enactment and provider – are determined by the spheres within which they are enacted. It should be noted, however, that the spheres are not exclusive and can overlap. In a sanctuary city such as San Francisco, for example, social sanctuary and discursive sanctuary can all be taking place at the same time if the city adopts comprehensive non-cooperation polices, churches harbour irregular immigrants, and both discursively claim to be sanctuaries (Mancina 2012). We will come back to this point below Table 1.

1. Territorial sanctuary is provided by governments of different sub-national territorial units. It is enacted through laws and public policies. Particularly important in this context are ‘firewall’ policies, that is, policies of non-cooperation with immigration law enforcement authorities. We will analyse the relationship between sanctuary and firewall policies in more detail in the epilogue to this special issue (Bauböck and Mourão Permoser 2023). Examples of territorial sanctuary can be found in both Europe and the United States, but one should be aware of the fact that not all subnational jurisdictions that declare themselves to be sanctuaries actually achieve the material effects that we would associate with territorial sanctuary. There is very significant variation of the level of protection granted to irregular migrants in different self-proclaimed sanctuary cities and regions, as the articles by Humphris (2023), Bazurli and de Graauw (2023) and Wyn Edwards and Wisthaler (2023) show. At the lower end of effectiveness, territorial sanctuary can be reduced to purely discursive sanctuary when a public declaration of being a sanctuary is not followed up with specific protection policies.

2. Social sanctuary refers to physical spaces within which immigration law enforcement is blocked. The concept of social sanctuary accounts for the ‘sublocal’ level of particular institutions (churches, hospitals, schools), ships and private homes where these do not merely offer services but serve also as shelters. Social sanctuary is provided by organisations, networks, activists in civil society and individuals who shelter irregular migrants. Social sanctuary is thus enacted through social practices. In the case of churches and more hierarchically organised civil society organisations, social practices can also be backed up by intra-organizational policies. For example, a hospital might enact an intra-organizational firewall policy that prohibits hospital personnel from asking for the legal status of patients and discriminating between them on this basis. In this case, that particular hospital would become a sanctuary in the sense of providing a safe space where irregular migrants can seek medical treatment free from fear of being deported. Although we do not include cases of church asylum or of rescue ships in this special issue, they would also fit into this category (Mourão Permoser 2022b, 2022a; Mann and Mourão Permoser 2022), as do homes where migrants and refugees are temporarily protected from immigration law enforcement (Elsrud, Lundberg, and Söderman 2023).

3. Discursive sanctuary refers to public communication spheres in which sanctuary is symbolically proclaimed for territorial or social spaces, by governments at different territorial levels, organisations, social movements and political activists. In our view, it is unlikely that discursive practices alone will be able to constitute a physical space or territory as offlimits for immigration law enforcement, thereby achieving the aim of temporary suspension of state authority that is the first defining feature of sanctuary. In order for a physical space or jurisdiction to be effectively considered a sanctuary more than declarations are needed. Nevertheless, it seems to us highly relevant that certain territorial jurisdictions or movements publicly proclaim sanctuary for irregular migrants and rejected asylum seekers. Doing so is a speech act that signals not only opposition to a government policy but also a commitment to counteract it by providing protection. Such a declaration can in and of itself affect the life-chances and perspectices of migrants in the territory. Consider the converse example. A city or organisation that implements policies aimed at shielding migrants from immigration law enforcement but publicly denies the intention of being a sanctuary. Also this is a performative speech act – one that is likely to have limiting consequences for the effective protection of migrants as it refrains from mobilising public support. We, therefore, suggest that discursive sanctuary should be included as a distinct phenomenon that can precede or accompany territorial and social sanctuary or remain limited to public communication acts.

Table 1. Three spheres of sanctuary.

Sphere	Space	Enactment	Provider
Territorial Sanctuary	Territorial jurisdiction	Public policies, laws and regulations	State actors: e.g. cities, counties, regional governments
Social Sanctuary	Physical space	Social practices, intra-organizational codes-of-conduct	Non-state actors: e.g. civil society organisations, churches, networks of activists, individuals
Discursive Sanctuary	Public communication	Discursive practices, symbolic measures	All: e.g. local or regional governments, organisations, social movements, activists

5. Using the typology

When using this typology to assess different empirical case studies, analysts should be aware of the fact that, in each particular case, different spheres of sanctuary may be either congruent or divergent. For example, when sanctuary cities publicly declare their territory as a sanctuary and at the same time adopt public policies that are meant to protect irregular migrants and ensure their non-deportability, sanctuary is being enacted in both the territorial and the discursive spheres. By contrast, sometimes the spheres can be disconnected. For example, social sanctuary may be provided clandestinely by hiding migrants without announcing this action to the wider public. Cities can also enact sanctuary only in the discursive sphere, but not the territorial one, as we discussed above.

One way to compare between the different spheres is to assess them in terms of their effectiveness. When applying this criterion, we can assume that sanctuary enacted in the territorial sphere by governments can be most effective in providing protection because of the instruments of administrative power and coercion that only government institutions can wield (in a well-functioning state). We expect sanctuary in the social sphere to be more precarious for the same reason. Its success depends on governments’ (in)effectiveness with regard to their law enforcement powers, their ignorance (in the case of clandestine sanctuary) or their toleration of such practices. Finally, we expect that sanctuary in the discursive sphere alone cannot provide effective protection.

Nevertheless, there may be particular reasons why a certain form of social sanctuary is highly effective, despite not being entrenched in legal rules and governmental policies. For example, in some countries mainstream churches are able to use their special status within the society to ensure that the government will tolerate their sheltering of irregular migrants. The socially and culturally entrenched expectation that law enforcement officers ought not to conduct arrests within church premises may be so strong that it trumps other legal considerations (Mourão Permoser 2022b). Social sanctuary practices may also be enabled by the judicial branch of the national government, as illustrated by the French Conseil d’État invoking the value of ‘fraternité’ when decriminalising solidarity actions of sheltering undocumented migrants crossing the Italian-French border (Boudou 2023).

Another way to compare between the different spheres is to consider whether they require distinct normative justifications. In the epilogue, we will return to this issue. We will argue that, because of their differences in terms of the level of contestation and the type of actors that are involved in providing the sanctuary, territorial and social sanctuary also differ in terms of which normative justifications are prima facie most plausible to legitimize sanctuary as a form of political action. In particular, we will argue that civil disobedience could be a plausible justification for social sanctuary, but not for territorial sanctuary, where the actors involved are governmental actors (Bauböck and Mourão Permoser 2023). In the latter sphere, sanctuary is likely to be justified by reference of territorial autonomy and the strengthening of a residence-based conception of local citizenship (Bauböck and Orgad 2020).

Finally, and related to that, the three spheres of sanctuary can also be compared with respect to their ability to create alternative (less statist) forms of membership. Sanctuary refers to conceptions of membership in two different ways. First, those who provide sanctuary are engaging in a contestation of laws that they deem unjust or morally deficient, thereby performing what they consider to be a civic duty. They are, therefore, ‘enacting’ citizenship (Isin and Nielsen 2008). Second, actors providing sanctuary seek to create the necessary conditions for irregularized immigrants to participate in society and access civil rights and social services without fear of persecution. Also those taking up sanctuary are sometimes engaged in a political struggle for migrants’ rights. In doing so, these actors question bounded conceptions of membership tied to national belonging or to the possession of a state-authorized residence permit. If they are effective in reaching their goal, sanctuary policies and practices have the potential to inaugurate a form of membership that is independent of legal status (Villazor 2010; Oomen 2019; de Graauw 2021). In that case, sanctuaries can be important sites for the development of ‘urban citizenship’ (Bauböck 2003), or for the empowerment of migrants to engage in ‘lived citizenship’ (Lister 2007; Kallio, Wood, and Häkli 2020).

6. The articles in this special issue

This special issue is composed of five empirical articles analysing specific cases of sanctuary in Europe and an epilogue. The empirical studies bring the conceptual framework outlined in this introductory article to life by investigating and explaining, for each specific case, how sanctuary emerged, to which extent it is able to create a space of protection, which means are used to enact the sanctuary, and who are the actors involved. In doing so, they give flesh to the elements that make up our typology, so that at the end a richly detailed picture of the spheres of sanctuary emerges – one that makes clear both the possibilities and the limits of each sphere, as well as revealing the ways in which they sometimes overlap. Also noteworthy is the fact that all the empirical contributions in this special issue broach the two normative debates outlined above, namely, the issue of whether sanctuary contributes to creating a more inclusive form of membership for immigrants, as well as the degree to which sanctuary is confrontational vis-a-vis national and international laws, and how the contestatory element can be normatively justified. In the epilogue, we will pick up on these issues and we will put sanctuary into perspective by comparing it to two alternative inclusive responses to the presence of unauthorised migrants in the territory: firewalls and regularizations.

The remainder of this introductory article briefly summarises each of the contributions in this special issue. We start with two contributions that focus social sanctuary in non-urban contexts. Boudou analyses a case of social sanctuary in the border region between Italy and France, while Elsrud, Lundberg and Söderman analyse transversal sanctuary networks that were initiated in rural areas of Sweden. We then move on to discussing cases of urban and regional sanctuary. The articles by Humpris and by Wyn Edwards and Wisthaler explore two cases in which sanctuary is enacted primarily in the discursive sphere within the same national context of the United Kingdom, namely in the city of Sheffield and in Wales. The fifth empirical article, by Bazurli and de Graauw, compares four sanctuary cities – San Francisco, Houston, Barcelona and Milan. They develop an analytical framework that explains why some of these cities have adopted substantive non-cooperation policies and constituted themselves into territorial sanctuaries, whereas others enact sanctuary only discursively.

Benjamin Boudou examines sanctuary spaces and practices in the Roya Valley, at the border between Italy and France, and more particularly the journey of its most well-known activist, the farmer Cédric Herrou. He shows that, although Herrou’s first actions could be best described as humanitarian and hospitable assistance towards migrants, they gradually evolved into more structured solidarity actions. These led to a decision by the Conseil d’État invoking the French constitutional value of fraternity that the legislator must better distinguish disinterested support from human smuggling motivated by profit. Boudou argues that the concepts of fraternity, hospitality and solidarity identify different repertoires of sanctuary that respectively fulfil functions of signalling, protecting and community-building. These different functions, he argues, when articulated together, constitute a social sanctuary or network of safety.

Torun Elsrud, Anna Lundberg and Emma Söderman analyse what they term ‘transversal sanctuary enactments’ in Swedish society, by which they mean bottom-up sanctuary alliances and networks engaged in resisting deportation of irregularized immigrants. Their study draws on empirical data collected in the transdisciplinary research initiative ‘Asylum Commission’ and is based on interviews with 90 people, including asylum seekers and people voluntarily or professionally involved in support work. One of the central themes analysed in this contribution is the issue of trust and its relationship to sanctuary. In contrast to Boudou, who highlights the ways in which social sanctuary can have a positive spill-over onto the institutional sphere, Elsrud, Lundberg and Söderman emphasise how social sanctuary arises out of, and sometimes also contributes to, a dramatic loss of trust in democratic institutions.

Rachel Humphris investigates the case of the UK city of Sheffield that proclaimed itself as a ‘City of Sanctuary’. Based on her long-term ethnographic fieldwork in Sheffield, Humphris argues that the self-affirmation to be a sanctuary city can be interpreted as a mobilising metaphor, which is given significance at particular moments and can be used either in a radical contentious or in a conformist way, to ameliorate frictions in urban governance. She traces how and why the ‘City of Sanctuary’ term emerged, who labelled the city in this way, for what purpose and with what effects. Humphris argues that in Sheffield sanctuary was enacted mainly in the discursive sphere and remained largely a conformist practice, whereby the local administration seeks to welcome migrants but resists pushes by civil society to engage in non-cooperation.

Catrin Wyn Edwards and Verena Wisthaler explore sanctuary policies and practices for refugees and asylum seekers in Wales. They argue that the Welsh government’s approach to sanctuary is richly symbolic and is used strategically to compensate for the lack of formal legislative competencies in the field of immigration and asylum, as well as to position the Welsh government in opposition to the UK government and its ‘hostile environment’ policy. In contrast with Humphris’ analysis of discursive sanctuary in Sheffield, Wyn Edwards and Wisthaler argue that, in Wales, discursive sanctuary has had tangible effects. They contend that discursive sanctuary has been beneficial both for migrants and for the affirmation of Welsh identity within a unitary state, as well as facilitating concerted action on the part of relevant governmental and non-governmental actors within the region.

Raffaele Bazurli and Els de Graauw present evidence from four American and European cities – San Francisco, Houston, Barcelona, and Milan – and develop a framework for explaining variation in the types of sanctuary policies. The article seeks to explain whether policies bring symbolic or substantive benefits to undocumented immigrants and whether they conform to, or confront, the goals of national immigration policies. In order to do so, Bazurli and de Graauw combine existing theoretical approaches to local immigration policymaking that emphasise the relevance of city-specific contextual factors and of the supra-local context in which different cities are embedded. They show how a combination of these factors have yielded substantive-confrontational sanctuary policies in San Francisco, symbolic-conformist policies in Houston, substantive-confrontational policies in Barcelona and symbolic-conformist policies in Milan.

Rainer Bauböck and Julia Mourão Permoser close the special issue with an epilogue. We discuss three inclusive responses to unauthorised immigration: sanctuaries, firewalls and regularizations. Although these responses are motivated by similar concerns, they differ with regard to their effects for migrants, the kind of actors that adopt them and their implicit normative justifications. The analysis connects the so far largely separate research topics of sanctuary, on the one hand, and firewalls and regularization, on the other hand. It aims to lay the grounds for deeper mutual engagement between social scientists, normative theorists and policy actors by conceptually comparing the characteristic features of the three responses to irregular migration and normative justifications that can be invoked in their defence.

Acknowledgement

Earlier versions of this paper were presented at the ECPR General Conference 2021 (online) and at the ECPR General Conference 2022 in Innsbruck. We would like to thank all participants in these events, all authors who have contributed to our Special Issue, as well as the anonymous reviewer who has commented on this piece for their cooperation and their helpful feedback. We would also like to thank Oleksandra Terentyeva for providing comments on an earlier version of this article. Finally, as editors, we would like to express our gratitude to all the scholars who have agreed to act as anonymous reviewers for all the other contributions in this Special Issue.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported in part by the FWF Austrian Science Fund [‘Migration as Morality Politics’, Grant V 743-G].

Notes

1 Hayarpi Tamrazyan’s poems are available here: https://gedichtenvanhayarpi.wordpress.com/. (Own translation)

2 The term sanctuary is more common in the United States, Canada and the UK. In Europe, we also encounter other labels such as ‘city of refuge’, ‘welcoming city’ or ‘solidarity city’ (Bauder and Gonzalez 2018).

3 Compare Brubaker’s critical discussion of the conceptual overstreching of “diaspora” (Brubaker 2005).

4 We thank a reviewer for raising this point.

5 The association with distinct types of public and private providers is similarly strong for both territorial and social sanctuary, as it is already built into the definition of these manifestations. Only discursive sanctuary can be enacted by providers of any kind.