When Numbers Lie: The Hidden Costs of UNHCR’s Humanitarian Finances
Guest Post by Dr. Ceren Yüksel
When Numbers Lie: The Hidden Costs of UNHCR's Humanitarian Finances
This article examines data bias in UNHCR's humanitarian finances for international refugee protection, prioritising image management for donor confidence and public support over truthful evaluations of systemic flaws and operational inefficiencies.
Defining Data Orders and Classifications in UNHCR
UNHCR's 'care' work is legal. It legalises individuals as refugees (and asylum seekers) who would otherwise be categorised as illegal immigrants, in return for funding from donor states and the public to funnel them into various UNHCR protection and assistance services when national governments delegate their responsibility to the organisation to administer the UN Refugee Convention. Individuals in this pseudo-citizenship with UNHCR become 'data subjects', with their personal details and experiences in their entire offline asylum journey to be recorded in the organisation's population database, ProGres v4, and other databases.
As a former UNHCR professional, having served in various data-related capacities from 2005 to 2021, including regional and global roles, I will focus on data's two key roles in guiding the organisation's humanitarian finance for refugee protection. First, data inform the planning, implementation, monitoring, and coordination of UNHCR's humanitarian action. By and large, the Country Operations Plan (COP) is essentially the key administrative process whereby UNHCR country offices collect detailed records of the types and impact of services provided to refugees based on the expenditures. These services may be directly implemented, facilitated through commercial contracts with vendors, or carried out via partnerships with agencies funded by UNHCR. Data collected at the local level are then harnessed at the global level to present the fiscal needs to the donors to close the gap between the intent of the refugee law and the experiences of those being quantified in data (the rights-based indicators of the law) for UNHCR's multi-billion-dollar yearly budget. Second, at the tail-end of this data production cycle, 'data subjects' are aggregated into online 'subjects of data' as part of widely circulated digital media content for humanitarian communication with one specific goal: demonstrating value for money.
As a case in point, appeal for donations is updated regularly or immediately following a significant crisis on UNHCR's global website. Using a standardised template, datasets and images change, presenting a focused, consistent perspective to varied audiences (governments, donors, scholars, and the broader public) that the 'humanitarian action' has either already been provided or is imminent through UNHCR's legal, moral, and humanitarian duty to know, care for, and represent the best interests of refugees for their protection. For example, "pay 50 GBP so UNHCR can provide thermal blankets and sleeping mats to a family of refugees." Simply put, data on the appeal, expressed in precise mathematical terms, provide evidence on circumstances of the person(s) depicted in the picture and what UNHCR does to alleviate their suffering.
The partial and purposive role of data as evidence emerges from UNHCR's internal data quantification process, which is not a neutral act. Rather, it is moulded by decisions of professionals who mediate the classifications, institutional interests, and the power dynamics of the global humanitarian sector with subjective judgments and biases embedded in data's three core elements: 1) task processes (operating procedures related to international refugee law); 2) systems (where data are stored, managed, and accessed); and 3) people (end users in and people served by UNHCR).
Processes: (Un) Seen and (Un) Counted in Refugee Protection
The decision making process of categorising a person as a numerical value worthy of international protection under the refugee law across multiple hierarchies with varied practices, responsibilities, and roles, from determining their status to compiling aggregate figures for resource mobilisation, is complex. Scholars have long raised concerns about UNHCR's accountability deficit within this complex network of distributed responsibility, examining the applicability of global administrative law to the organisation's decisions and acts, the potential for setting up an external review mechanism, and the possibility for UNHCR to scale down its humanitarian work to focus on supervising governments' compliance with the Refugee Convention under its core protection mandate. While this paradigm shift is yet to materialise with the ephemeral pushes from scholars, advocates, or sporadic media exposures of UNHCR-related scandals, the organisation's accountability towards refugees is entirely assessed through the optics of its COP finances, with the Global Trends Report and the Global Report – along with their annexed figures and interactive websites – offering an aggregate executive summary that frames the international protection agenda around the financial resources to be allocated to UNHCR.
What UNHCR produces for this de facto external review mechanism also presents an internal opportunity to take stock of its own and host governments' failures for potential course correction. But the potential pitfalls of an organisation assessing and demonstrating the impact of its own work by producing its own evidence is a culture of success at all costs with conflicting interests. The priority shifts from admitting and recording what needs to be corrected to correcting the data through layers of approvals and protocols to meet targets and metrics that look good on paper. In the absence of enforcement authority to hold states accountable for refugee law compliance, UNHCR's moral authority to encourage governments depends on how well it can mitigate its fear that donor states could withdraw funding and host countries could withdraw hospitality, halting its operations.i This misplaced accountability leads to quantitative oversimplifications that serve image management, producing fragmented and selective knowledge of refugees' misaligned circumstances with prescribed outcomes of refugee law, and a failure to record or take responsibility for UNHCR's direct or indirect acts of violations or omissions that cause them.
Systems: Collating Refugee Protection Data to Rally International Fiscal Support
Refugee law forms the basis for the data that UNHCR must record in its 'systems' to manage its activities and demonstrate evidence to prove their impact with accuracy, reliability, and validity. While the COVID pandemic intensified the use of digital media technologies in the delivery of humanitarian aid, UNHCR was already under pressure from donors, driven by domestic demands for a more efficient aid system, to show effective use of funds amid record-breaking displacement. The Grand Bargain during the 2016 World Humanitarian Summit (WHS) kick-started a craze over artificial intelligence, applications, interactive dashboards, and platforms featuring rankings, indexes, and metrics. The bargain for UNHCR was more flexible funding in exchange for better and more data.
Post-WHS, UNHCR made strides to bolster its data capabilities, establishing partnerships with major tech companies, taking part in high-level global events, and eventually declaring the Data Transformation Strategy in 2019 with a commitment to position itself as a leader in refugee data by 2025. But internally, UNHCR had been grappling with what is colloquially known as the 'data problem', a fragmented and non-interoperable data landscape, with costly projects that did not address or solve the root causes for quite some time. Technological advancements offered new data management possibilities, but without a unified data strategy, UNHCR had failed to retire or upgrade systems properly. Outdated systems coexisted with new ones, causing compatibility issues and data silos with unnecessary expenditures. When I left the organisation in mid-2021, UNHCR had identified 35 global databases, and countless others at regional and country levels. Each database used different formats, terminologies, and quality standards, resulting in conflicting statistics and mismatched taxonomies.
Results-based management (RBM) links an organisation's budget to data-driven performance results. Flawed input data ('garbage in') leads to unreliable outputs ('garbage out') for performance analysis. UNHCR's latest RBM platform, COMPASS, introduced in April 2021, does not address the 'data problem'. It improves the data outputs published across various interactive web interfaces without any contextual information on, e.g., data sources, methodologies, system documentation, data provenance, quality control measures, change logs, and version control. Some sensitive or confidential data are intentionally left out following a robust anonymisation and de-identification process to avoid the potential for harmful profiling of refugee communities. However, the absence of such public data is different from the missing datasets due to unsystematic or no record-keeping of operational inefficiencies and failures. In other words, certain data and how they are produced should be available to facilitate engagement or scrutiny from external parties, but they are not. The core issue of this data gap is not one of technical feasibility. Altering data infrastructures and unifying UNHCR's data for more inclusive and transparent applications is within reach. The crucial missing component is the political will to reform data infrastructures, allowing for issues to be viewed and indexed from a perspective that is different from what is being documented and authorised by UNHCR for external sharing. The absence of publicly available information on UNHCR's 'data problem' is a testament to the current state of this political will.
People: The Witness Behind the Data
Even if UNHCR resolved all the technical challenges since my 2021 departure, a crucial setback remains: officials' biased intent to convert refugees' real-life experiences into data that showcase their interventions as life-saving.
Nearly two decades ago, Harrell-Bond situated the scripted witnessing of humanitarian professionals as "microcosm for the ill-starred relations between (western) humanitarian 'charity' and its target populations," at the very centre of which the power of the 'giver' and 'receiver' sits at a macro level.ii In simple words, it would be exceptional for a UNHCR professional to admit and denounce malpractices to an external party with a desire to protect the funding at all costs. Since Harrell-Bond's influential critique, which demanded a re-evaluation of UNHCR's operations for its humanitarian work to be humane, the organisation has amassed more technological affordances to proliferate its official narratives. The new technologies have not replaced or altered these longstanding issues of power and control; they operate within them. The question is not whether UNHCR professionals lack dedication, integrity, or ethical standards in this passive witnessing, but rather that these practices are imbued with power dynamics of an untouchable mindset in a bureaucratic system that blurs the distinction between their objective witnessing of politically incorrect events and what UNHCR communicates subjectively in public.iii
In the absence of accessible and effective ways for refugees to seek remedial measures for direct or indirect acts of violations or omissions by UNHCR, the failure to admit what does not work by those involved in the process of data quantification creates an information gap that obscures knowledge about what necessitates corrective action.iv This gap inhibits the potential for change that people under UNHCR's care never experience with any meaningful reform. Each data set is an outcome of a digital loop that ends where it starts with a narrative on how well UNHCR does what it does through the same historical layers of silencing. Inevitably, inconsistencies emerge between the lived realities of forcibly displaced persons in the offline physical world and their digitised representations from local to global.
Looking Forward…
The crux of the problem is not lack of better and more data for sufficient funding. It lies instead in the absence of effective strategies and policies that take into account the unadmitted. The unadmitted is not the financial but the actual human cost in the implementation of international refugee law around the massive amount of funds and trust placed on UNHCR. Many digital media technologies are likely to emerge fast beyond the writing of this article, mediating what is fiction and what is real through the abstraction that happens within this expanding loop of UNHCR's digital computing affordances. Real change will have to begin from within UNHCR with an honest and transparent dialogue that breaks the historical amnesia across all hierarchies about actions that are unadmitted, unspoken, or not immediately evident in data, rather than fixing the data to maintain a consistent narrative that outperforms what needs to change.
While the wait for a paradigm shift towards a more accountable refugee protection continues, as a former UNHCR employee, I join the ranks of scholars and advocates calling for systemised efforts external to and independent of UNHCR to address the organisation's accountability deficit through the design of a digital archival prototype: Refugee Data Minder. The archive organises elements of inconsistency as manifestations of information gaps relating to rights violations informed by anonymised testimonies of UNHCR staff, academic literature, and advocates alongside UNHCR's data. It proposes an end-to-end external communication network to pose critical questions about systemic injustices and underlying structures of unevenness in the protection of refugees lost in this gap.
The pace at which new technologies become outdated at UNHCR is dizzying, exceeding the financial, technical, and non-technical capabilities that a single scholar or advocate can manage alone without external collective momentum. The true merit of the archive does not lie in its immediate outcome, but rather in the critical perspective it encourages, calling for a new mode of thinking through data towards a more inclusive, transparent, accountable, and equitable humanitarian landscape. Moving forward, while no specific tool can quantify every inconsistency in UNHCR's data, progress is achievable through a gradual, layer-by-layer approach, just like building a house. To realise its full potential, future layers need to be built collaboratively with scholars, practitioners, and advocates from diverse fields. Each new piece of data would add another 'brick' to the structure as an essential step towards thinking about what else data could do and who else data could serve beyond UNHCR's mediated representations.
Dr. Ceren Yüksel | System Architect/Founder | Refugee Data Minder | contact@refugeedataminder.com | http://linkedin.com/in/yukselceren