Abstract
Executive Summary
Irregular migration at the Canada-US border, along with associated dangers, has received increased attention since the implementation of the Canada-US Safe Third Country Agreement. Reports by non-governmental organizations highlight the lethality of irregular migration at this border. Nevertheless, not only do reports fail to provide estimates of migrant fatalities, but there are also no official data sources on deaths produced by authorities in either country. Using this silence as a starting point, we ask: How many migrants perish while crossing the Canada-US border? To what extent, if at all, are US and Canadian officials tracking border-crosser deaths in this region? What can we learn from media reporting on this issue? We addressed these questions by (1) submitting public records requests to federal authorities in Canada and the United States, and (2) conducting an exhaustive search of media coverage on migrant deaths published by Canadian and US newspapers. Records requests resulted in just one migrant fatality recorded by the Royal Canadian Mounted Police. Yet, media reports confirmed 40 migrant fatalities, which most likely constitute an undercount. Given the overall underwhelming attention paid to deaths at the Canada-US border, we conclude by discussing the factors contributing to the lack of visibility of this ongoing human tragedy. Considering our study’s findings, we provide the following policy recommendations:
● The federal governments of the United States and Canada should create a classification system—like the one utilized by the Pima County Office of the Medical Examiner in Tucson, Arizona—that local officials in jurisdictions along the Canada-US border could use to categorize decedents as suspected irregular border-crossers.
● Increased communication, collaboration, and transparency between local authorities involved in the investigation of migrant fatalities at the Canada-US border and federal immigration officials in both countries. For example, authorities should liaise with their counterparts on the other side of the border for any migrant fatality investigation. We suggest the establishment of a working group modeled on the informal Working Group on Accounting for Migrant Deaths along the US-Mexico Border.
● Address jurisdictional challenges in the Canadian context that lead to a vacuum of official accountability in tracking migrant fatalities on Canadian territory and report on migrant fatalities at the border on an annual basis. For example, our research suggests that officials at the national headquarters of the Royal Canadian Mounted Police (RCMP) do not have access to data about fatalities investigated by their local detachments across the country. Migrant deaths at the Canada-US border are likely investigated by a range of agencies, including RCMP and local police forces that may be administered by provincial, municipal, and Indigenous governments.
● The US Missing Persons and Unidentified Remains Act of 2019, which is public law, requires that US Customs and Border Protection (CBP) annually report to the US Congress the number and location of migrant fatalities that occur at the U.S.-Mexico border, and that these estimates be disclosed publicly within 30 days of each reporting period, the latter of which CBP currently does not do. The Missing Persons and Unidentified Remains Act should be amended to include migrant fatalities at the US-Canada Border. Furthermore, CBP should adjust its public reporting of deaths at the US-Mexico border to comply with the law.
Keywords
Introduction
Attention to the plight of irregular border crossers at the Canada-United States border has grown steadily in recent years (Markusoff 2018). Changes to border policy, rising xenophobia in both countries, and a wave of migrant deaths have heightened media and societal concern about the border between the two countries. In March 2023, the Canada-US Safe Third Country Agreement (STCA) was extended to block asylum seekers entering one of the two countries from the other at any point across the entire land border, instead of restricting this to official ports of entry. Combined with the return of President Donald Trump to the White House in 2025, which set off heightened surveillance and enforcement on the Canadian side, STCA changes have driven migrants who do not meet the terms of the agreement to choose less visible, and often riskier, geographic locations when crossing the Canada-US border in either direction. Alongside highly publicized instances of the arrest and deportation of would-be migrants, the shifting policy context has been accompanied by a number of deaths in transit. In one tragic example, four members of the Patel family died while crossing into the United States from Manitoba during severe winter weather in January 2022 (Alam 2023).
Even as the attention of Canadian media and society has been drawn to irregular border crossings and instances of lives lost at the border, the details of the deaths remain cloudy, particularly in terms of the official response. Non-governmental organizations in both countries warn of the dangers associated with clandestine crossings in an era of heightened border enforcement; however, as we discuss below, our systematic review of NGO reports found no concrete estimates of border-crosser deaths near this border. In a similar vein, empirical data on border-crosser deaths along the Canada-US border are virtually non-existent in official sources. Given these clear trends, this paper poses the following interrelated research questions: How many migrants perish while crossing the Canada-US border? To what extent, if at all, are US and Canadian officials actively tracking border-crosser deaths in this region? What can we learn from media reporting about this issue? And, given the overall underwhelming attention paid to deaths at the Canada-US border, what explains the lack of visibility within Canada and the United States of this ongoing human tragedy?
As we set out on this research, we are aware of the long-standing concern and interrogation of the dangers of the Canada-US border by critical researchers. Writing a decade ago, for example, Helleiner (2013, 1508–1510) drew our attention to the rarely discussed dangers and even deaths that occur around the Canada-US border and that exist in tension with longstanding perceptions of this as a “benign” border context. She drew on media coverage and interviews with youth who had grown up in the Niagara region of southern Ontario to consider how both groups constructed unauthorized border crossings and the extent to which these perceptions influenced the reception of state-led border enforcement efforts locally. Helleiner’s piece sought to surface the structural violence of how the Canada-US border is managed and policed, including how border enforcement efforts are framed and understood. Our project takes up some of Helleiner’s concerns within the contemporary context.
This paper proceeds as follows: we begin by providing a high-level overview of the policy context that governs the Canada-US border and outlining our theoretical discussion of the invisibilization of migrant deaths. From there, we draw on data from US and Canadian officials to illustrate that substantial migration occurs between the United States and Canada, including clandestine migration attempts as well as the movement of individuals seeking asylum in each country. Next, we discuss the results of a review of media coverage of border-crosser deaths in US and Canadian news outlets from 1980 to 2024 and follow by describing the extent to which officials in each country actively track and publicly report on border-crosser deaths along the Canada-US border. We conclude by discussing the implications of our findings and articulate how they constitute a case of the invisibility of death, and by outlining concrete policy recommendations aimed at increasing public awareness and official recognition of migrant fatalities near the Canada-US border.
Reviewing the Policy Context
An important context of addressing our questions about migrant deaths is the increasingly collaborative approach to policing and enforcement of the shared border between Canada and the United States. A factor that distinguishes this border from the US-Mexico border is that irregular migration is taking place in both directions: from the US into Canada and from Canada into the US. Cooperation intensified in the aftermath of the events of September 11, 2001, when the two governments signed the Smart Border Declaration in 2002. One of its provisions reinforced existing collaborative border enforcement efforts in the form of Integrated Border Enforcement Teams (IBET). Begun as a pilot project in 1995, the IBET expanded under the Smart Border Accord with 23 teams covering 15 border regions; the teams coordinate on security matters and share information. The integrated approach to border governance was further reinforced in 2011, when the two countries launched the Beyond the Border Action Plan, which framed the border as a joint effort of control focused on securing the two countries rather than on the security and safety of people seeking to cross (Government of Canada 2018).
The Canada-US Safe Third Country Agreement (STCA) that came into force in 2004 is another legacy of the Smart Border Declaration. This agreement requires asylum seekers to make their claims in the first of the two countries they enter, unless they qualify for one of the exceptions. 1 The STCA reshaped the landscape of asylum seeking in relation to the shared border because of the unusual geography of its application for the first two decades it was in force: its terms only applied at official ports of entry. If a person seeking asylum crossed the border at an unofficial point and lodged their claim inland, they were not subject to the exclusion of this policy. This pushed migration journeys to unofficial and often remote and dangerous crossing points as illustrated by the recent rise in Border Patrol apprehensions between ports of entry as well as increased interceptions of irregular border-crossers by the RCMP. As we demonstrate below, these are sites where many of the border-crosser deaths we were able to enumerate occurred. The patchwork application of the STCA to the boundary line prior to 2023 also helps to account for the Canadian data on interceptions and “irregular border-crossers” discussed in the next section.
And yet, existing scholarship demonstrates that policies aimed at deterring migration do not stop people from crossing but instead shift patterns of mobility, pushing people to more remote and dangerous crossings, while also incentivizing smuggling operations (Rubio-Goldsmith et al. 2006, 31; Martínez et al. 2014, 263; Martínez 2016, 104; Gammeltoft-Hansen and Tan 2017, 37–38; Slack and Martínez 2018, 171; Boyce, Launius and Chambers 2019, 24; Molnar 2024). In other words, what we know about the consequences of restricting human mobility is that it not only changes the routes that are available but also leads to increased human suffering and death. Moreover, these consequences are masked by the seemingly routine operation of policies through which states obfuscate responsibility for their role in the outcomes. In this sense, we have seen simultaneously a deepening of collaboration in relation to border enforcement and an abdication of responsibility for the outcomes and impacts of this approach, including limitations of the rights of people on the move, raising the question of who is accountable for these impacts that include deaths at the border.
Drawing from the work of critical migration and border studies scholars, we make a case for an institutional theory of invisible death at the Canada-US border. “Invisibilization” here refers to the idea that the visibility of migrant deaths at the border is, in itself, not simply an artifact of statistical (non-)recording, but is rather produced through embedded power relations within political institutions nominally operating to produce ‘apolitical’ knowledge. From a Foucauldian perspective, authorities determine “irregular border-crossers” as making abnormal entries into the state by building, surveilling, and breaking down their subjectivity (Franko-Aas 2011, 337; Foucault 2004). At the same time, the construction of this problematic subjectivity requires that they become internalized to the state’s power in order that their presence can be administered via fused juridical-executive authority in achieving the state’s population management objectives (Maillet et al. 2018, 145). The result of this process is a rendering of limited social, political, and legal personhood akin to Butler’s (2004, 60) depiction of detainees or Agamben’s (2005) “state of exception” in camps for the purposes of attempting to control unauthorized migrant movement at the Canada-US border. In the context of our work, then, the failure to acknowledge and track deaths along the Canada-US border illustrates a disregard for the lives of non-citizens.
In this way, the (in)visibility of death at the border reflects power relations within political institutions whose statistical recording procedures are seen as apolitical due to their bureaucratic or humanitarian nature. But as Foucault (2014) argues in On the Government of the Living, even knowledge production that seeks to reveal what is already believed to be known (as opposed to what has been verified or falsified) directly embeds power in relation to truth; that is, the invisibilization of migrant deaths itself is not simply a process of statistical record-keeping but represents the product of a form of political technology in governance. Other scholars make parallel observations on this quality of invisibility of the implications of border policy choices to our own. Hiemstra (2019, 46), for instance, notes how the extension of US border policing activities through time and space—that is, beyond the territory of the country itself - helps to conceal their direct role in the impacts of these enforcement activities that violate human rights, distract from their policy failures, and evade their international obligations. Gilbert’s (2019, 426) analysis of the elasticity of borders applies similar ideas to the Canada-US border, arguing that existing policies, such as the STCA, expand the reach of state bordering practices even as this expansion allows state officials to deny accountability for the political outcomes of their policies and limit migrants’ rights in these expanded border spaces.
Invisibilization captures the ways in which outcomes are produced through policy processes. In this case, we argue that the invisibility of migrant deaths along the Canada-US border is a strategic outcome, a deliberate choice, rather than simply attributable to the non-counting or non-reporting of these deaths or the purported absence of official data. In this sense, policy choices produce particular outcomes - including dangers and deaths - that come to be seen as justified through the lens of what is understood as the normal and neutral application of laws and administrative practices.
Canada-US Border Crossings: Snapshots From Public Records
To provide context for the discussion of border-crosser deaths that follows, we begin with an overview of the publicly available data in the US and Canada related to “encounters” of migrants at the border, which, in the US context, includes apprehensions, inadmissibles, and expulsions. 2 This offers a sense of the broader enforcement context in which we attempt to make sense of the question of migrant deaths at the Canada-US border. US Customs and Border Protection (CBP) data from FY 2020–2024 suggest that US officials encountered 16 times as many migrants at the southern border with Mexico as at the northern border with Canada. Importantly, during this period, 80–95 percent of encounters at the southern border occurred between ports of entry by US Border Patrol (USBP) officers, which included Title 8 apprehensions and Title 42 expulsions (USCBP 2025a) The opposite was true at the northern border, where approximately 90 percent of encounters annually occurred at ports of entry where the Office of Field Operations (OFO) has jurisdiction, consisting mostly of Title 8 inadmissibles and Title 42 expulsions. Inadmissibles consist of (1) individuals seeking to enter the U.S. immigration system, including non-Canadian asylum seekers, (2) individuals denied entry due to missing documentation or who are deemed inadmissible, and 3) individuals who withdraw an application for admission and return to their countries of origin, including many Canadians. Though encounters by OFO have made up the majority of CBP encounters along the Canada-US border, encounters by USBP between ports of entry rose after 2022 (USCBP 2025a), suggesting that the share of individuals attempting to cross the border clandestinely increased.
Migration From North to South: US Customs and Border Protection Apprehensions at the US-Canada Border
US Customs and Border Protection (CBP), a federal agency within the US Department of Homeland Security, comprises two primary immigration enforcement entities: the CBP Office of Field Operations (OFO) and Border Patrol. In the context of the land borders the United States shares with Mexico and Canada, OFO is tasked with enforcement activities at official ports of entry, while Border Patrol carries out enforcement actions between ports of entry, including patrolling remote areas of the borderlands and staffing checkpoints within the 100-mile border zone.
Migration scholars focusing on the US-Mexico border have historically relied on Border Patrol apprehension statistics as a proxy for undocumented migration flows at the US’s southwestern border (Espenshade 1995, 198; Durand and Massey 2004; Martínez et al. 2017, 258). For instance, Border Patrol was responsible for between 72 and 93 percent of migrant encounters along the US-Mexico border from 2019 to 2024 (USCBP 2025b). However, we were surprised to learn that the context is substantially different at the Canada-US border. Table 1, which illustrates total encounters carried out by both CBP entities, communicates three critical findings. First, unlike at the US-Mexico border, Border Patrol is responsible for a surprisingly low percentage of encounters along the Canada-US border. Second, though Border Patrol encounters a smaller share of migrants at the northern border compared to the southern border, this proportion doubled from 5.9 percent in 2023 to 11.9 percent in 2024. Third, and perhaps most importantly, the total number of migrants encountered by CBP, which includes encounters by both USBP and OFO, has increased rapidly over the past five years, rising from roughly 32,000 in 2020 to 198,929 in 2024, representing a 514 percent increase.
U.S. Customs and Border Protection Encounters along the U.S.-Canada Border.
Note: Percentages may not sum to 100 due to rounding.
USBP = U.S. Border Patrol (between ports of entry); USBP encounters include Title 8 apprehensions and Title 42 expulsions.
OFO = Office of Field Operations (at ports of entry); OFO encounters include Title 8 inadmissibles and Title 42 expulsions.
While it is possible that 2020 and 2021 represented anomalies for immigration enforcement actions due to the COVID-19 pandemic, Border Patrol encounters along the Canada-US border had been steadily declining since the early 2000s (see Figure 1). Unfortunately, OFO encounters data for this border are not publicly available beyond the most recent five fiscal years. Nevertheless, Table 1 and Figure 1 collectively illustrate that north-to-south migration at the Canada-US border has increased in recent years among both clandestine migrants and asylum seekers. Moreover, Figure 1 demonstrates that Border Patrol encounters in 2024 were the highest on record since 2000.
Total U.S. Border Patrol Northern Encounters by Fiscal Year.
Table 2 disaggregates CBP encounters along the Canada-US border by country of citizenship, illustrating data for both USBP and OFO. The overall number and proportion of encounters of migrants from India and “Other” countries have increased, while the proportion of Canadians encountered has declined in recent years.
US Customs and Border Protection Encounters along the US-Canada Border by Citizenship.
Note: Percentages may not sum to 100 due to rounding.
USBP encounters include Title 8 apprehensions and Title 42 expulsions, while OFO encounters include Title 8 inadmissibles and Title 42 expulsions.
However, we can gain additional insights by further disaggregating these encounters by specific CBP components, as reflected in Table 3. The top panel in Table 3 illustrates the countries of origin among encounters carried out by OFO at ports of entry, which includes both Title 8 inadmissibles and Title 42 expulsions. In contrast, the bottom panel provides this information for encounters by Border Patrol between ports of entry, which includes Title 8 apprehensions and Title 42 expulsions.
US Customs and Border Protection Encounters along the US-Canada Border by Citizenship and Component.
Note: Percentages may not sum to 100 due to rounding, OFO encounters include Title 8 inadmissibles and Title 42 expulsions, while USBP encounters include Title 8 apprehensions and Title 42 expulsions.
We find substantial differences in terms of the representation of countries of origin by CBP component. For example, in 2023 and 2024, Canadians accounted for nearly 21 percent of encounters at ports of entry, but less than 2 percent of encounters between ports of entry. Meanwhile, Mexican nationals constituted just 3.5 percent of encounters at ports of entry, but 48.7 percent and 14.8 percent of encounters between ports of entry in 2023 and 2024 3 , respectively. Chinese and Filipino migrants, on the other hand, are seldom encountered by Border Patrol compared to being stopped by OFO at ports of entry. Indian nationals constituted a similar proportion of encounters at ports of entry and between ports of entry in 2023. Nevertheless, the proportion of US Border Patrol encounters consisting of Indian nationals increased exponentially in 2024, representing a significant contributor to the overall doubling of Border Patrol encounters at the Canada-US border from 2023 to 2024. Presumably, individuals crossing between ports of entry, to either avoid detection by US officials or to turn themselves in to initiate the asylum process, should be at the highest risk of death while in transit. Therefore, it stands to reason that migrants from India, Mexico, and “Other” countries should account for the largest share of those perishing near the Canada-US border while traveling southbound.
Migration From South to North: RCMP Interceptions of “Irregular Border Crossers,” at the Canada-US Border
In Canada, while the Canada Border Services Agency (CBSA) is responsible for enforcing the border at official ports of entry, the Royal Canadian Mounted Police (RCMP) supports this role through enforcement between ports of entry. Consequently, those who make “irregular” border crossings (i.e., cross the Canada-US border between ports of entry) become subject to border control via RCMP “interceptions.” According to the Government of Canada (2024a), “RCMP interceptions” in their public data specifically refer to asylum seekers apprehended between ports of entry. This division of labor between CBSA and RCMP reflects the 2015 federal Ministerial Responsibilities Under the Immigration and Refugee Protection Act Order, whereby human smuggling and irregular arrivals fall under the jurisdiction of the Ministry of Public Safety and Emergency Preparedness (Government of Canada 2023b). We coalesce and briefly analyze the RCMP interceptions data here.
It is important to note that the Canadian government does not make the same kinds of records available to the public as are available in the US. Interestingly, officials did start tracking what they termed “interceptions” when the number of individuals crossing the border between ports of entry increased dramatically starting in 2017. As of June 2025, the Government of Canada provides data on RCMP interceptions during the time period of January 2017 to April 2025; these data include information on the time (month) and geography (province/territory) of interception of irregular border-crossers at the Canada-US border. According to these data, at least 119,694 asylum seekers irregularly crossed the border during this time. As shown in Table 4, 96 percent of these asylum seekers crossed the border in Quebec, reflecting the importance and operation of the de facto border crossing point at Roxham Road, where the RCMP operated infrastructure to process asylum seekers.
RCMP Interceptions of Irregular Border-Crossers at the Canada-US Border, by year and geography.
Source: RCMP (2017-2024).
The data for 2024 only include the month of January.
Interceptions of asylum seekers at the Canada-US border can be grouped into four main phases (Figure 2). From 2017 to 2019, there were substantial numbers of irregular border-crossers intercepted by the RCMP, which were facilitated by the Trump administration’s policies against asylum seekers in the US, as well as the Trudeau administration’s processing of asylum seekers at the Roxham Road unofficial port of entry. From 2020 to 2022, the Quarantine Act prohibited such processing and, more generally, the arrival of asylum seekers between ports of entry, resulting in a sharp decrease in recorded numbers of asylum seekers intercepted by the RCMP. A third phase began from 2022 to 2023, when the Government of Canada lifted the quarantine restrictions and once again allowed processing to occur between ports of entry. The fourth and final phase runs from began in 2023 to the present during the implementation of the STCA’s additional protocol, which once again prohibited asylum seeking between ports of entry.
RCMP Interceptions of Irregular Border-Crossers at the Canada-US Border, by Month.
IRB Irregular Border Crosser Refugee Claims
Additional data on “irregular border crossers” at the Canada-US border are available from the Immigration and Refugee Board of Canada (IRB 2024), which publishes statistics about this category on a quarterly basis. As of June 2025, these data cover the period of February 2017 to March 2024. In particular, the IRB reported that during this time they had processed at least 90,309 refugee claims by people who made irregular border-crossings at the US-Canada border, with an additional backlog of another 23,942 claims yet to be determined at the end of March 2025. In general, the claim acceptance rate among irregular border-crossers making refugee claims in Canada during this time was about 60 percent (with a rejection rate of about 29 percent); this compared to about 62 percent for all asylum seekers making refugee claims with IRB during this time. This acceptance rate for irregular border-crossers, moreover, increased to about 73 percent by 2024.
The IRB also records information on the top countries of origin for asylum seekers who arrived via irregular border crossings. According to these data, the countries with the highest rate of intake for refugee claims by asylum seekers who made irregular border crossings at the Canada-US border were Haiti (about 21 percent of all claims), Nigeria (16 percent), Colombia (10 percent), Turkey (9 percent), Pakistan (4 percent), Venezuela (3 percent), the Democratic Republic of the Congo (3 percent), Angola (3 percent), Afghanistan (3 percent) and Sudan (2 percent); all other countries made up about 27 percent of such claims. In general, the geography of countries of origin for those who made irregular border crossings was highly varied.
The official data sources we discussed in this section illustrate that irregular migration is a social phenomenon that not only occurs at the United States’ southern border but also along the Canada-US border, with individuals from diverse countries migrating from both north to south and south to north, including clandestine migrants and aspiring asylum seekers crossing between official ports of entry. Moreover, bidirectional irregular migration between these countries seems to have increased in the years following the COVID-19 pandemic. Persistent irregular migration at Canada’s southern border carries a real risk of death, particularly when individuals attempt to migrate between official ports of entry to avoid detention by US and Canadian officials.
Documented Deaths at the Canada-US Border: Public Records and Media Reports
The questions that motivate our research are straightforward: how many migrants die crossing between Canada and the United States, and how do the two governments track and report on these deaths? To enumerate border-crosser deaths along the Canada-US border, we first engaged in Access to Information requests to the Royal Canadian Mounted Police (RCMP) in Canada and Freedom of Information Act requests to US Customs and Border Protection (CBP) in the US, both of which led to official denial of information on deaths. We then turned to a review of academic articles and of reports published in the last 20 years by non-governmental organizations (NGO). No systematic analysis of death at the Canada-US border has been published to date in the academic scholarship. Despite mentions of the lethality of irregular crossings between the United States and Canada, no specific deaths were accounted for in 35 reports by five different NGOs. Finally, we conducted a systematic review of Canadian and US print media through NexisUni from 1980 to 2024, which allowed us to identify 40 deaths.
Requests for Public Records from Government Agencies
As part of the cross-checking of data from multiple sources, we placed an Access to Information and Privacy (ATIP) request in Canada and a Freedom of Information Act (FOIA) request in the United States, seeking official estimates of migrant deaths tracked by both countries. Both processes resulted in lengthy delays, and after two years, we received information on only one case from Canadian authorities, with no cases from their US counterparts.
The Canadian Access to Information Act gives a period of 30 days for response to requests, but allows for extensions at the discretion of access officers (Brownlee and Walby 2015). A request placed with the RCMP in July 2022 was answered but indicated we were to provide a list of each RCMP detachment subject to the claim since, according to an RCMP access to information officer, “in many instances the investigative information is strictly available from the unit conducting the investigation” (RCMP 2022). The RCMP ignored a request to provide a map of detachment locations, and so the request required a review on our part of the 722 RCMP detachments nationwide, and the creation of a list of 94 of those that cover a section of the border. We were then asked to limit the request to statistics, in order to avoid requesting personal information. After agreeing to this change, communication fell silent, despite multiple emails on our part. In January 2023, seven months after submitting the original request, we asked for an investigation into the request through the Office of the Information Commissioner. The claim was assigned to an investigator in June 2023, and in July 2023 the RCMP replied, indication that information on border crossing deaths was held by local police forces, rather than the RCMP.
The same request was delivered in the United States through the Freedom of Information Act, requesting records of migrant deaths at the northern border held by the US Customs and Border Protection (CBP) and/or the US Border Patrol. Four months after the September 2022 request, the FOIA office replied only that no fees were required. In November 2023, fourteen months into a process that, by law, should have been resolved within 20 working days (Hiemstra and Conlon 2021, 668), CBP responded with a link to a public website that did not, in fact, contain any information on deaths along the Canada-US border. Rather, the link provided was simply to CBP’s “Border Rescues and Mortality Portal,” with which the authors were already quite familiar, as it provides US Border Patrol migrant death estimates for the United States’s southwestern border for the fiscal years 2018–2022.
In summary, formal requests for information from both the Canadian and US governments led to official denials that they held information on deaths at the Canada-US border.
Review of Academic and Advocacy Publications
A review of the academic literature confirmed that a systematic analysis of death at the Canada-US border has not been published to date. While Helleiner (2013, 1508, 2016) extensively assessed the danger of unauthorized border crossings in the Niagara region, and Bauer (1997, 680) reviewed deaths in the context of human smuggling operations in the 1990s, a review of cases and trends in migrant deaths on both sides of the border has not been conducted before now. Similarly, non-governmental organizations advocating for migrant rights have not compiled data on known deaths at the Canada-US border. We reviewed decades’ worth of reports published by Amnesty International, the Canadian Council of Churches, the Canadian Council for Refugees, and the Harvard Immigration and Refugee Clinical Program, and found no information on specific deaths.
Systematic Review of Canadian and US Media Coverage
The most productive data source turned out to be media coverage of deaths along the border. Using the NexisUni database to access articles published by US and Canadian news and media outlets between 1980 and 2024, we were able to document a total of 40 deaths. We tracked migrant- and death event-level data based on information garnered from each media report, although some data were not available for every case. We recorded the age, sex, country of origin, cause of death, location of death (i.e., geography), and year/fiscal year of death. We also documented the crossing direction of the migrant (Canada to US; US to Canada), the US Border Patrol sector corresponding to each death, the side of the border on which the migrant’s remains were recovered (Canada, US, or Border Crossing), and whether media reports mentioned smuggling as being suspected during migration.
Though we recorded just 40 media-reported deaths of migrants along the Canada-US border, we present our descriptive results from those data in Table 5 below. Based on the 34 migrants whose age was reported by the media, the mean age of border-crossers at the time of their death was 29.7 for the entire sample. Three young migrants did not have their sex reported, but had a mean age of 3.3. For country of origin, 10 migrants (25%) were from India, five (12.5%) were from the United States, four (10%) were from Romania, four (10%) had a country of origin that was unknown or not reported on, three (7.5%) were from Mexico, two (5%) were from Pakistan, and two (5) were from Senegal. One migrant (2.5%) was from each of the following countries: the Dominican Republic, Ghana, South Korea, Zimbabwe, the United Kingdom, Peru, South Africa, and Haiti. Drowning was the most prevalent cause of death in these reports, with 23 (57.5%) of 40 migrants dying as a result. Hypothermia was the next most prevalent cause of death with eight (20%) migrants dying from the condition, followed by five (12.5%) deaths because of miscellaneous (i.e., “Other”) occurrences, three (7.5%) deaths as a result of border enforcement-related incidents, and one (2.5%) migrant dying from a medical condition. The five miscellaneous occurrences included a migrant being crushed by a train, a mother accidentally smothering her infant child while hiding from Border Patrol, a migrant becoming entangled in the driveshaft of a bus, and a suicide. The three border enforcement-related incidents included two migrants being fatally shot by Border Patrol, and one migrant dying as a result of a high-speed automobile chase with Border Patrol and a subsequent vehicle crash.
Descriptive Statistics for Media-Reported Migrant Deaths.
Note: N = 40.
The most prevalent US Border Patrol sector where media-reported migrants died was the Buffalo Sector, an area where 21 (52.5%) migrants died, 12 of whom drowned in the St. Lawrence River/Akwesasne territory, a site that media reports prominently discussed as a popular area for smugglers to cross with migrants. 2023 was the fiscal year with the most media-reported migrant deaths at 9 (22.5%). Thirty (75%) migrants were crossing from Canada to the US, with 22 (55%) migrants’ remains recovered on the Canadian side of the border, six (15%) migrants’ remains recovered on the US side, six migrants whose remains were not recovered (15%), and four (10%) whose remains were recovered at a border crossing. Nineteen (47.5%) media reports wrote that smuggling was either suspected to have facilitated or did facilitate migrants’ journeys. While only seven reports disclosed an estimated cost of smuggling, the average of those estimates was $4,178.57.
The Invisibilization of Death
Our research reveals that people from around the world are dying on a regular basis while attempting to cross the Canada-US border. Our best understanding of these deaths is found in media reports, but we suggest that not all cases attract the attention of the media and as a result, this is likely an undercount of deaths at this border. Due to the absence of official data, we do not know how many people are dying at this border or under what conditions. Here we discuss some of the factors that produce this lack of data about and awareness of migrant deaths at the Canada-US border, which we characterize as the result of invisibility as a policy choice. As noted earlier, the past two decades of shared border enforcement between Canada and the US have been characterized by a deepening of collaboration even as responsibility for the dangers that ensue is abdicated, obfuscated, and hidden. Invisibilization of migrant deaths, then, is not accidental or unintended but rather a deliberate outcome of decisions to not record and/or not report on data. We argue that lack of prioritization by government agencies and the low transparency of public institutions with responsibility for border enforcement contribute to the invisibilization of death along the Canada-US border. Indeed, they are the bureaucratic practices through which both states effect this policy choice.
Lack of Priority by Government Agencies
The US Border Patrol has tracked migrant deaths along the US-Mexico border since fiscal year 1998 as part of the agency’s Border Safety Initiative Tracking System (BSITS). In fact, after waiting for fourteen months, our FOIA request to CBP inquiring about migrant deaths at the US-Canada border resulted in the agency simply providing a link to a public website that reports on fatalities at the US-Mexico border from 2018-2022. However, based on our research on the Canada-US border, it is quite evident that neither US nor Canadian officials are systematically tracking border-crosser deaths in the region.
Although the US Border Patrol does track migrant deaths along the southwest border, several studies drawing on local medical examiner and medical investigator reports, including studies conducted by the Binational Migration Institute at the University of Arizona, have found that Border Patrol has grossly undercounted migrant deaths across the US-Mexico border (Leutert et al. 2020, 7; Martínez et al. 2021, 258; Leutert 2024, 279; No More Deaths 2024). For example, from 2014 to 2022, the Pima County Office of the Medical Examiner in Tucson, Arizona, documented at least 668 migrant deaths that were not accounted for in the Border Patrol’s estimates in the Tucson Sector. This a major problem considering there is no other independent systematic effort to track and enumerate migrant deaths across the US-Mexico border. If policymakers, journalists, or even the public want to get a sense of migrant death fatalities border-wide, they often turn to the Border Patrol’s flawed estimates. By virtue of coming from the federal government, these estimates are granted an element of legitimacy that other sources may not be afforded. Leutert (2024, 280–281) suggests several factors might help explain this discrepancy, including whether Border Patrol was the reporting agency or was on the scene when the remains were recovered, the condition of the remains—skeletonized remains are more likely to be undercounted by Border Patrol—and whether or not the decedent had any identifying documents on them when their remains were recovered.
Nevertheless, despite the reasons for this gross undercount, Border Patrol has had numerous opportunities to reconcile its database with estimates from local authorities such as the Pima County Office of the Medical Examiner in Tucson, Arizona. In the case of the Tucson Sector of the Border Patrol, it has failed to do so because, as a 2022 US Government Accountability Office report detailed, the individuals tasked with “coordinating with external entities to obtain data on migrant deaths” are not the same individuals tasked with entering the data into the BSITS system (US Government Accountability Office [GAO] 2022, 17). There has also been “confusion as to the definitions of deaths and rescues, and that because of supervisor turnover, not all Border Patrol supervisors are familiar with BSITS” (GAO 2022, 17). This is even though Border Patrol is required to report annual migrant death estimates to the US Congress and make them publicly available as part of the agency’s Missing Migrant Program (MMP), 4 which was launched in 2017, as well as the Missing Persons and Unidentified Remains Act (MPUR) of 2019 (2020, 5). In short, based on data provided to the public, the US Border Patrol continues to engage in a systematic undercount of deaths along the US-Mexico border because it is not an organizational priority, and it lacks the will to do so. In sum, though the 1998 BSITS, the 2017 MMP, and the 2019 MPUR mandate the Border Patrol to track migrant deaths along the US-Mexico border, the agency has yet to make this an institutional priority.
Where does this leave the tracking of deaths occurring at the Canada-US border? We argue that what we are seeing, or failing to see, along the northern border of the US is symptomatic of this lack of prioritization when it comes to enumerating the dead, which constitutes a form of “invisible death.”
Low Transparency of Public Institutions
This lack of prioritization observed in relation to the tracking of border deaths by US officials at the more high-profile southern border is reinforced by the challenges we faced in response to our requests for information from public institutions charged with border enforcement on both sides of the border. As noted above, our requests to Canadian and US institutions for official data on migrant deaths were met with delays and, ultimately, claims that the data do not exist at the federal level.
Non-existent data and delays beyond legally-mandated timeframes join a long list of structural problems within both the Access to Information and Privacy (ATIP) and Freedom of Information Act (FOIA) systems. In their analysis of reluctantly delivered and heavily redacted government documents on immigration detention contracts in the United States, Hiemstra and Conlon (2021, 669) argue that the absence of data through redaction or non-provision holds its own significance, since to withhold information demonstrates its importance to the processes under scrutiny as well as to the operation of state power. Reflecting on three research projects pertaining to detention and border enforcement, Maillet, Mountz, and Williams (2017, 929) present adaptive, but ethically complicated, solutions to the difficulty of accessing research data. For Williams (2018, 11–12), whose access to information requests on migrant boat interceptions by the European Union returned hundreds of pages of redacted material, the answer lay in filling in data gaps through research in additional sources. Williams’ methodology, repeated as part of the research for this paper, produced a dataset where state institutions had tried to prevent its release (see also Williams and Mountz 2018, 84).
In Canada, researchers regularly decry systemic and politicized problems with the ATIP model. National print media publication The Globe and Mail recently responded to this scenario with the launch of a multi-year project dubbed “Secret Canada.” A Globe and Mail audit of every government ministry and department in the country found that just 19 percent of all information requests in 2021 were granted in full while 37 percent were delivered with redacted results. When results were challenged, 59 percent of rulings found that information had been wrongfully withheld. The audit, together with hundreds of expert interviews, led the authors to conclude that “political leaders have effectively created a freedom of information regime that incentivizes institutions to keep public records secret” (Doolittle and Cardoso 2023). Monaghan (2015) similarly observes that barriers to public information access include the political control of information, systemic time delays, excessive fees, superficial and arbitrary document retrieval, and redaction practices that abuse legally-mandated parameters. In the US context, Hiemstra and Conlon (2021, 678) hold that the redaction and removal of information from their claims did not reflect individual US bureaucrats “acting with some larger aim of protecting the government. . .or publicly asserting the dominance of the state. In the Canadian context, however, Monaghan (2015) documents practices that run from ministerial review of information prior to its release, to the normative context of information officers sharing workspaces and supervisors with the people who create documentation subject to ATIP requests. That ATIP requests are filled by members of the same institution holding the information points to a conflict of interest structured into the system.
The delayed response to our requests and the absence of information on migrant deaths recorded at the Canada-US border doubtlessly corresponds with problems larger than any delays or errors related to these specific requests. A refusal to provide data or communicate with researchers reflects not only systems in disarray through underfunding and inadequate support (Cardoso, Curry, and Doolittle 2023; Doolittle and Cardoso 2023), but also information systems that are intended to prevent democratic scrutiny while actively projecting state power (Monaghan 2015; Hiemstra and Conlon 2021, 680). As this applies to the death of migrants crossing the Canada-US border, a scenario of regular fatalities that frequently finds itself far from the press cycle and public concern becomes further invisibilized through the non-provision of government data.
Conclusion
Our research produced the first-ever compilation and review of media coverage of deaths at the Canada-US border. Unfortunately, the lack of official data on deaths means that the cases reported in media sources cannot be understood to represent or capture all cases. Importantly, previous research by Helleiner (2013, 1516–1517) pointed to knowledge by Canadian officials of deaths in the Niagara region of the border that were not reported on by the media.
We have identified at least 40 migrant deaths along the Canada-US border from 1980 to 2024, with 17 of these deaths having occurred during the 2020–2024 period, which corresponds with the influx of irregular migrants apprehended between official ports of entry by both US and Canadian officials during this same period. Moreover, 10 of the 40 deceased individuals were Indian nationals, who have made up an increasing proportion of migrants apprehended by US authorities in recent years. However, this identification of deaths at the border is likely an undercount as our data represent only cases that have received media attention. Ultimately, we do not know how many people have died crossing the Canada-US border in the period under investigation and, under current state practices, we cannot know. The US Border Patrol collects data on deaths at the U.S. southern border with Mexico, although these are systematically undercounted, whereas deaths that occur at the northern border are not a priority under US policy. Even if the US Border Patrol is in fact actively collecting data on these deaths, the agency (1) does not make the information publicly available, and (2) has failed to provide the information despite FOIA requests.
Our research illustrates that media reports confirm at least 40 individuals have lost their lives while crossing the Canada-US border. Nevertheless, public institutions in both countries denied that they had data on migrant deaths at the border. In this sense, the generation of reliable data on deaths at the Canada-US border is systematically denied by the Canadian and US governments through policy decisions that prevent data collection and make deaths invisible.
It is crucial to connect this invisibilization of migrant deaths to the policy context in which it has unfolded. The Canada-US Safe Third Country Agreement (STCA) is especially germane to understanding patterns and experiences of migration across the shared border, given its focus on the management of asylum seeking in a shared North American geopolitical space. The announcement in March 2023 that the two countries had renegotiated the agreement to apply to the entire land border offers a window into the structural violence produced by this policy. The Canadian government’s Regulatory Impact Analysis Statement that accompanied the announcement identified the anticipated “costs or risks” of the expanded policy in this way:
Asylum seekers . . . may look to evade application of the STCA by attempting to cross in more remote areas, and remain undetected until they are eligible to make a claim for asylum. In these instances, claimants may face increased danger, such as involvement with human smugglers and may be at risk for physical, mental, or financial abuse. They may also face risks from exposure to extreme weather conditions if they cross at remote locations or fail to secure access to shelter (Government of Canada 2023b).
Here, Canadian officials acknowledge that extending the policy to the entire Canada-US border will create more dangerous situations for people attempting to cross the border that they would not face in the absence of the policy. The excerpt is jarring precisely because it explicitly names the implications of pursuing this operational change to the policy even as we understand that Canadian - and US - officials have already moved ahead with its implementation. Indeed, an April 2023 Public Safety Canada memo warned that expanding the STCA “would likely fuel smuggling networks and encourage people to seek more dangerous, remote crossing routes” (cited in Bronskill 2023). This illustrates the ways in which policy choices produce particular outcomes, including danger and deaths, while also directing our attention to the invisibilization of the impacts of these policies.
Considering our study’s findings, we provide the following policy recommendations:
The federal governments of the United States and Canada should create a classification system—modeled after the approach utilized by the Pima County Office of the Medical Examiner in Tucson, Arizona—that local officials in jurisdictions along the Canada-US border could use to categorize decedents as suspected irregular border-crossers.
Increase communication, collaboration, and transparency between local authorities involved in the investigation of migrant fatalities near the Canada-US border and federal immigration officials in both countries. For example, authorities should liaise with their counterparts on the other side of the border for any migrant fatality investigation. We suggest the establishment of a working group modeled on the informal Working Group on Accounting for Migrant Deaths along the US-Mexico Border (see e.g., Kerwin and Martínez 2024).
Address jurisdictional challenges in the Canadian context that lead to a vacuum of official accountability in tracking migrant fatalities on Canadian territory and report on migrant fatalities at the border on an annual basis. For example, our research suggests that officials at the national headquarters of the Royal Canadian Mounted Police (RCMP) do not have access to data about fatalities investigated by their local detachments across the country. Migrant deaths at the Canada-US border are likely investigated by a range of agencies, including RCMP and local police forces that may be administered by provincial, municipal, and Indigenous governments.
The US Missing Persons and Unidentified Remains Act of 2019, which is public law, requires that US Customs and Border Protection (CBP) annually report to the US Congress the number and location of migrant fatalities that occur at the US-Mexico border, and that these estimates be disclosed publicly within 30 days of each reporting period, the latter of which CBP currently does not do. The Missing Persons and Unidentified Remains Act should be amended to include migrant fatalities at the US-Canada Border. Furthermore, CBP should adjust its public reporting of deaths at the US-Mexico border to comply with the law.
This paper represents a step toward an illumination of data on migrant deaths, to uncover figures and analyze trends while countering the official silence of the state. We argue that considering this institutional invisibilization of migrant deaths in relation to border enforcement policies that contemplate the risk of death as an anticipated consequence clarifies that deaths at the border are a price that Canadian and US officials have decided they are willing to pay in exchange for the political win of being seen to attempt to manage migration. There is reason for grave concern. The increasing number of border-crosser apprehensions at the Canada-US border demonstrated in US and Canadian data examined above points to the region as a significant space of migration and border enforcement. This trend also increases the likelihood of injury and death as people seeking safety and security by crossing this border attempt to evade border enforcement efforts.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was undertaken, in part, thanks to funding from the Canada Research Chairs Program.
ORCID iDs
1
People may qualify for an exception if: they have a close family member who is a citizen, permanent resident, protected person, or accepted refugee; they are an unaccompanied minor; they hold a valid visa or are not required to hold a visa to enter Canada while they would be to enter the US; they have been charged with or convicted of a crime that would subject them to the death penalty in the US or another country (
).
2
The term “encounters” was used by the US Department of Homeland Security while Title 42 expulsions were in effect from March 2020 to May 2023. Encounters included individuals apprehended between official ports of entry by the US Border Patrol (USBP) pursuant to Title 8, those deemed inadmissible at ports of entry by Office of Field Operations (OFO) under Title 8, and those expelled pursuant to Title 42 by both USBP and OFO (Office of Homeland Security Statistics 2025; also see for a discussion of Title 8 apprehensions and Title 42 expulsions).
3
4
The US Department of Homeland Security changed the name of the “Missing Migrant Program” to the “Missing Alien Program” at the beginning of Donald J. Trump’s second term as US President.