trictly Border Cases’

On October 28, 2020, two workers arrived from Hawai’i in the US Army Garrison at Kwajalein on the Marshall Islands and tested positive for COVID-19. Authorities quickly assured the public that these were “strictly border cases”: since the workers entered quarantine at the border, they posed no risk of community spread to the islands. On November 15, three more Marshallese workers returning home from Hawai’i had positive tests. These cases brought an end to the Marshall Islands’ run as one of the only countries in the world with no known COVID cases.

While the Marshall Islands have only recently begun to reckon with COVID’s arrival, the pandemic has throttled Marshallese communities in the diaspora for nearly a year. In Northwest Arkansas, home to the largest Marshallese population in the diaspora, Marshallese COVID rates and mortality rates have been astronomical since March 2020: while Marshall Islanders make up no more than three percent of Northwest Arkansas’s population, by mid-June they accounted for half of all COVID deaths in the area. In Springdale, Marshallese community leaders have been working around the clock to protect and provide for their community members. Skyrocketing COVID rates have also besieged other Pacific Islander communities around the country: by May, Pacific Islanders in California, Oregon, and Utah were dying from COVID at nearly twice the rate of the overall state populations, and by July, Pacific Islanders were being hospitalized with COVID at up to 10 times the rates of other groups.

At first glance, these places—Northwest Arkansas, with its vertigo-inducingly high COVID rate, and the Marshall Islands, with its exceedingly low one—look like polar opposites. Yet Northwest Arkansas, a new immigrant destination, and the Marshall Islands, a former US territory, are inextricably linked by the skeins of US empire. The two sites are connected not only by the human topographies of diaspora, but also by colonial ideologies that understand islands as isolated and distant (based in environmental determinism) and races as categorically and genetically distinct (based in biological determinism, a cornerstone of eugenics). At the heart of these paired discourses and the practices they undergird is the belief that vulnerability is intrinsic to certain places and populations. Therefore, a geographical critique of environmental determinism and biological determinism—one that focuses on their entanglements—must identify the material histories and power dynamics of US empire that have produced certain people and places as vulnerable. Furthermore, geographic attention to the mobilities engendered by empire reveals that US empire produces islanders’ vulnerability not (only) through distancing and abandonment, but also through exposure, intimacy, and contact.

In Black Marxism, Cedric Robinson (2000, p. 16, emphasis added) observed that the “tendency of European civilization through capitalism was thus not to homogenize but to differentiateto exaggerate regional, subcultural, and dialectical differences into ‘racial’ ones”. Exaggerating regional differences into ‘racial’ ones, of course, has also been a key project of the eugenics movement. Scholars like Laura Briggs (2002) have shown how eugenic thought has long informed US policies implemented on islands, not just through health and population control measures but also around military operations like nuclear testing. In other words, racial concepts are always spatial concepts, as Robinson, Ruthie Gilmore (2002), and others have shown.

What does it mean that racial concepts are spatial concepts? And why does it matter? As I argue in this essay, US empire produces vulnerability unevenly in both people (racial) and places (spatial), and then deploys discourses of crisis and disaster—presented as natural phenomena—in order to erase its own role in manufacturing them. Yet as Neil Smith (2006) observed, in his indictment of the US government’s abandonment of Hurricane Katrina’s (largely Black) victims, “it is generally accepted among environmental geographers that there is no such thing as a natural disaster. In every phase and aspect of a disaster—causes, vulnerability, preparedness, results and response, and reconstruction—the contours of disaster and the difference between who lives and who dies is to a greater or lesser extent a social calculus.” If we read Smith’s essay as an insight not only into environmental disasters but also into public health crises—recognizing that the two are often coterminous—we can begin to connect how the state produces vulnerability to disaster in places and people through the coupled practices of exposure and abandonment. We can then unmask the logics of environmental determinism and biological determinism in disaster discourses, which locate the origins of risk respectively in places and people themselves.

Crises lay bare the existing structures of our world, exposing preexisting fault lines and uneven topographies. Again, Neil Smith: “[D]isasters don’t simply flatten landscapes, washing them smooth. Rather they deepen and erode the ruts of social difference they encounter.” A global disaster, COVID-19 has both produced and revealed a horrifyingly uneven topography that has been anything but natural, prompting us to ask “whether colonialism itself might not be best understood as a kind of disaster” (Bonilla, 2021, p. 1). This uneven topography results not only from the real-time effects of organized abandonment (Gilmore, 2008; Harvey, 1989) but also from sedimented historical layers of imperial violence, which are also layers of racial violence. A closer look at the pandemic’s effects on the Marshall Islands and the Marshallese diaspora, and at longstanding discourses now being dusted off to explain these effects, reveal once again how US empire produces vulnerability for islands and islanders.

Preexisting Conditions: US Nuclear Testing and Exposure in the Marshall Islands

For Marshall Islanders, the COVID pandemic is a point of crisis in an attenuated history of slow violence (Nixon, 2011). The US military presence and nuclear testing in the Marshall Islands created the conditions of Marshall Islanders’ present-day health risks, even as US military officials have justified occupation of the islands as protecting their inhabitants from encroachment by (other) foreign invaders. In 1946, the US took on the Marshall Islands as UN trust territory, then promptly used the islands in an extensive and devastating nuclear testing campaign. The US military’s selection of Bikini Atoll for nuclear testing was premised upon the idea of the islands’ remoteness, as one US military officer succinctly conveyed: “We just took out dozens of maps and started looking for remote sites. After checking the Atlantic, we moved to the west coast and just kept looking” (Davis, 2005, p. 613). When asked about the search for a test site, he responded: “It looks like pretty far away is going to be the answer.” In short, the US portrayed the Marshall Islands as geographically distant—and the Marshall Islanders’ as racially distant or Other—in order to justify a terrifying kind of intimacy, as US doctors and scientists confidently probed the affected lands, waters, and human and non-human bodies in the nuclear fallout zone.

Image 1: A medical inspection of a Marshallese woman by an American doctor. Source: Nuclear Savage: The Islands of Secret Project 4.1 (Adam Jonas Horowitz, 2011; Creative Commons)

The US military uses of the Marshall Islands and Marshallese people produced vulnerability not only to the nuclear testing’s immediate after-effects—including cancers, infertility, and other health issues—but also to future exposure. At Runit Dome, for example, the US deposited “more than 3.1 million cubic feet — or 35 Olympic-sized swimming pools — of US-produced radioactive soil and debris” (Rust, 2019). Now, as the islands’ first documented COVID cases at Kwajalein make clear, risk often comes from the Marshall Islands’ interface with the ‘outside world’—from their closeness to US imperial power, rather than their isolation or remoteness from it.

Formally independent since 1986, the Republic of the Marshall Islands (RMI) still maintains deep political, economic, and military ties to the US, ties laid out in the US-RMI Compact of Free Association, which, among other things, granted the US the right to deny other countries military access to the islands and surrounding waters. The COFA also laid out a unique legal status for citizens of the Freely Associated States (FAS)—the Marshall Islands, Micronesia, and Palau—which allowed them to live in the US indefinitely without a visa. This legal status would follow Marshall Islanders out into the diaspora as they left the islands in droves after 1986, seeking educational and employment opportunities, and, more recently, due to concerns over climate change and sea-level rise. Many headed for Northwest Arkansas, for its relatively low cost of living and availability of jobs in the poultry industry. There, they have experienced new forms of racism, what Gilmore (2007) calls state-sanctioned vulnerability to premature death (Gilmore, 2007).

Food Work, Care Work, and ‘Close Contacts’

While the COFA’s migration provision offered a conditional set of benefits for Marshall Islanders, it also exacerbated the health-related vulnerabilities that shape Marshall Islanders’ lives in diaspora. In Northwest Arkansas, Marshall Islanders make up a disproportionately large segment of the poultry workforce, working in Tyson plants where physically close working conditions have been shown to pose an elevated risk of COVID exposure. For Marshallese poultry workers and other legally precarious immigrants, legal status also functions as a barrier to accessing safe labor conditions, unemployment and disability benefits, and healthcare.

Image 2: New protective barriers between Tyson Foods workers at the company's Springdale, Arkansas plant on April 24, 2020. Source: Tyson Foods/Taylor Borden and Associated Press

The vulnerability of immigrant food workers to COVID has a stark geography: meatpacking plants, food processing facilities, and farms have been outbreak hot-spots since the earliest days of the pandemic. While the Trump administration issued an executive order declaring meat and poultry processing as ‘critical infrastructure’ in April 2020, it issued no directive to grant workers access to protective equipment or to ensure social distancing in meatpacking plants. Industry leaders have been similarly cavalier about worker safety: Tyson managers in Iowa recently faced a lawsuit for placing bets on how many workers would die from COVID (a manager later said the wager was meant as a “morale boost”).

COVID exposure risk is also higher for workers in nursing and elder care, where Marshall Islanders and other immigrants are disproportionately represented in the workforce. These occupations put immigrant workers at far greater risk of COVID exposure than the general population, and self-isolating is a luxury many workers cannot afford. These risks are further exacerbated by weak labor protections: decades of union-busting in the region have led to limited family and medical leave, weak protections against dismissal, and major impediments to worker organizing. Thus, workers who are expected to quarantine must make the impossible choice between a paycheck and their own health and safety—very literally, given community mortality rates, choosing between their livelihood and their lives. In short, intimacy, contact, and exposure are unavoidable in the kinds of work often done by Marshall Islanders living in the US. In this case, intimate closeness to US empire has also meant closeness to the risk of exposure to COVID.

The effect of this intimacy is the very opposite of discreteness, despite potentially misleading impressions given by reports on ‘at-risk’ demographic groups: unsafe working and living conditions, not inherent or inherited physical traits (i.e. ‘preexisting conditions’), have determined COVID vulnerability for Black, Latinx, and Indigenous communities in the US, and the same is true for Pacific Islanders (Pulido, 2000). And yet, the notion of separateness persists: despite all we now know about community spread, rates of mask-wearing have been appallingly low across much of the US South, even in towns like Springdale where COVID has decimated entire neighborhoods. Resistance to masking belies the fact that people in communities are connected in very intimate, corporeal ways. As contact-tracing as a mapping project makes painfully clear, bordered demographic (e.g., racial) categories are ineffective containers of risk.

Race as Risk: Against a Eugenicist Public Health

In Northwest Arkansas, Marshall Islanders’ vulnerability to COVID is produced not by an unfortunate sequence of natural events, but by a one-two punch of labor precarity and health vulnerability, both exacerbated by class and non-citizenship. While Marshall Islanders’ legal status in the US provides some benefits unavailable to undocumented immigrants, it also carries significant restrictions, most notably ineligibility for Medicaid, a stipulation put into place by the Personal Responsibility and Work Opportunity Act (PRWORA) in 1994. While this stipulation was just overturned by Congress in December, the long-lasting effects of state abandonment persist: most Marshall Islanders in the US lack health insurance, and many cannot access basic healthcare.

Marshallese activists in the US consistently identify Medicaid ineligibility as one of the community’s biggest challenges, linking chronic health problems like diabetes and high blood pressure—conditions associated with higher rates of COVID complications and mortality, at least for some groups—to the tentacled, intergenerational effects of US nuclear testing (Duke, 2017; Hallgren et al., 2015). Marshall Islanders’ ‘preexisting’ health conditions, in other words, should be read as topographical markers of underlying imperial violence, experimentation, and exclusion, not as inherent. Marshall Islanders’ labor conditions and health care restrictions demonstrate that vulnerability is politically produced: it is a function of the simultaneous unwillingness of the US to protect its most vulnerable populations and its encouragement of their exploitation in the guise of work opportunities. And yet there persists the perception that group vulnerability to COVID exposure, illness, and mortality is an inherent quality of certain demographic groups—in this case, ‘racial’ groups—or at least that their at-risk-ness stems from a somehow similarly politically-neutral, matter-of-fact confluence of risk factors and preexisting conditions.

By now, bells should be going off: we know full well to be critical of logics naturalizing group-differentiated vulnerability. Vulnerability to premature death, as Ruth Wilson Gilmore (2007) has shown us, is the product of an organized, state-sanctioned or extralegal process—a product specifically intended to be exploited. Vulnerability to premature death, therefore, is not the result of demographic categorization, nor of one’s unfortunate origin in a naturally vulnerable place like an island: it is produced (or permitted) by the state as part of a racist and racializing project that seeks to differentiate and categorize. To differentiate, not to homogenize, in order to exploit.

Put more simply, COVID mortality rates are not just benignly given by demographic categories. As a social determinants of health framework shows, if we account for ‘race’—and related factors like immigration status—such variables are correlated with higher rates of COVID-related illness and death. This data does not prove that BIPOC[1] generally, or Pacific Islanders specifically, are inherently or genetically more likely to contract COVID, or to get sick and die from it. They are more exposed to the risk of COVID because of the state-sanctioned conditions of work and living. To attribute their vulnerability to their race or geographical origin is a eugenicist argument, as it attributes disparities to groups’ supposedly inherent qualities rather than associating them with the systems that created those disparities in the first place. A racial concept, and a spatial one. Eugenicist ideas connecting race, place, and empire should come as no surprise, given their origins: Sir Frances Galton, the founder of eugenics and noted geography enthusiast, was feverishly imagining the imperial possibilities of islands more than a century ago. Since then, the seduction of introducing (and then containing) risk to islands and islanders has remained strong.

No Island is an Island: Geographies of Risk and Vulnerability

US uses of islands, both historically and presently, create vulnerability for islanders and then naturalize that vulnerability by framing it as inherent to them. Buttressing these practices is the notion that islands are inherently remote and isolated, rather than connected. This notion is so pervasive that the term ‘island’ itself has become a metaphor for isolation. This idea has been embraced by everyone from artists like Gauguin and science fiction writers—think of H.G. Wells’ The Island of Dr. Moreau, or Jurassic Park—to authors like Jared Diamond (Correia, 2013) and military strategists testing bombs on the islands and atolls of Bikini, Tinian, and Vieques.

Image 3: US soldiers on a nuclear clean-up mission on Enewetak Atoll in the Marshall Islands, ca. 1977-1979Source: ABC News Australia, Mark Willacy (and Creative Commons).

This discourse is double-edged: it has been used to argue that islands’ isolated nature—their very ‘island-ness’—makes them both safer and more vulnerable. During the pandemic, we have seen a zealous resurgence of discourses portraying islands as safe spaces. Take the case of New Zealand, which has led the world in its commendable response to the pandemic, reducing COVID cases on the island to zero by this past June. And, as recently as September 2019, a startlingly prescient article in the journal Risk Analysis, entitled “The Prioritization of Island Nations as Refuges from Extreme Pandemics,” examined islands with “resilience-relevant” domains (Boyd and Wilson). Per these discourses, islands’ remoteness is the very reason for their security; safety from threat is a feature essential to the landform.

Conversely, imperial powers have long portrayed islands as vulnerable to external threats to justify their own military presence there. For example, the COFA’s ‘exclusive strategic denial’ clause safeguards the US military presence in the Pacific until 2086, revealing a Cold War ‘power vacuum’ logic that presents the US as protecting the Freely Associated States against the various external threats of Japan, the USSR, and, most recently, China. This past month, Alexander B. Gray, former Director for Oceania & Indo-Pacific Security at the White House National Security Council, called for the expansion of US influence in the Pacific in an article entitled, paradoxically, “How the US Can Protect the Sovereignty of the Smallest Pacific Islands.” Gray’s concern about China’s “aggressive influence campaigns” in the region compelled him to extend a compassionate gaze to islands currently beyond the US military’s formal grasp: “Particularly vulnerable are Nauru, Tuvalu, and Kiribati, whose exceptionally small size and geographic isolation make them especially susceptible to outside coercion” (2021, emphasis added). These discourses of islands’ smallness, isolation, and susceptibility to coercion will be familiar to those who have lived under imperialism in the Pacific Islands or the Caribbean, two archipelagic regions with long histories of successive and ongoing occupation. In both regions, generations of imperial actors have weaponized the language of vulnerability to describe islands and then justify heavy-handed imperial policies on that basis. Again—and this is important—notions of islands as inherently safe from, or vulnerable to, threat are not contradictory: both are borne of a longstanding imperial desire to frame islands as remote and isolated, all while holding them close enough to touch.

Either way, the notion of islands as remote is patently absurd, at best relative. Islands are never remote to their inhabitants, and their remoteness is a function of available means of communication. Moreover, islands have never been completely isolated, much as their residents might at times wish them to be. Repeatedly, we see just how much island spaces and communities are deeply interconnected: people travel for purposes of migration, commerce, work, and tourism. Some of these routes and connections predate colonial occupations and some of them have been ‘ruts of social difference’ carved by colonialism. Islands are anything but disconnected orbs, floating in empty space. In his famous essay “Our Sea of Islands,” Epeli Hau’ofa (1994) excoriated Westerners for the colonialist underpinnings of this ideology, and yet we continue to see similar (mis)representations of islands, decades on.

Of course, for all the US’s efforts to present islands as remote and isolated, it has demonstrated a paired desire to keep them close and accessible. The US has an extensive history of using islands as sites to contain people with health risks. Very often, these same health risks have been held up as justifications for restrictions on immigration, asylum, and citizenship (Loyd et al., 2016; Lopez, 2015; Paik, 2013). The notion of risk is another double-edged sword that has been wielded in paradoxical ways: it has often been the most vulnerable, at-risk populations that are presented as posing a risk to the general US public, as potential contaminants to the body politic.

But risk is not an innate feature of certain climes or landforms, any more than it is inherent to certain phenotypes. Risk travels to occupied and colonized islands, carried there by US mainland residents (as tourists, military members, and venture capitalists) and the grooves carved by colonialism enable people’s movement to escape risk on islands, or to seek pleasure there. As a result, islands’ ability to self-isolate during a pandemic is hamstrung by external geopolitical and (neo)colonial interests. While island states like New Zealand and Taiwan have been lauded for their COVID responses, islands with colonial ties to the US like the Marshall Islands, Guam, and Puerto Rico are limited in the protectionist measures they can take against the virus, particularly when it arrives via the military (for example, in an outbreak aboard an aircraft carrier) or tourists from the US mainland. The Marshall Islands, for its part, is enormously reliant upon US Compact funds, which have comprised up to 55% of the RMI’s federal budget (Stege, 2004). These neocolonial structures inform Marshall Islands’ experience of the pandemic, and its ability to enact policy responses to it, revealing the ongoing presence of US empire in the Pacific.

The Perils of Imperial Intimacy

Image 4: US medical officer inspecting Marshallese man after US nuclear testing, c. 1950s. Source: Talking Stick TV and Seattle Community Media.

As these imperial histories and geographies show, there are no such thing as ‘strictly border cases’: they exist only in the imperial fantasies of militaries attempting to contain nuclear risk to the Marshall Islands, or in the minds of Arkansas’s immigration NIMBY-ists who hope to safely employ Marshall Islanders to cut their meat and care for their loved ones while the Marshallese are burying their own. Borders, after all, are not only spaces of separation; they are sites of intimate touch. The link between intimacy and exposure is a powerful one, especially in the context of colonial systems and colonial imaginations. Similarly, there is no such thing as an inherently vulnerable place: vulnerability is made by the slow-and-fast violence of imperialism, eugenicist policies, and organized abandonment. The ‘afterlives’ of empire continue to actively produce vulnerability among islanders, through new colonial disasters like COVID that reveal just how connected islands really are.

[1] Black, Indigenous, and People of Color


This essay was written with the generous insights and editorial support of Rich Nisa, Ben Rubin, Mazen Labban, the Society of Fellows workshop at Dartmouth College, and Jan M. Padios’s “Introduction to Asian American and Pacific Islander Studies” class at Williams College. To support the work of Marshallese-led organizations in Arkansas, please explore the Arkansas Coalition of Marshallese (ACOM) and the Marshallese Education Initiative (MEI), and the Marshallese COVID-19 Task Force, led by Dr. Sheldon Riklon. To learn more about the ongoing impacts of the COVID-19 pandemic on Asians, Asian Americans, Native Hawaiians, and Pacific Islanders in the US, visit the AAPI COVID-19 Project.


Bonilla Y (2020) The coloniality of disaster: Race, empire, and the temporal logics of emergency in Puerto Rico, USA. Political Geography 78: 102181.
Briggs L (2002) Reproducing Empire: Race, Sex, Science, and US Imperialism in Puerto Rico. Berkeley: University of California Press.
Correia D (2013) F**k Jared Diamond. Capitalism Nature Socialism 24(4): 1–6.
Duke MR (2017) Neocolonialism and health care access among Marshall Islanders in the United States. Medical Anthropology Quarterly 31(3): 422-439.
Gilmore RW (2002) Fatal couplings of power and difference: Notes on racism and geography. The professional geographer 54(1): 15-24.
Gilmore RW (2007) Golden Gulag. Berkeley: University of California Press.
Gilmore RW (1999) You have dislodged a boulder: Mothers and prisoners in the post Keynesian California landscape. Transforming Anthropology 8(1‐2): 12-38.
Hallgren EA, McElfish PA and Rubon-Chutaro J (2015) Barriers and opportunities: A community-based participatory research study of health beliefs related to diabetes in a US Marshallese community. The Diabetes Educator 41(1): 86-94.
Harvey D (1989) The Limits to Capital. Chicago: University of Chicago Press.
Hau'Ofa E (1994) Our Sea of Islands. The Contemporary Pacific 148-161.
Lopez PJ (2015) Clumsy beginnings: from ‘modernizing mission’ to humanitarianism in the US occupation of Haiti (1915–34). Environment and Planning A 47(11): 2240-2256.
Loyd JM, Mitchell-Eaton E and Mountz A (2016) The militarization of islands and migration: Tracing human mobility through US bases in the Caribbean and the Pacific. Political Geography 53: 65-75.
Nixon R (2011) Slow Violence and the Environmentalism of the Poor. Cambridge: Harvard University Press.
Paik AN (2013) Carceral quarantine at Guantánamo: Legacies of US imprisonment of Haitian refugees, 1991–1994. Radical History Review 115: 142-168.
Pulido L (2000) Rethinking environmental racism: White privilege and urban development in Southern California. Annals of the Association of American Geographers 90(1): 12-40.
Robinson CJ (2000) Black Marxism: The Making of the Black Radical Tradition. Chapel Hill and London: University of North Carolina Press.
Smith N (2006) There’s no such thing as a natural disaster. Social Science Research Council 11. Available at: https://items.ssrc.org/understanding-katrina/theres-no-such-thing-as-a-natural-disaster/
Stege KE (2004) Marshall islands. The Contemporary Pacific 16(1): 126-132.

Emily Mitchell-Eaton is a feminist political geographer whose work examines migration and geographic mobilities across US empire. She is currently a Visiting Assistant Professor of Women’s, Gender, and Sexuality Studies at Williams College, and will be joining Colgate University’s Geography Department as a Visiting Assistant Professor in Fall 2021.