Global Issues
Algorithms Are Quietly Erasing Disability Rights -By Fransiscus Nanga Roka
Courts should be unequivocal. All algorithms should meet minimum legal safeguards for the use of algorithms in public benefits decisions, transparency, auditability, and the ability to challenge them in basic but meaningful ways. Anything less would violate both due process and the equal protection clause. In addition, disability rights jurisprudence must change to challenge a new form of exclusion that works not through discrimination per se but statistical abstraction.
What a lawsuit publicly revealed in Nebraska last May was that the true threat, far worse than a bureaucratic error, had been allowed to silently take root, that legal judgment had been supplanted by algorithmic authority. Dane County presented this disability score in the case of Bessey v. Nebraska Department of Health and Human Services, essentially only stripping 40% away state funded home care from a man with developmental disabilities not due to any medical reassessment or hearing but an arbitrary computational score.
The interRAI assessment system in which Nebraskas employs is a common but troubling governmental temptation: to turn human need into data, and then treat the data as somehow excluding or neutral truth. However, algorithms, are not neutral arbitrators. You are software and these are policy choices: what matters, what gets erased from the record, whose suffering is easy to see.
The stakes for Nicholas Bessey are existential. Minimizing the home based care work is not a mere discomfort in design but is a structural incitement to institutionalization. The Americans with Disabilities Act (ADA) is explicit, especially under the Olmstead doctrine, that people with disabilities have a right to be in the least restrictive environment. And yet here, an algorithm quietly nullifies that right, no explanation, no justification and no recourse.
At its constitutional core, due process requires that individuals know the rules applied to them and are afforded a right of appeal from decisions impacting their rights. But how do you cross-examine an algorithm? How is it possible to appeal a score produced by proprietary logic, undisclosed weighting and hidden assumptions? More than automating eligibility, Nebraska built a buffer between decision-making and accountability.
The state may argue efficiency. Perhaps consistency, scale, or cost management. Efficiency is not an excuse for opacity when it comes to fundamental rights. In so far, even the logic behind economy is doubtful itself. The system is effectively pushing people into institutional care, which is generally much more expensive and a little shown to be less restrictive by making home-based services less accessible. There is indeed a major fallacy in the big-slice thesis, since prudence may be an illegitimate and economically selfish position.
More troubling is the precedent. But of course if algorithmic systems can reframe need as disability in Medicaid, what will stop them from expanding into other contexts — child welfare, criminal prosecution or sentencing, public housing? It’s all patterned – prediction machines promise objectivity, but mirror systemic prejudice while dodging conventional jail protections.
Bessey, at its heart, is not just the story of a single man and a single state. This makes for a constitutional stress test for the digital welfare state. If we delegate decisions to systems that cannot explain themselves, and over which no human being is ultimately responsible, can civil rights even survive?
Courts should be unequivocal. All algorithms should meet minimum legal safeguards for the use of algorithms in public benefits decisions, transparency, auditability, and the ability to challenge them in basic but meaningful ways. Anything less would violate both due process and the equal protection clause. In addition, disability rights jurisprudence must change to challenge a new form of exclusion that works not through discrimination per se but statistical abstraction.
This is a larger moral breakdown. In this model, by using automated scoring in matters of care over human judgement, governments may be getting closer to redefining dignity from a principle into something that can fluctuate. No longer, does it matter whether the state helps or not; it only matters, if the state acknowledges the humanity of those who are governed.
Nebraska does not simply compute need with its algorithm. It determines whose life is worth living in the community and whose is not.
A decision no machine should ever be equipped to make.
Fransiscus Nanga Roka
Faculty of Law University 17 August 1945 Surabaya and Managing Partner Law Firm Victorious Indonesia
