In his article, “Quantifying Vulnerability: Humanitarian Datafication and the Neophilia of Integrated Power,” Malay Firoz interrogates the increasing datafication of humanitarian aid practices. Based on ethnographic fieldwork with refugees and aid workers in Lebanon and Jordan, he explores how the algorithmic quantification of vulnerability assessments impacts on refugees, aid workers, as well as on the institutional make-up of humanitarian aid. He critically argues that the use of black-boxed technologies not only heightens refugees' precarious position, but also wrestles power away from local organizations to more central levels in the aid sector. In this author interview, Firoz reflects on his fieldwork experience, anthropological outlooks on the future of humanitarianism, and how his research relates to discussions of data extractivism at large.

Interview

Steffen Hornemann: You have researched various aspects of aid and humanitarianism in the context of Syrian and Palestinian refugees in Jordan and Lebanon. Can you tell us what has brought you to this topic, and what has directed your attention towards the role of technology in this article?

Malay Firoz: My interest in humanitarian aid emerged, as often happens with ethnographic research, in a rather serendipitous fashion. When I commenced my doctoral dissertation research in 2013, I had originally intended to explore how Western non-governmental organizations’ (NGOs) programs in northern Iraq were predicated on legitimating imperial intervention and regime change as a necessary prelude to democratic governance. However, when the security situation in Iraq destabilized with the rise of the Islamic State and fieldwork became logistically difficult, I turned my attention instead to the prospects for Iraqi refugees in Jordan. By the time I started fieldwork in Amman in the summer of 2013, the Syrian refugee crisis had already burgeoned across the region, and I gradually became more interested in humanitarian policy discourse—in particular, the recent turn toward “resilience” as a guiding paradigm for humanitarian action. My book manuscript based on this research traces how resilience-based aid programs draw humanitarians into deeper entanglements with the developmental prerogatives of asylum states and amplify contradictions between the human rights of refugees and citizens as exclusive subjects of biopolitical concern. Looking back on this project, I often think that I did not choose my research topic, but rather, that my topic chose me, because I was present at the right time and place to grapple with its logics.

My attention to questions of humanitarian technology was equally unanticipated. I wanted to accompany humanitarian aid organizations on household visits to examine the challenges facing Syrian refugees in a particularly neglected governorate of Lebanon. I expected the assessments to be technical and formulaic and was primarily invested in analyzing how Lebanon’s regional inequalities shaped disparate refugee experiences. During the visits, however, I found the assessment process to be surprisingly subjective and impressionistic, which piqued my anthropological curiosity about the ways in which social understandings of vulnerability condition expert assessments. This is well-trodden terrain in humanitarian and development studies, but as I delved deeper into the topic, I realized that vulnerability assessments were part of a broader debate in the aid world about the promises and perils of technology in humanitarian action. The United Nation’s Syria Crisis Response had served as a watershed moment for experimenting with new modes of data collection and service delivery, from e-cards to biometrics. Yet, as with similar debates about innovation in other sectors, concerns around technology’s adequacy often displaced deeper questions about the way it invariably reconfigures—whether inadvertently or by design—social relations among the range of actors it connects, occluding critical engagement with the pernicious implications of data extractivism for refugee rights. It’s this story of what technology does to humanitarianism rather than what it does for humanitarianism, to paraphrase a piece I quote by Kristin Sandvik and colleagues (2014), that I wanted to explore in my article for Cultural Anthropology (CA).

SH: You mention “important convergences” of refugees with “consumers in a data marketplace or … rights-bearing citizens of a welfare regime” (350). Could you elaborate on how these groups experience datafication in similar ways, and in different ways, and what this can tell us about datafication and techno-utopian visions of societal change at large?

MF: Thank you for drawing attention to that sentence, as it was intended to offer an important qualification to my argument. When reviewing scholarship in the anthropology of data, I realized that the question of how “datafied” subjects are emplaced within data infrastructures is inseparable from their subjectification by institutions of governance. In one sense, there are strong resonances between how data is extracted from refugees by humanitarian organizations and from citizen-consumers by state agencies and Big Tech companies. Both categories of subjects experience datafication in similar ways as an exercise in opacity, disempowered from making informed choices about how and by whom their data is used. Both groups are conscripted into the logics of datafication by the demands of modern living, and unable to meaningfully withdraw consent or safeguard their privacy. However, the question of consent is further complicated in humanitarian contexts by the fact that the humanitarian mandate is laden with what I’ve elsewhere called a discourse of “moral exceptionalism,” which sanctifies humanitarian practices with an aura of life-saving urgency and suspends some of the critical questions we need to ask about the right to privacy and protection against data extractivism.

Take, for instance, the well-established public conversation about the dystopian nexus between state surveillance, social media, and monetization of online data. While Facebook, X (formerly known as Twitter), and the National Security Agency certainly appear as archetypal villains in any story about diminishing prospects for democracy, the enormous databanks maintained by United Nations (UN) agencies rarely merit the same degree of public debate. After all, humanitarians are neither ostensibly agents of state interests—although they are invariably co-opted by them—nor are they governed by the profit motive. As the false presumption therefore goes, better data simply leads to better aid distribution. At the same time, refugees are scarcely empowered to preserve their data rights when their eligibility to receive aid is often contingent upon their “voluntary” compliance with humanitarian data protocols. In my CA article, I illustrate this point with examples from the UN High Commissioner for Refugees (UNHCR) in Jordan and the World Food Programme (WFP) in Yemen, both of which have mandated biometric verification for their cash and food assistance programs. States are no different in the way they institute data extraction into their citizenry’s every encounter with public bureaucracy, but my point is that the discursive barriers to transparency and accountability are uniquely difficult to surmount in a humanitarian enterprise that operates through claims of moral largesse rather than political responsibility.

These analytical distinctions notwithstanding, there is a dialectical relation between refugees and citizen-consumers as political subjects. As I mention toward the end of the CA article, refugees also serve as canaries in the coal mine, revealing the forms of datafied power that can be recircuited and adapted for use with vulnerable populations in other settings. This is a pattern evident throughout colonial history, where the techniques and technologies of violence originally developed in peripheral zones of exception were eventually weaponized against the metropole’s own internal Others. The adoption by United States law enforcement of military strategies developed in imperial wars abroad is but one example of this. Humanitarian data may emerge in the near future as another. Furthermore, as the aid sector warms to business partnerships with the private sector, it is precisely Big Tech companies that invariably design and build the infrastructure undergirding humanitarian data. For example, UNHCR’s Profile Global Registration System (proGres), which revolutionized mobile refugee registration in 2003, was developed in partnership with Microsoft. When thinking about techno-utopian futurities, therefore, data extractivism is another instance in which displacement—understood in broader terms as the condition of dispossession, precaritization, and rightlessness endemic to colonial modernity—is becoming an increasingly generalized predicament rather than an exceptional status for vulnerable communities.

SH: In addition to centralizing power, it also seems like the integration of data systems increasingly fragments humanitarian practices. You write, for instance, about how Mahmoud collects data without much regard for what happens with those data. Can you say more about this fragmentation? I’m also wondering how it might contribute to the intransparency not only of the technologies in use, but also of the institutional networks they’re embedded in. As a corollary, alongside the ways in which refugees experience vulnerability scoring as algorithmic violence, how do international non-governmental (INGO) workers also lack insight and control over the algorithms that quantify vulnerability? How do they experience and navigate this “blackboxing” of the technologies they use in their work?

MF: This is such a great question! You’re right that fragmentation operates concurrently with the integration of data systems and the concomitant centralization of power in the aid sector. In one sense, humanitarian practices have been growing more fragmented since the professionalization of the industry in the 1980s, with increasingly specialized modes of service delivery, standardized evaluation metrics, and delegated regulatory agreements with third-party actors. Mahmoud’s sense of distance from the final outcome of the data he collects is therefore not all that different from any public-facing bureaucrat, or client-facing corporate employee, who can only intervene in one node of a much larger chain of processes they do not control. The more aid institutions come to resemble such a faceless bureaucratic machine, the more humanitarian labor is consequently fragmented.

The reason Mahmoud was so striking to me was not because his experience of fragmentation was unique, but because it contradicted the aura of moralism so often associated with humanitarian labor, and revealed instead a different and perhaps more sober disposition toward that labor. I’m not at all suggesting that moral callings played no role in Mahmoud’s professional choices, but rather, that the humanitarian workplace can equally involve other more prosaic experiences and motivations that are not adequately addressed by moralist framings. It’s also important to add that this disposition is not apposite to “local” aid workers alone. Indeed, a British aid practitioner I once spoke with laughed at the very idea of aid work as moral work. Most of us go into these jobs, he told me, because the pay is good, you get to live in a big house, and you even get a car with a driver. How many other jobs in this day and age can promise that? My point is that both laudatory and critical accounts of humanitarian action are often so overdetermined by the aid sector’s self-representation as a moral enterprise that the institutional throughlines between the humanitarian workplace and other forms of fragmented, alienated, and classed labor can often be missed.

That said, there is also a specific purchase to the way technological innovation fragments humanitarian labor. The problem here isn’t limited to the opacity of the technological apparatus alone but, as you correctly identify, with the institutional networks in which that apparatus is embedded. In particular, the adoption of proprietary technologies—such as the databases in which vulnerability data is stored, or the algorithms used to process it, or the user interfaces through which it can be accessed—function as gatekeeping mechanisms that not only reinforce the regulatory powers of the international agencies controlling these technologies, but also wrest agency away from partner organizations that are contractually bound to use these technologies. As a result, aid workers at smaller NGOs are deprived of the ability to adapt these technologies to their operating environments or render them more meaningfully accountable to refugee demands. This is the story I discuss in my CA article of Rania’s NGO, an implementing partner for WFP’s food voucher program in Lebanon. Such NGOs collected vulnerability data about refugees to determine their eligibility for the voucher program, but had no control over how WFP used the data or who eventually received assistance. Proprietary technologies of datafication thus reconfigure relations within the humanitarian system in ways that “blackbox” their functions and render them opaque to public scrutiny. My point is not that aid organizations should renounce technological innovation altogether, but that humanitarian principles need to govern both whom technology assists as well as how it assists them.

Aid workers responded to the blackboxing of humanitarian technology and the fragmentation of their labor in a variety of ways. A common refrain among my interlocutors was that the humanitarian sector was failing to meet the needs of its beneficiaries, but where precisely they placed the blame for this failure was often conditioned by their nationality, professional experience, and the mandate of the organizations they served. Many were deeply frustrated with the structural inequities between UN agencies, international NGOs, and local NGOs, which determined who held sovereign domain over policymaking and how accountable they were to their partners down the chain of delegation. Some decried these institutional arrangements for hindering their ability to meet their ethical responsibilities, while others sounded more resigned that they were simply making the best of a bad situation. In that sense, one aid worker’s description of her job as a mere “numbers game,” as I discuss in the CA article, invokes both the alienating proceduralism of humanitarian labor as well as the sector’s minimalist mantra that “something is better than nothing at all.” Returning to the institutional throughlines I alluded to earlier between humanitarianism and other forms of fragmented labor, my interlocutors’ responses were reminiscent of sentiments articulated by employees in countless other white-collar bureaucracies, from law to healthcare to my own field of higher education. How many of us, after all, haven’t at some point felt that our employers impede the very performance by which we’re measured? In my book manuscript, I treat these utterances as a veritable archive of humanitarian ethics, articulated not through the institution’s ostensible commitments but by its practitioners’ auto-critiques of the way those commitments are invariably betrayed.

Surya Pratap Deka: Writing about the lived experiences of refugees and the systemic limitations of humanitarian practices often involves deep ethical and emotional engagement. Could you share some of the ethical and emotional experiences, both highs and lows, that arose for you personally while working with and writing about refugees' lived realities and the limits of humanitarian practices?

MF: Thank you for that question. My experience conducting fieldwork in Jordan and Lebanon felt strangely fractured. On the one hand, my interviews with Syrian families often left me completely eviscerated as I grappled with the sheer hopelessness of their legal and economic predicaments, the vast gap between their needs and the support available, and their diminishing prospects for a just return to Syria. It remains to be seen whether the fall of the Assad regime in December 2024 might substantively alter that future. I also remember seething at the incongruency I witnessed between the inflated estimations of humanitarian giving in both academic and NGO circles and the grinding poverty in which most refugees lived. Moreover, while the aid sector’s focus was predominantly on Syrian refugees, other refugee groups such as Palestinians, Iraqis, and Sudanese did not fare any better, yet their lives were effectively invisibilized in policy conversations about what was owed to vulnerable communities in the region. I sometimes questioned whether my own research was guilty of reproducing the provincialism of expatriate “experts” by centering their policy formulations of resilience, rather than the forms of resilience practiced daily by refugees themselves. However, I reminded myself that studying powerful actors is instructive precisely because it reveals how governing institutions reproduce themselves, even when they seem to “fail” their own mandates.

On the other hand, if I can gesture to a countervailing “high” during my research, it would undoubtedly be associated with the more personal register of living and learning in the Middle East. As an Indian national, my access to Jordan and Lebanon was severely restricted by the limitations of my passport, such that the access I did secure always felt like a precarious privilege afforded through the selective grace of others. The cognitive dissonance of this experience forced me to reckon—in sometimes painful ways—with my own emplacement in the apartheid logics undergirding the international passport regime. At the same time, this emplacement also offered me a generative opportunity to build other concepts of ethnographic solidarity than the ones I had been trained to anticipate. Over the course of twenty months in the field, I stumbled upon an unlikely translation with my Jordanian, Lebanese, Syrian, and Palestinian interlocutors, founded not on an expressly defined politics shared across cultural distance, but on a more intuitively embodied cohabitation of the contradictions of postcolonial nationhood, conditioned by its experiences of racialization, its sensibilities of immobility, and the mutual reckonings produced by that inheritance. These translations would often appear outside the frame of formal interviews, in the stories we shared of being interpellated by a Western gaze, or the jokes we traded about the absurdities of our own contexts. To have understood this inheritance, to have grown conscious of its intimate and daily workings, was perhaps the most profound takeaway from my research and has shaped all my writing since.

SPD: Given the challenges you highlight regarding the potential for increased control in humanitarian contexts, do you remain optimistic that we can “bring the oligarchic electronic atmosphere under democratic control” (Duffield, 2019), especially ‘in mass displacement contexts involving non-citizens and stateless persons’ (366)? What would a hopeful path forward look like in this regard?

SH: And what role, if any, do you see for anthropology in finding answers to the harm that datafication does in particular contexts?

MF: I take the charge of anthropology as bearing witness to the world, as it unfolds in all its narratological incoherence, through a liberationist lens grounded in the discipline’s methodological commitments. Anthropological skepticism toward technocratic expertise and quantitative metrics, while well-trodden terrain in the discipline, remains relevant to public conversations in an era when techno-utopian fantasies abound and multiply. At a pragmatic level, it is possible to make limited but meaningful interventions in the extractive relations of humanitarian data, from promoting a refugee right to data, to cultivating greater caution with proprietary algorithms and private-sector partnerships. Imagine a world, for instance, in which aid organizations are intentional about employing open-source tools whose security protocols are accessible to public inspection. Imagine a world in which the choice to “innovate” is not simply the means to a purported public good, but itself accountable to the public good. At the heart of these steps is an effort not to reject the promissory note of technology, but to subject it to greater democratic visibility and accountability. These are already ongoing conversations within the aid world, and while there are ultimately no humanitarian solutions to humanitarian problems writ large, such conversations can nevertheless shape the professional culture around datafication.

That said, I must confess that I don’t feel very optimistic about the prospects for refugee rights in general. As climate change intensifies with each passing year and ecological catastrophe appears increasingly unavoidable, we can already anticipate that the rates of global displacement will also accelerate. Ironically, the popularity of resilience discourse in international policymaking is arguably itself a manifestation of growing institutional acceptance that climate-induced crises are the new normal to which the world’s vulnerable populations must adapt themselves. Liberal democratic governments have proven themselves incapable and unwilling to tackle these problems, but with the global resurgence of right-wing neo-fascist populism, it is increasingly difficult to imagine a future where the securitized and carceral logics of our present politics are not further entrenched, with their attendant implications for the uses and abuses of technology. That is why I describe my genre of ethnographic writing in my book manuscript as dwelling in the “tragic mode,” where tragedy connotes what Raymond Williams (1979, 209) describes as “the loss of hope; the slowly settling loss of any acceptable future.” Making space to grieve about the world that is unfolding, I find, is just as important as finding the courage to confront it.