Inclusion by Design? The Janus Face of Match’In 2.0: Family Complexity, Unequal Vulnerabilities, and Accountability

Digital tools increasingly shape asylum and migration governance, among others in the allocation of refugees to local communities. In this area of application, algorithmic matching systems promise efficiency, transparency, and improved integration outcomes, yet they also risk reproducing existing inequalities and generating new forms of exclusion. This paper examines these tensions through the case of Match’In, an algorithm-based decision-support tool developed in Germany to complement quota-based refugee allocation by systematically incorporating integration-relevant needs and preferences. Empirically, the paper relies on findings from Match’In 1.0 (2021–2025), a pilot project implemented in cooperation with several German federal states and municipalities. Drawing upon substantial practical experience and evaluation insights, Match’In 1.0 offers an illustrative example of how algorithmic recommendations can enhance the accuracy of allocation decisions while remaining firmly embedded within legal frameworks and human discretion. Concurrently, the pilot phase has exposed structural constraints and normative trade-offs intrinsic to the implementation of digital tools within administrative systems, particularly with regard to participation, transparency, and accountability. Conceptually, the paper uses these insights to frame Match’In 2.0 as a “janus-faced” development process. While the next iteration seeks to build on previous lessons, it foregrounds a central challenge for algorithmic matching: the increasing complexity of the target group. The expansion of the analytical lens from the individual to the family unit brings to light multiple and intersecting forms of inequality that are challenging to operationalise within the confines of algorithmic systems and to implement as part of the existing allocation system. These include age-specific needs across different life stages, intra-family power relations affecting women and children, differentiated vulnerabilities related to health, disability, and care responsibilities, as well as inequalities linked to access, comprehension, and expression within digitalised procedures. The paper's approach to complexity departs from treating it as a mere technical obstacle to be resolved through optimisation. Instead, it conceptualises complexity as a normative and political issue that shapes processes of inclusion and exclusion in the allocation of asylum. Methodologically, it reflects on the limits of participation and informed consent under restrictive institutional conditions and discusses how participatory and reflexive approaches can nevertheless inform the design and evaluation of digital governance tools. The present paper contributes to broader debates on algorithmic governance, inequality, and accountability in migration policy by combining empirical insights from Match’In 1.0 with a conceptually open and forward-looking discussion of Match’In 2.0. It argues for a cautious and reflexive understanding of “inclusion by design” that recognises family complexity, differentiated vulnerabilities, and power relations as constitutive—rather than residual—elements of algorithmic asylum governance.