Against Reverence

Some of the most groundbreaking innovation in digital identity research lies in the work of early career scholars, who develop key theoretical concepts to navigate the field. In this post I expand on how work conducted by early career colleagues, met before and during this summer of conferences, has deeply informed my understanding of digital ID.

I was raised in an academic field, Information Systems, where “standing on the shoulders of giants” is a not-so-silent, overarching precept. For as much as it can be questioned, the precept informs the field: senior editor panels at conferences see packed rooms of early-career colleagues, striving to find a space to grasp wisdom from seniors and even, in the wildest dreams, bag the chance to ask a question to the “big ones”. Innovative, thoughtful ideas are brutally bashed for insufficient reference to “established” work, with the more or less (often less) kind invitation to return to the drawing table. The idea of seniority as reverence has affected the field so much that, in my observation, we even refrain from questioning it anymore. When bullied by a senior scholar in front of a whole academic seminar audience, simply for daring to ask a question, 28-year-old me cried all the way through the bus route to my north London sublet, thinking that what happened was just right. If the senior implied that my question was dumb, so that definitely was. And I deserved being laughed at by the whole seminar room, a laugh I won’t forget, in fact I can still hear it.

What I didn’t know back then is how wrong the colleague was. And not the person specifically, but the established, deeply ingrained idea that academic knowledge, to be accepted, should come from all-established “seniors”. A recurring thought after coming back from conferences, where the packed-room-for-seniors-panel situation is a regular (and a whole post should be written on the demographics of such panels, and the hierarchies it reinforces), last year I wrote a Medium post on the topic. I reiterate it this year, at the dawn of one of the largest convenings in my field: scientific knowledge comes from mutual learning, rather than from the blind reverence that the field ingrains in us. And as I work on finishing my first book, Unfair ID, I am beyond blessed with learning from colleagues that, having come recently to the field, are bringing the most important innovations in it. Below are just some of my notes from recent, massively enriching conference encounters.

As someone who started studing digital ID during her PhD, understanding biometric tech markets has always been a key challenge, and one that I’ve never completely gotten around. The digital identity solutions market is forecast to grow from US$23.3 billion in 2021 to US$49.5 billion by 2026: and at the same time, technologies of biometric identification are diffusing rapidly in state and humanitarian domains, where they purport to provide the type of technology required by development objectives, epitomised by SDG 16:9 on legal identity. A report on the status of biometric humanitarianism released just this month is clear on the matter: private tech, on which large humanitarian schemes are already predicated, is becoming even more key to ensure the authentication-authorisation nexus on which targeted schemes are based. Two years ago a blog post by Aaron Martin and Linnet Taylor, written just after the Identity Week in London, noted how private players’ narratives explicitly depicted human development as a business opportunity, leaving a big question mark on the interests of the protected.

It was in my effort to get around the logics of biometric markets that I met Carolina Polito, whose work with Cristina Alaimo was presented at the European Conference on Information Systems in Kristiansand. Their paper “The Politics of Biometric Technologies: Border Control and the Making of Data Citizens in Africa” uses the notion of “politics of the biometric artefact” to understand the role of private contractors in the establishment of border control markets. Developing the idea of “Do artifacts have politics?” written by Langdon Winner in 1980, the work presented by Polito moves significantly beyond it: the notion of “politics of biometric artefacts” is deployed to understand the complex enmeshment of border control with economic profit, inviting questions on “who benefits” from the marketed deployment of advanced tech recognition in borders. Another recently released report, “Artificial Intelligence: The New Frontier of the EU Border Externalisation Strategy”, explores the role of technologies marketed as AI for border externalisation, a widely used strategy to stop migration in EU member states.

More knowledge is developed on topics that, such as the effects of biometrics on social protection programmes, have long been established in the field. A widely research digital ID platform, India’s Aadhaar as incorporated in the country’s largest social protection programmes, has seen debate among researchers on the extent and measurability – rather than actual presence – of exclusions associated to biometric recognition. Limited work exists, instead, on the Aadhaar-Enabled Payment System (AePS), an infrastructure that allows using Aadhaar as ID verification to access one’s bank account and perform key operations. Placed in the domain of cash transactions, work on AePS is yet to discuss the presence and depth of exclusions associated to it, and the infrastructural and societal drivers associated to more or less successful use of the service.

Ration shop, Kolar district, Karnataka, April 2018

Especially, limited research has delved into the transaction failures that have affected AePS. It is here that Malavika Raghavan, currently a PhD candidate at the LSE, has published detailed insights on the levels at which AePS transaction failures occur: her work notes three levels – consumer/BC, infrastructure, and bank – at which failure can take place, quantitatively estimating the majority of failures to take place at the infrastructural level. Noting that infrastructure-level transactions, which include biometric mismatches, are prominent, Raghavan sheds light on an important side of biometric architectures: the very material, impoverishing consequences that failures embedded in the technology may yield. Costanza-Chock’s (2020) important book on Design Justice comes to mind, reminding how injustice, rather than incidental, can be inscribed in the very body of the technology.

Still on the topic of biometrically assisted social protection, prosperity scoring – embedded, paradigmatically, in Colombia’s System of Identification of Social Program Beneficiaries (Sisbén), – has gained foothold as a route to designing targeted anti-poverty systems. Sisbén, a unified household vulnerability index built to identify poor households, is based on a continuum vulnerability score (0-100), assigned to each household in the light of socioeconomic variables. Calculation of the index is based on three components: a socioeconomic survey to collect data on households; a welfare measure to assess vulnerabilities; and software to perform household level calculations.

It was thanks to Joan Lopez, currently a PhD researcher at Tilburg University, that I encountered Sisbén, and the problematic evolution that the system known as Ingreso Solidario – Solidarity Income – has displayed with its establishment during the COVID-19 pandemic. The goal of the new subsidy programme was to identify subjects not served by other schemes: data from across government databases were hence crossed to identify needful households, resulting in opaque combinations of data from different repositories. But in the machine-led process of entitlement assignation, the combination of information to make entitlement-assigning decisions remained blackboxed to users, resulting in uncertainty on key aspects of a decision strongly marking the prosperity scoring of households and the effects resulting from it.

As the colleagues featured in these notes, a vast group of researchers I learn from every day are conducting doctoral or postdoctoral work. My privilege is to be in the position to remain constantly updated just by getting to hear their work, discuss and cite it. None of these scientific advantages is made possible by positions that silo research into seniority categories. In fact, richness lies in drawing on an amazing pool of colleagues from across domains, geographies and fields every day.

My happiness from having joined the adventure of Sociotechs, a new podcast launching in August with Tejas Kotha, stems exactly from the need of siloed barriers to be broken. In a series of episodes currently in production, we speak to colleagues from NGOs, PhD programmes, academic institutions, and other domains to achieve one overarching goal: learning from each other on topics pertaining to technology and society, looking straight at the bases of data-induced harm and routes to overcoming them. Only through mutual learning can this be achieved.

And to the senior academic from my opening story, I want to tell they were so wrong. That early career research can effectively produce not only important results, but those that drive the field and shape it into what it is. The field needs to do a lot more to acknowledge the importance of innovation brought by early career work, and begin making sense of the huge loss connected to not doing so.