Abstract
Over the last two decades, there has been a growing use of predictive technologies to determine who is allowed at the border from visitors to immigrants to asylum seekers. With their credulous pledge to eliminate irregular entries at the border, these novel automated systems appeal to state and non-state actors who justify their use in the name of national security or efficient management of borders.
This paper examines in what ways claims to predictable borders via automated technologies are crafted in bureaucratic institutions. We focus on two projects by the UN Refugee Agency (UNHCR) that use machine learning and data science techniques to examine how the UNHCR claims and performs predictability at the border.
We draw on Louise Amoore’s concept of “deep border,” which is the result of deep neural network algorithms to not only grasp “representations from data” but also to “generate meaning” from varied forms of digital information (2021:2). We ask: What does the micro-work of building deep borders through predictive analytics look like at the UNHCR? How do officials at the Agency reconcile the risks of these technologies with the benefits they claim to receive from predictability? In what ways does the “deep border” crafted by the UNHCR deepen, if at all, their humanitarian work?
Drawing on a content analysis of policy documents and interviews with key actors involved in these projects, we argue that the so-called success of a predictable border does not solely derive from its technical capacity to accurately predict numbers, but also from creating a semblance of a predictable border within the organization.
We identify three practices of UNHCR officials that help justify and maintain the claim that a predictable border is possible: constantly seeking novel variables and data, continually maintaining opacity, and quickly shifting models to adapt to changing circumstances. These socio-cultural practices constitute the micro-work of building a predictable border inside the Agency even when the technical reality of predictive analytics does not live up to bureaucratic imaginaries.
Making up the Predictable Border
Burcu Baykurt & Alphoncina Lyamuya, University of Massachusetts Amherst, US
About the event series
“STS-MIGTEC circle” is a small format which takes place once a month and which serves to reflect jointly on work-in-progress contributions related to the themes of interest to STS MIGTEC. The idea is to create a safe space for probing and experimenting with ideas, arguments, attempts of analysis, sense-making of empirical material.