Columbia Daily reports of hacking and contentious debates about privacy or the lack of it are now commonplace. Reports are already on the front page of the papers about the fact that Facebook is even now alerting the FBI that certain groups have been formed to try to manipulate voters before the 2018 midterm elections in the USA, and they are likely being fomented by Russian governmental actors. Concerns about the violence fanned by social media applications like Facebook’s WhatsApp in India, Indonesia and elsewhere have forced the fumbling Facebook team to limit “shares” to no more than twenty at a time, as if that’s a solution. Nonetheless, the scariest thing I have read recently was a book called Automating Inequality documenting the “digital poorhouse” by Virginia Eubanks.
Mentioning this book to some of the tech team of Action Network, the great social action and mobilization tools and applications organization before their retreat on the eve of the Netroots gathering in New Orleans, where I was asked to make some opening remarks, they at first thought I was talking about the digital divide which is standard issue for lower income families. “Internet for All,” as ACORN’s campaign is called in Canada or our $10 per month fights with Comcast and others in the USA are certainly focused on that issue, but Eubanks’ “digital poorhouse” is something much, much worse.
Here we are talking about digital tools, artificial intelligence, and mindless algorithms in many cases being weaponized against lower income and working families in precarious situations. Proving her proposition, Eubanks meticulously dissects several case studies.
The first was the effort by Indiana’s Republican Governor Mitch Daniels to make eligibility determinations for social service programs in that state, whether food stamps, welfare, or Medicare, through a more than half-billion contract with IBM. The intention was to make the process virtually mechanical, cutting down the labor costs and discounting the experience of social service professions in the state system and by computerizing the process reduce beneficiaries. The results several years later were such a horror that even Daniels was finally forced to concede that it was an unholy mess ending up in contentious litigation between Indiana and IBM. Tragically, even as the experiment was a disaster, it succeeded in both reducing the state payroll and, even worse, dramatically reducing the number of beneficiaries of public support programs in the state. Paperwork was lost. The process was complicated. Appeals grew exponentially as desperate families tried to keep from being arbitrarily rejected and then struggled to get recertified. People were harmed. People died. Families when hungry.
In another case, Eubanks looked at the problems in the Pittsburgh/Allegany County system of child and youth services. Even with perhaps better intentions, data was weaponized and the error rate was extreme. The algorithms assigned worse risks than common sense would ever have determined. Children were separated from families. Mothers were shamed.
Eubanks was not waving the Luddite’s banner. She was cautioning that these tech hammers are pretending that all situations – meaning people – are the same nails, and the impact on lower income families is not only extreme, but given the way data works, permanent. The “digital poorhouse” is caging a class of people and in her works, “automating inequality” by using all of these tools and weapons against the poor.
***
Thanks to KABF.