1 Introduction

Cities have become key players in all manner of policy areas concerned with the mobility of humans, labor, and capital. They co-govern settlement and integration programs, help administer temporary foreign work regimes, bolster migrant civic engagement, and provide access to core social services such as health, education, transit, and housing. Many cities also contradict national immigration policies through sanctuary policies and other strategies of inclusion that help growing numbers of non-status or unauthorized migrants navigate exclusionary national and provincial/state laws. The sheer scale of local involvement in migration changes the way we understand cities, borders, citizenship, and constitutions.Footnote ¹ This is especially true of federal states, which promise local autonomy but which, in truth, have established records of delimiting municipal authority and redirecting centrifugal forces in the service of nation-building.Footnote ² Sanctuary policies and other aspects of local migration governance provide opportunities to reflect on the robustly democratic heritage of federalism, including its capacity for managing the tensions, contradictions, and occasional violence that erupts when a plurality of political communities occupy the same physical space.

But federalism comes in many forms, and the form that predominates in doctrine represents a different history – a different set of functions.Footnote ³ In the context of Canada and the United States, judicial interventions in disputes about jurisdiction are premised on the twin myths of dual sovereignty and political neutrality. The former describes sovereignty as a finite resource, exhaustively divided between federal and local scales with each level of government reigning supreme in its allotted sphere. The latter holds that the role of courts is to police this division of powers by enforcing the plain language of constitutional text. In this way, disputes about migration center around how to classify the “core” subject matter of a policy and then to determine which level of government is “naturally” authorized to govern it. Given that the act of defining is the act of deciding, and that the judges who ultimately decide these questions are appointed by federal governments, the myth of judicial impartiality was indispensable to the core function of federalism: preserving political stability. At root, the symbolic depoliticization of judicial interventions serves to avoid a reckoning with the political bases for choices about who has power over what issues.

Sanctuary city policies are not well served by this kind of federalism. They are not simply matters of migration or local administration but are also part of the broader history of city-building, urban resistance to racial and economic inequalities, and the habitual political disenfranchisement of municipalities- these are issues that require confronting the political choices underlining distributions of authority and that will not be resolved under a facade of a pristine constitutional equilibrium. But federalism doctrine is a hard habit to beat. It is everywhere in academic, policy, and public discourse about sanctuary cities, which reduce it to questions of whether cities can deliver on their promise of providing safe space within which federal authority has no sway. The very use of the term “sanctuary” draws from the same motifs as federalism, which recommends that “the best way to protect minorities is to give them an exit option.”Footnote ⁴

This chapter is concerned with the limitations of approaching sanctuary cities through the lens of federalism doctrine. One way of doing this would be to join with others in exploring how municipalities, local public institutions, and non-state actors have assumed jurisdiction over broad aspects of migration without challenging federal sovereignty over citizenship and borders. Rose Cuison Villazor and Pratheepan Gulasekaram have recently done this through a careful study of how trans-local sanctuary “networks” composed of churches, educational institutions, unions, and other institutions scale locally generated resistance up to the federal scale, invoking federal statutory and administrative law to unravel immigration enforcement from within.Footnote ⁵ I will approach this issue from the opposite angle, which is how federal immigration authorities scale down to the local level, indirectly using local laws and powers to amplify their own jurisdiction within and through the city without directly challenging local sovereignty. Somewhat like sanctuary networks, the result is the movement of locally generated data identifying non-status migrants to the federal scale and, through the subsequent management of migrant populations, the conscription of local authority in the service of immigration enforcement. Spatially mobile border regimes cross from national to local and back again in spaces of shared jurisdiction, without directly contesting the precept of dual sovereignty.

The chapter uses the example of urban securitization in Canadian sanctuary cities to explore how federal immigration authorities have extended their reach beyond their jurisdictional grasp by tapping into the wellspring of locally generated data on populations and individual persons. This occurs in many ways, but my focus will be on partnerships between federal immigration authorities and local police. On the one hand, Canadian immigration authorities lack the operational capacity to conduct robust inland enforcement or to independently acquire data on non-status migrants. On the other hand, local police have drawn from logics of risk management and predictive policing to expand their access to the personal information of migrants through arrests, detentions, streets checks, and their access to information “hubs” in such areas as health, education, and social work.Footnote ⁶ Through interviews with local police agencies in Ontario, I outline the rationale for sharing these data with immigration authorities and the ways they use jurisdiction to avoid democratic accountability. The resulting picture is one of a sanctuary city where the punitive logics of surveillance, control, exclusion, and banishment operate with the greatest intensity.Footnote ⁷ The realities of securitization establish quite clearly that federalism’s promise of exit from national sovereignty (or, more accurately, sovereign power) is not to be had – not for migrants or sanctuary cities.

The chapter is organized as follows. In Section 1, I survey the weaknesses of federalism as applied to sanctuary cities, using as examples two leading theoretical perspectives on sanctuary cities in federal states: urban political economy and urban citizenship. In Section 2, I examine other ways of thinking about jurisdiction, focusing on the case of urban securitization. In Sections 3–7, I use data sharing between local police and federal immigration authorities in Canada to examine how federalism both facilitates and obscures shared jurisdiction over the border. I conclude by reflecting on the implications this has for sanctuary cities.

2 Sanctuary Cities in Federal States

It would be useful to begin with a review of sanctuary city policies in the United States, which have generated the most concrete and detailed scholarly record. Although sanctuary practices and policies in this setting are clearly concerned with rights, scholars, policymakers, and jurists predominantly approach them by reference to immigration federalism.Footnote ⁸ On this basis, sanctuary cities sit within subnational sovereign spheres, migration sits within the federal sphere, and jurisdictional conflicts emerge only when one level of government trespasses onto the space of the other. It should be noted that disputes are almost always connected with the question of data and who “controls” it: can local governments constitutionally withhold locally generated information about immigration status from immigration authorities and, phrased from the other angle, can federal governments compel disclosure of this information? Other questions emerge, to be sure, but control over data and whether to classify it as “local” or “national” is the starting point of any effective sanctuary policy or, for that matter, any border enforcement regime. Without it, governments cannot effectively implement, evaluate, refine, or account for policy.

The law of federalism provides a deceptively simple approach to the exceptionally complex problems posed by data politics. Cities may pass sanctuary laws if those laws are “truly” local in character and are only incidentally concerned with immigration. In turn, federal governments may not compel cities to conduct or facilitate immigration enforcement. These questions are settled by reference to precedent established in relation to the Tenth Amendment anti-commandeering clause, which protects local governments from being conscripted into administering federal policies, and the doctrine of preemption, which allows the federal government to override local laws under a range of conditions, including if a local law is an “obstacle to the accomplishment and execution of the full purposes and objectives of Congress.”Footnote ⁹ Federal courts have consistently ruled that the federal government cannot compel state and local cooperation in immigration enforcement, upholding sanctuary city and state laws that preclude, among other things, the sharing of locally generated data with federal authorities, even if that data pertains to immigration status.Footnote ¹⁰ It should be noted, though, that anti-sanctuary laws have also passed constitutional muster. The authority of Charter cities to disobey California sanctuary laws was recently upheld, as was a Texan law forbidding cities from passing sanctuary policies.

Legal scholars have done an excellent job analyzing this jurisprudence,Footnote ¹¹ but my interest is in how the concepts and categories of federalism doctrine have found their way into social science analyses of sanctuary. One example is what I will loosely term “urban political economy.” On this view, the fiscal and political capacities of cities are the primary variables explaining the nature and efficacy of sanctuary and other local access policies.Footnote ¹² Els de Graauw’s rich empirical work documents an official consensus among municipal officials that sanctuary is at most concerned with providing precarious and non-status inhabitants of a city access to services and rights to which they are already entitled as a matter of local law. But the legal authority to provide access does not settle the question of whether municipalities will either actively remove barriers to access or defend their authority to withhold data from federal authorities. As De Grauuw notes, these decisions are made on the basis of the political and economic consequences of rebuffing federal authorities. While big cities such as Chicago, New York, and San Francisco successfully fought against the Trump administration’s attempt to withhold federal funds from sanctuary cities, this was an exceptional situation. In the long run, cities need to have workable relationships with federal governments in order to manage global policy issues, of which migration is one. As democratic institutions, municipalities must also be attentive to the will of the electorate, which can be decisive in how ambitious sanctuary policies will be.

Urban political economy has the merit of describing at least official consensus that cities are bound by federalism doctrine. Few municipalities see sanctuary as a step toward rescaling authority over citizenship from the national to the local sphere. But bundled within these descriptive claims are a series of political claims that have to be placed into the much larger history of the relationship between federalism and nation-building. It is true that cities lack the constitutional capacity to assume the degree of political and fiscal independence that would enable them to adopt more ambitious policies and practices, but it is also true that they have done precisely that in other areas, including health, the environment, and economic development.Footnote ¹³ And equally important from this point of view is that federal governments facilitate and tolerate expansions of municipal jurisdiction in these fields, sharing jurisdiction without concentrating on the final question of who is ultimately sovereign in these spheres. While migration is among the most sacred of subject-matters from the standpoint of nation-building, even here, federal governments recognize and encourage expansive municipal roles in migration policy, including settlement and integration, which are understood to be inseparable from health, education, labor, and so on.Footnote ¹⁴ We can recognize the reality of official consensus over the federal government’s claims to a monopoly over migration while also recognizing that the actual governance of migration is far more fluid and complex than this.

Urban citizenship theory offers a second account of sanctuary that picks up on this point. Best represented by critical geographer Harald Bauder, this perspective constructs sanctuary cities as a “scale of formal belonging,” which can “supersede regional and national scales.” They do so through a mix of social, political, and legal factors. The social and political factors include the bare fact of inhabitance and repeated social interactions organized around discourses of inclusion, which can produce postnational sociopolitical identities and alter the form and organization of political communities.Footnote ¹⁵ The legal aspect is trickier. Bauder argues that local jurisdiction over migration and citizenship can be grounded in lex domicilii, an ancient body of (private) international law that grounds jurisdiction (including rights) in the physical location of a party to a dispute. This domicile principle can be contrasted with lex patriae, which in the Westphalian system establishes jurisdiction by virtue of national citizenship. Other principles may also be deployed, including that of territoriality, where the location of the subject matter of a dispute determines which political community possesses jurisdiction over a dispute.

Traditionally, it is courts that would use the foregoing principles to determine which laws apply when a typically private law issue contains a “foreign” element. While frequently used to coordinate international struggles over the regulation of transnational disputes (i.e., “private international law”), these jurisdictional devices have played a prominent role in the stabilization of domestic cultural and territorial conflicts in federal states. In the early years of confederation in Switzerland, for example, courts wavered between lex domicilii and lex patriae in identifying which canton had jurisdiction over a private (though not public) law dispute.Footnote ¹⁶ Conflict of laws here played the role of softening the coerced inclusion of seven Catholic cantons into the federation in 1847, following their secession. Under this arrangement, subnational governments could regulate a range of subject matters (wills and estates, family law, torts, contracts, etc.) in the context of the internal movement of Swiss citizens across historically sovereign territories. Similarly, in the United States, courts in one state can sometimes apply their own laws to disputes that touch upon the jurisdiction of another state, but they can also apply the law of another state in its own courts based on the domicile principle. To make matters more complicated, states can also vary the scope or applicability of federal public law within their territory, including constitutional rights.

Lex domicilii has therefore been an essential ingredient to maintaining federal systems, precisely by conditioning the possibility of shared rather than exclusive jurisdiction. For this reason, it actually contradicts the premise of dual sovereignty. It is interesting that this core myth of federalism doctrine would feature in critical conceptions of sanctuary. Connecting law to postnational citizenship theory, Bauder argues that there is a distinctive “legal strength to implement policies” and “domicile rules of belonging” in both American and Canadian cities by virtue of their autonomy from federal governmentsFootnote ¹⁷ because the city is “a territorial legal entity at a different scale at which sovereignty is articulated.”Footnote ¹⁸ But when applied within federal states, the domicile principle of belonging has never been about upholding dual sovereignty. To the contrary, it has always been used along with other conflict of laws principles to engender stability through the undulation of sovereignty, and the means for this have been layerings of shared jurisdiction that perforate the hard lines of sovereign enclaves. But to be clear, this process has also been central to the nation-building enterprise; through compromise, the national political community is stabilized and fortified. This being so, it isn’t clear how the domicile principle can lay the basis for postnational political communities within or without a federal state.

I could say more about this, but I will summarize this section by saying that the legal strength of cities is reduced to dual sovereignty in each of the accounts canvassed here. Urban political economy sees federalism as a source of legal weakness, where sanctuary cities lack the authority to govern the core moral subject matter at issue: migrant rights qua migrant rights. But this obscures the realities of shared jurisdiction that flow from city-building in a broad range of policy domains that can be brought to bear in sanctuary cities. Federalism doctrine also limits our imagination in urban citizenship theory, which starts well enough by rejecting the nation-building premise that citizenship and migration are inherently federal in character. But if not for the cover provided by sovereign subnational governments, cities in federal states would not differ from those in unitary states in terms of their “legal strength” – and this cover is occasional and always conditional. Worse, the legal materials invoked to support the transfer of jurisdiction from federal to local scales have actually been instrumental in fortifying national political communities.

3 Urban Securitization in Federal States

One way of responding to these problems would be to abandon federalism altogether. Rosa Cuison Villazor and Pratheepan Gulasekaram take this route, arguing that “discussion of the term ‘sanctuary’ remains obsessed with state and local rights,” reducing it to a “federalism contest” that pits federal jurisdiction over migration with “the right of states to control their own affairs as independent, constitutional actors.”Footnote ¹⁹ This is a strong point: denied constitutional autonomy, municipalities and other local public institutions are so often seen merely as “creatures” of states or provinces. But Villazor and Gulasekaram wisely distinguish jurisdiction from sovereignty, noting that municipalities and other local public institutions wield considerable authority within key policy fields and are adept at protecting this authority against federal incursions. Institutions of note include universities, hospitals, schools, business organizations, religious organizations, and digital sanctuary networks. Like municipalities, these institutions draw on jurisdictional devices other than federalism, including constitutional rights, common law, administrative and regulatory law, and statutory regimes. Universities, for example, have common law rights to prevent access to campuses and are actually obligated to maintain the privacy of student information under the Federal Education Rights and Privacy Act.Footnote ²⁰ This example is especially important because the local authority in question is sourced in federal law, drawing a direct line between local public institutions and federal legislative bodies. This reminds us of the importance of the separation of powers, whereby federal executive actions, such as border enforcement, can be contained by invoking the limits built into federal statutes.

What emerges is a picture of jurisdiction that associates municipalities with local (nonstate) institutions as much as or more than higher levels of government, and which eschews the precepts of dual sovereignty inherent to federalism doctrine. Two matters of interest bear noting. First, local authority can be defended by reference to separation of powers within the federal scale, rather than simply the spaces the division of powers affords within the local scale. The sites of contestation vary, but resistance to inland border enforcement occurs in local and subnational institutions (division of powers) as well as within federal courts, legislatures, and administrative agencies, all of which impose checks and balances on pure executive power (separation of powers) at the federal scale. Second, we are reminded that nonstate actors possess legal authority and identities that shape the production, interpretation, and application of state law, with interesting implications for political conceptions of sovereignty and how we define “the city.” Local nonstate actors and nonmunicipal, local public institutions share in the production and interpretation of federal immigration law across multiple scales, often in tandem with municipal governments with whom they co-govern key policy fields.

But the legal pluralism we see in local spaces is content-neutral.Footnote ²¹ The legal modalities of sanctuary networks are not unique, as their antagonists – security professionals – are equally nimble, if not more so. The securitization of migration in and through local laws is by now well advanced, producing new practices, agents, and spaces of border control.Footnote ²² While on the surface concerned with the maintenance of borders, urban securitization actually represents the collapse of territorial divisions between “internal” and “external,” blurs jurisdictional lines, and betrays anxieties about the nature and future of the nation-state. It is worth lingering on this point before relating it back to sanctuary cities.

The so-called externalization of borders is a well-known process involving the “territorial and administrative expansion of a given state’s migration and border policy” to foreign states and jurisdictions.Footnote ²³ Central to this process is a set of intentions or habits through which state power is fortified by means of the dispersal, pooling, or integration of sovereignty in the international field. Interdiction, digitization, the collection and sharing of information, and other global preventive and deterrent measures are part of the process, as are regional mechanisms of “opening” borders for desirable migrants and “closing” them for the undesirables. While this process presents opportunities to harden borders, it also threatens them and reveals the incapacity of nation-states to manage their borders alone. The rise of populism and the breaking apart of the United Kingdom from the EU reflect well how regional and international integration can be perceived as a loss of sovereignty.Footnote ²⁴

Varying conceptions of bordering practices shed light on a similar process of border “internalization.” Critical border studies describe national borders, not as fixed territorial or juridical lines, but instead as a set of spatially mobile performances, practices and technologies of exclusion and inclusion, often operating in connection with racial, colonial, and economic hierarchies.Footnote ²⁵ They too involve forms of interdiction, data-production and sharing, and the integration of functionally disparate agencies in the management of human mobility. Tellingly, state officials and xenophobic populists worry about the loss of sovereignty when management of borders is shared with external sovereigns, but when borders shift internally, the illusion of dual sovereignty remains: Local power to enforce the border is “delegated,” while legal control over citizenship and migration remains in the hands of the federal government alone.

Federalism doctrine plays the role of maintaining this illusion, obscuring how sovereign practices, understandings, relations, and institutions (what we might call assemblages) cut across territorial and jurisdictional divides.Footnote ²⁶ It does so in part by drawing artificial distinctions between the fields of migration and security, when the two are coeval political constructs. In constitutional terms, the former is reserved for federal governments, while the latter is of concern to all governments and is not subject to the precepts of dual sovereignty. Shared jurisdiction over matters of crime and security is the key to maintaining the illusion. Associations between irregular migration (a federal matter) and criminality (a shared federal/local matter) create space for local police and federal immigration/border authorities to pool operations, funding, and jurisdiction over such matters as human trafficking and people smuggling, drug and arms trafficking, terrorism, and transnational organized crime. In this way, federal immigration authorities receive data collected by local police and security agents but, by virtue of the myths of federalism, they can insist on exclusive control over borders. In other words, the policing of irregular migrants is primarily about criminal law enforcement and has only incidental effects on federal immigration law and policy. As I will proceed to show, the Canadian context reveals the dynamics of shared jurisdiction over crime/security and shows that, far more than federalism doctrine, it is urban securitization that determines the legal strengths and weaknesses of the sanctuary city.

4 High in Demand and Short in Supply: Data Collection at the Federal Scale

It is best to start at the federal scale, before drilling down to the local scale. As with many other nations, the securitization of migration in Canada was deeply affected by a suite of legislative, policy, and operational changes made post-9/11.Footnote ²⁷ Just prior to 9/11, parliament amended the Immigration and Refugee Protection Act, introducing a range of preventive and deterrent measures and the partial dismantling of the refugee status determination system.Footnote ²⁸ But the events of 9/11 permitted even greater operational changes, the most significant of which was the creation of the Canada Border Services Agency (CBSA) in 2003. The CBSA was an amalgamation of disparate customs and border authorities that were, until this time, strewn across several ministries, including what was then termed Citizenship and Immigration (now called Immigration, Refugees and Citizenship), Customs and Revenue, and the Canada Food Inspection Agency. Parliament enacted the Canada Border Services Agency Act in 2005, which rendered the CBSA fully operational.

The CBSA is housed within Public Safety Canada (PSC). This is the core security and criminal justice ministry in the country, which contains Canada’s federal police agency, the Royal Canadian Mounted Police (RCMP), Corrections Canada, and Canada’s primary security intelligence agency, the Canadian Security Intelligence Service (CSIS).Footnote ²⁹ The CBSA is vested with broad authority to enforce both the Immigration and Refugee Protection Act and a wealth of criminal law statutes, including the Criminal Code. As a law-enforcement agency, the CBSA employs technologies and practices similar to those used by police services, through its Criminal Investigations Division. It also possesses limited security intelligence powers. The CBSA has authority to partner with international agencies, which it uses to gather information, facilitate deportations, and physically obstruct access to Canadian territory.Footnote ³⁰ It works regularly with the United States in this respect.Footnote ³¹

The CBSA is legally and functionally unique, being the only entity within PSC that is not subject to independent oversight and review. Unfortunately, it is the only federal agency empowered to conduct both policing and security intelligence operations. The one saving grace is that the CBSA does not have particularly strong security intelligence powers nor is it well positioned to conduct policing operations or, for that matter, border enforcement between ports of entry. This is evident in the pivotal role played by the RCMP in policing the Canada–US border between ports of entry to stem the inflow of asylum seekers between 2017 and 2020.Footnote ³² Meanwhile, CSIS remains the premier security intelligence service in the country, handling the most serious security files in the immigration context. The CBSA’s role is principally geared to staffing ports of entry and overseas liaison work, with only 6,500 uniformed officers; this leaves very little for inland enforcement.Footnote ³³

Unsurprisingly, information is one of the pillars of Canadian security. When it passed its first-ever national security policy in 2004, the federal government stated that the “key to providing greater security for Canadians and to getting the most out of our security expenditures is to co-ordinate and better integrate our efforts.”Footnote ³⁴ Ever since, it has tried to smooth the flow of information, both domestically and internationally. One especially important year was 2015, when parliament passed the Security of Canada Information Sharing Act. This law mandates the sharing of security-based information among at least seventeen federal institutions, with special focus on those operating out of PSC.Footnote ³⁵

But as net importer of intelligence, Canadian policing, border, and intelligence agencies have a long history of competing with each other and jealously guarding data and sources.Footnote ³⁶ This is evident in the absence of an official interoperable security database accessible by all federal agencies with security mandates; data are contained in institutional silos and shared only on request and following high-level authorization. There are a handful of field-specific interoperable databases, including some linking border control and policing, but very little is known about them. The largest one is the Global Case Management System, which is a database shared by the CBSA and Immigration, Refugees and Citizenship Canada. It contains personal information related to citizenship and immigration applications, including name, date of birth, country of birth, address, medical details, education, and criminal history. Another is the Canadian Police Information Centre (CPIC), which is a central database that contains information “about crimes and criminals.”Footnote ³⁷ The CBSA has access to this database. Managed by the RCMP, it is “the only national information-sharing system that links criminal justice and law enforcement partners across Canada and internationally.”Footnote ³⁸ The CPIC is interfaced with the US’s National Crime Information Centre, so that the American authorities have access to the CPIC (but not information regarding young offenders) and the CBSA (and Canadian police) have access to American criminal databases.

It is within this context one must approach both inland border enforcement and urban securitization. On the one hand, the CBSA is legally vested with immigration, criminal, and security intelligence powers and is located in the heart of Canadian security and criminal justice governance. On the other hand, it is set adrift in a sea of informational scarcity alongside much larger, more mature, better resourced, and politically adept security and policing agencies. As a result, it focuses its efforts on the physical border, various interdiction and externalization strategies, and smoothing deportation and inadmissibility processes. Full-scale inland border enforcement is limited by these incapacities, although it has found workarounds in the form of partnerships with local public institutions with access to data.

5 Data Sharing between Local and Federal Institutions

It should be noted at the outset that there is no uniform body of law governing the sharing of data between federal border agencies and local partners in Canada. What we find is a patchwork of provincial privacy legislation, common law, and the occasional Memorandum of Understanding (MoU) between the CBSA and provincial or municipal bodies.Footnote ³⁹ Provincial privacy legislation is the most comprehensive of these, providing discretionary power to local public institutions to disclose for the purposes of aiding investigations by “law enforcement” agencies. A typical wording is found in s. 42(1)(g) of Ontario’s Freedom of Information and Protection of Privacy Act, which states disclosure may be made:

The language is generally the same in other provinces, as is the express requirement or permission to comply with search warrants and other court orders.

Early efforts to secure data were rather brazen, producing powerful public backlash. In 2013, for example, the Vancouver Transit Police (VTP) detained Lucia Vega Jimenez for fare evasion, reporting her to the CBSA. Ms. Jimenez committed suicide in a CBSA holding cell in the Vancouver international airport. In 2015, the VTP terminated its MoU with the CBSA in response to public outrage and recognition that the VTP lacked a mandate to engage in immigration enforcement. Schools are another good example. In 2006, the Toronto Catholic District School Board allowed CBSA officers to enter school property to enforce immigration law. In one instance, they threatened to arrest two students unless their mother turned herself in, only to take all three into custody. In another instance, the CBSA apprehended two children and took them to a van carrying their mother and grandparents.Footnote ⁴⁰ Public outrage led to the passage of a “Don’t Ask Don’t Tell” policy in the Toronto District School Board.

Transportation authorities are a third example. In 2014, the Ministry of Transportation Ontario used powers of vehicle safety audits to stop trucks so that the CBSA could search them. One of the trucks carried twenty-one non-status passengers. The public reaction was swift and powerful, leading the ministry to terminate its partnership with the CBSA.Footnote ⁴¹ The Ministry of Transportation Ontario cited as one reason the fact that it lacked the statutory mandate to participate in border enforcement. In none of these examples did the CBSA have a warrant, and in most, they did not know the identity or location of the migrant until local authorities informed them.

It is worth pausing to consider the place of jurisdiction in these examples. Animated by migrant rights, resistance to disclosure centered around the rule of law, inflected by federalism. Each public institution noted earlier terminated its MoU with the CBSA because it recognized that open-ended sharing of data was inconsistent with the principles and purposes of its enabling legislation; privacy legislation and rights were a secondary consideration at best. The rule of law argument is underlined by dual sovereignty and the lack of any constitutional, much legislative, link between border enforcement and schools, local transport, and municipal transit. More precisely, the lack of shared jurisdiction among these institutions and the CBSA undermined the legal and political justification for systematic cooperation.

But this is not so with police, where there is shared jurisdiction both vertically with the CBSA and, increasingly, horizontally with schools, shelters, NGOs, and other local institutions. One prominent example is the criminalization of human trafficking and the regulation of sex work. Following a series of constitutional challenges, aspects of sex work have been decriminalized, with provinces and localities filling the jurisdictional space with a bevy of regulations, including municipal zoning by-laws and labor laws.Footnote ⁴² But the most relevant are sweeping antihuman trafficking laws that associate sex work with international and transnational criminality. Multimillion dollar provincial and federal antitrafficking strategies establish concrete partnerships between local police and the CBSA, which rest on the belief that human trafficking is predominantly international and associated with irregular migration. The federal government launched a five-year antitrafficking strategy in 2019, spending $75 million over six years, principally in the shared area of criminal and immigration law.Footnote ⁴³ Many provincial governments provide similar levels of funding to police as well as to NGOs, which help collect and share data for the purposes of aiding law enforcement investigations.Footnote ⁴⁴ Migrant rights organizations note that these powers are “anti-sex work, anti-migrant, and racist,” co-opt grassroots organizations through fiscal incentives and language of care, and roll out “heightened surveillance capabilities” directed at racialized migrant sex workers.Footnote ⁴⁵

Shared jurisdiction is key, cementing horizontal information-sharing partnerships between local NGOs, public institutions, and police, and vertical data sharing with the CBSA through linkages between crime, security, and the border. But it is the tip of the iceberg. According to an Access to Information request filed by No One is Illegal, the police in the Greater Toronto Area made 4,392 out of 10,700 calls to the CBSA’s “Warrant Response Centre” between November 4, 2014, and June 28, 2015, which local police officers (or any other law-enforcement officer) can use to provide the CBSA with information about a person who the officer believes or merely suspects lacks status.Footnote ⁴⁶ The Toronto Police Service (TPS) made 75% (3,278) of those calls, which is more than the RCMP (1,197), and greater “than the police services of Montreal, Quebec City, Ottawa, Calgary, Edmonton, and Vancouver combined (2,729).”Footnote ⁴⁷ What is more, “status checks” were the most common reasons for calls – 83.35% in the case of the TPS as against a national average of 72%.Footnote ⁴⁸ The Service de Police de la Ville de Montréal has also been an active collaborator, making 2,632 in 2015, 2,872 in 2016, and 3,608 in 2017.Footnote ⁴⁹ It should be underlined that more than 83% of these calls were status checks and not responsive to a federal arrest warrant.

The policing of non-status migrants has escalated while most formal partnerships between local and federal institutions have been dismantled. Indeed, police involvement in border enforcement is the most rigorous in the two most high-profile and reputedly most robust sanctuary cities in the country: Toronto and Montréal. While the rule of law and federalism have had an impact in many areas, they have been completely ineffectual in the context of policing, based precisely on the appearance of shared jurisdiction over irregular migration, when seen as a matter of crime and security. In Section 6, I will examine how local police understand the basis and scope of this jurisdiction.

6 Shared Jurisdiction over Crime, Security, and Migration

In 2017–2018, I participated in an empirical study of the local policing of non-status migrants in Ontario, led by Mia Hershkowitz, and joined by Harald Bauder.Footnote ⁵⁰ We interviewed eleven high-ranking officers (chiefs, super-intendants, etc.) in eight municipalities in Ontario (including Toronto) and asked officers to speak to their role in sharing information with the CBSA, how they perceived their role in border control, their perceptions of sanctuary policies, and their interpretations of Ontario’s Police Services Act (PSA) in these areas. While I cannot go into the details of study here, it is worth noting all officers we spoke with admitted to the routine sharing of information in the absence of a federal arrest warrant, and thought that this sharing was required by the PSA. For background, s. 5(1) of Ontario Regulation 265/98 states:

Case law is clear that disclosure in the absence of a court order is discretionary and not mandatory, and the CBSA must request specific information in the context of a specific investigation; the law does not allow for discretionary disclosure in the context of “[m]ere suspicion, conjecture, hypothesis or ‘fishing expeditions.’”Footnote ⁵¹

The first theme that emerged from the interviews was that police do not consider disclosure a form of border enforcement.Footnote ⁵² Officers insisted involvement is merely bureaucratic – a routine pushing of a file to the requisite agency. One officer stated that when non-status migrants:

This attitude aligns with associations between irregular migration and criminality, where even victims are referred to the CBSA, especially if an officer thinks the underlying crime is one of a national or international character, such as human trafficking or people smuggling.

The local boundedness of the policing of migrants is also established by reference to provincial statutes. Another officer stated:

Another officer stated:

It bears repeating that the statistics of information sharing noted earlier indicate that police are almost never executing immigration warrants but instead are proactively calling the CBSA to conduct “status checks.” If a warrant is issued, it’s only because of disclosure, not the other way around. Clearly, a warrant cannot retroactively justify the very disclosure the warrant is predicated upon.

The Toronto Police Service has provided a number of public statements that shed more light on the legal rationale for cooperation. In a 2017 hearing before the TPS civilian review body, the Toronto Police Services Board (TPSB), then Chief Mark Saunders stated the PSA and privacy legislation:

This is a strained interpretation of the PSA, which nowhere states that disclosure of identifying information to the CBSA or any federal law enforcement agency is required. It is understood, of course, that police must comply with a court order. But like provincial privacy legislation, the PSA provides discretion to disclose in any other context. Legislation specifically outlines the interests in nondisclosure that police must weigh before deciding whether or not to disclose. Section 6 of the PSA states that the decision to disclose must be based on considerations of “what is consistent with the law and the public interest.” The “law and the public interest” is defined in part through s. 4(2) of the PSA, which defines “adequate and effective police services” as including: crime prevention and assisting victims and witnesses. The principles and purposes policing itemized under s. 1 of the PSA include the following:

1. 1. The need to ensure the safety and security of all persons and property in Ontario.

2. 2. The importance of safeguarding the fundamental rights guaranteed by the Canadian Charter of Rights and Freedoms and the Human Rights Code.

3. 3. The need for cooperation between the providers of police services and the communities they serve.

4. 4. The importance of respect for victims of crime and understanding of their needs.

5. 5. The need for sensitivity to the pluralistic, multiracial, and multicultural character of Ontario society.

Disclosures to the CBSA contravenes many of these principles and, to this extent, violates the rule of law. Principles 3 and 5 are undermined by the documented fact that police cooperation with immigration authorities makes it far less likely migrant communities will report crime.Footnote ⁵⁷ Principles 2–5 are undermined by the fact that racial profiling is often a feature of the policing of non-status migrants in Canada.Footnote ⁵⁸ In fact, the analogous practice of “carding” (i.e., random street checks of mostly racialized minorities) was recently prohibited through provincial changes to the PSA regulations in 2017. A 2018 review of carding legislation led by Ontario Court of Appeal Justice Michael Tulloch recommended even clearer restrictions on the powers of police to collect personal information, including the outright prohibition of asking for information for arbitrary reasons.Footnote ⁵⁹ The Supreme Court of Canada recently referenced this report, among others, when it recognized systemic racism in policing.Footnote ⁶⁰

What we are left with is the fact that the scale of disclosure cannot be justified by reference to law or, more precisely, shared jurisdiction over criminal law. Most of the time, disclosure is made in the absence of a legal obligation and, worse, contrary to the principles and purposes of enabling legislation – just as it is when schools, transportation ministries, and municipal transit corporations share data. How is it that police can get away with flagrant contraventions of the rule of law?

There are two answers relevant to this chapter. The first is that security trumps federalism in the field of jurisdiction. Defending cooperation with the CBSA, Chief Saunders claimed that the TPS, “as a member of the law enforcement and public security community, respects and supports the mandate of other law enforcement agencies, like the C.B.S.A.”Footnote ⁶¹ We can glean much from this statement, when we disaggregate the terms “law enforcement” and “public security.” After all, the TPS does not partner with park wardens, by-law officers, or Canada Revenue Agency officers with the same enthusiasm as it partners with the CBSA, yet all are law enforcement agencies. The real community is not law-enforcement officers but, as Chief Saunders states, the “public security community.” This community is defined by a shared role in the management of threats to state and citizen, where legal distinctions between criminal law/immigration law and criminal/migrant break down and jurisdiction can be extended beyond what is permitted within the four corners of provincial legislation.

This has distinctive importance in federal states because local police can draw authority from federal law and, more to the point, they can use federal authority to transgress the provincial laws that confer them with most of their power. The federal law they draw from includes substantive criminal law, which now includes entire sections of the Immigration and Refugee Protection Act as well as the range of regulations designed to tamp down on aspects of irregular migration that are defined as international and transnational crimes. Through these laws, local police partner with federal agencies in the investigation of human trafficking, people smuggling, terrorism, drug trafficking, cybercrime, and so on. They are also part of broader circuits of security, defined by the insatiable hunger for information and, in the past several decades, a risk-based obsession with “pre-crime.”Footnote ⁶² Proximity to data gives them purpose and power, an opportunity to aggrandize themselves – to secure more funding, influence, and prestige.

But in democracies, police are compelled to use legal arguments to justify themselves. Long on discretion and short on accountability, police easily dissemble their role in the extralegal facets of the carceral state, where inside/outside and local/national are blurred every bit as much, and for much the same reasons, as migrant/criminal. This raises the second point: We see the double sense in which sanctuary policies are “provincial” – not just because cities are creatures of provinces but also because the dynamics of political struggle unfold outside of jurisdictional remit of discrete bodies or levels of government. The police and other security professionals are now serious political actors in their own right and know they have “the ability to act as more or less autonomous agents.”Footnote ⁶³ Police are not accountable to representative government, which can at best influence policing through civilian oversight and review commissions. In 2015, Toronto City Council tried to assert influence through an independent civilian oversight body, the TPSB, asking it to investigate and then report back on the following issues: (1) statistics on the number of non-status persons reported to the CBSA, (2) agreements that exists between the TPS and the CBSA, (3) practical implementation of sanctuary policies, and (4) and the possibility of amending the PSA to regulate that police officers only report immigration status to the CBSA when directed by the courts after a conviction has been registered.Footnote ⁶⁴ These demands have been ignored, with the TPS steadfastly refusing to change its practices and reminding the TPSB and the City of Toronto that they lack jurisdiction to direct operational changes.Footnote ⁶⁵ One participant in our study stated:

For policing the issue is that we are bound to respond to statute violations related to the criminal code, any other federal statute, along with any other statute at the provincial level, we don’t have the luxury of being able to turn to a municipality and say, “okay we are going to adopt your philosophies and your principles,” because our practices are not dictated by the municipality, it is exclusively the realm of the province. The province decides what we will and will not do. As a result, the province has decided that we will enforce federal and provincial statutes.Footnote ⁶⁶

But provincial law actually constrains the local policing of migrants. The irony is unmistakable: police violate provincial law but use its shadow as a bulwark against subsequent democratic accountability, using one form of jurisdiction to prevent an accounting of the absence of another.

7 Conclusion

Reckoning with urban securitization offers an important inroad into the nature and limits of sanctuary cities in federal states. As the urban assumes a decisively municipal character in sanctuary scholarship, we become preoccupied with a constitutional order historically geared toward nation-building and against city-building. Through federalism, the city appears as epiphenomenal, utterly dependent on national and subnational governments for political and legal authority. While this is partly true, sanctuary cities are not reducible to municipalities nor is municipal authority reducible to formal constitutional law or enabling statutes. Cities play greater roles in a range of policy fields that intersect with and include migration, through patterns of shared jurisdiction that reflect a transformation of the nation-state and the emergence of other global and trans-local political communities. As Saskia Sassen has shown, these parallel processes occur precisely through the concepts and institutions of the nation-state, so it is no surprise to see federal governments carrying this process forward, even if they don’t fully appreciate what they’re doing.Footnote ⁶⁷ However, attempts to theorize the sanctuary city in this (possibly) emergent context of urban citizenship draw heavily on the cumbersome and ultimately unproductive language of dual sovereignty. While no doubt a reality that must be contended with, federalism doctrine produced by courts is so domineering as to obscure what critical federalism theory can offer with respect to the democratic potential of local communities in this transformative setting.

Shifting away from sovereignty toward (shared) jurisdiction reveals that local institutions can and do wield more authority than at first meets the eye. This chapter has been less concerned with exploring the empirical bases of this point in terms of sanctuary policies and networks, as with how the legal strength of the sanctuary city is affected by the parallel legal modalities of urban securitization. The Canadian experience shows that local police have also acquired considerable authority over the governance of migration through the shared jurisdiction produced by the criminalization of migration. The key commodity is data. Far from being inert, data “is generative of new forms of power relations and politics at different and interconnected scales.”Footnote ⁶⁸ By virtue of their access to data and the logics of risk management and predictive policing, local police are now key players in border enforcement. The spatial mobility of the border goes hand in hand with the legal mobility of police, who seem able to feely cross the boundaries set by federalism, rights, and the rule of law. It bears noting this process is also facilitated by the federal government, which remains confident it can share jurisdiction but retain sovereignty over the border.

This is all to say that federalism remains relevant, of course, but circling back to the legal strength of sanctuary cities, the question isn’t whether cities do or do not govern migration, but whether they can protect against the disclosure of locally generated information. This is a jurisdictional question that engages not only federalism but also transversal normative framings related to security, rights, the rule of law, common law, administrative law, and so on. The fact is that, through security, migration is already governed at the level of the city while the illusions of dual sovereignty leave sanctuary cities ill-equipped to implement their policies.