Civic News
Tech + Government Month 2022

How does DC government use tech to make decisions?

The new "Screened & Scored in DC" report gives an overview of what agencies are using automated decision-making, and how it impacts residents.

If you simply live in the digital world, you’ve probably come across automated decision-making.

As described by DC’s Enid Zhou, automated decision-making (ADM) is the use of data-driven or rules-based approaches to either replace or inform decision-making. This could mean anything from a score sheet to something as complex as machine learning. But it shows up constantly in Netflix show recommendations, suggested groups to join on Facebook and recommended products on consumer sites.

DC government, as it turns out, is no exception to the tech. A new report from the watchdog Electronic Privacy Information Center (EPIC) examined how much the District of Columbia is outsourcing its government decisions to such choice-making systems. All around the city, the “Screened & Scored in the District of Columbia” report found, automated decision-making shows up and directs children on what schools to attend, informs medical decisions and influences policing protocol.

Zhou, senior counsel at EPIC, told Technical.ly that the org spent 14 months investigating these systems in DC as part of a larger project on the topic. In total, they found 29 systems from over 20 agencies thanks to Freedom of Information Act requests, publicly accessible information, academic research, news articles and other sources.

ADM systems, Zhou said, aren’t inherently good or bad when used by governments. Rather, the determination of ADM’s suitability for government work relies on how the systems are developed and used on residents.

A headshot of Enid Zhou

Enid Zhou. (Courtesy photo)

“Our position really is that people should be aware of how these systems work, in what context, who is using them and whether they’re developed by third parties or not, if that kind of relationship is set in place,” Zhou said.

While human work is also inherently biased, the report argues that ADM tools have their own limitations. Public agencies in DC use them to determine things like prioritizing applicants among the unhoused community for housing resources, as well as detecting unemployment fraud. But these uses can only go so far, and the report states that some agencies and tech companies even block audits of their ADM tools over claims that allowing the public to scrutinize them would hurt a competitive position or lead to harm.

“In fact, using ADM systems conceal structural inequities, providing an illusion of progress by replacing a biased individual with an equally biased algorithm,” the report states. “ADM systems learn what is normal by studying the world that exists, including its patterns of discrimination. These systems inevitably make predictions and recommendations that maintain the status quo.”

With the report, Zhou said, EPIC hopes that DC residents can see the critical need to improve oversights of automated systems. She said that during the research stage, many agencies weren’t even aware of what automated decision-making was or lacked a clear definition. For many, she realized, there is no clear, consistent definition; She hopes the report can change that and, in turn, how agencies and organizations use automated decision-making.

“We would really love for them to know that automated decision-making plays into their work and overlaps and bleeds in some way,” Zhou said. “So we’re just hoping to use this report to help them in their work — especially, especially when they’re helping other people.”

Read the full report

This editorial article is a part of Tech + Government Month 2022 of Technical.ly's editorial calendar. This month’s theme is underwritten by Spotify. This story was independently reported and not reviewed by Spotify before publication.

Companies: District of Columbia

Before you go...

Please consider supporting Technical.ly to keep our independent journalism strong. Unlike most business-focused media outlets, we don’t have a paywall. Instead, we count on your personal and organizational support.

Our services Preferred partners The journalism fund
Engagement

Join our growing Slack community

Join 5,000 tech professionals and entrepreneurs in our community Slack today!

Trending

Biotech, high chairs and dirty laundry dominate DC Startup and Tech Week’s pitch competition

Not all jobs are the same. Why do workforce agencies treat them like they are?

After nearly a decade, the federal program for immigrant entrepreneurs is finally working

Block the bots or feed them facts? How Technical.ly uses AI in journalism

Technically Media