Evidence is a policymaker’s biggest weapon

Curated on Posted on by Stefaan Verhulst

Report by Jacquelyn Zhang: “Fundamentally, public policy is supposed to address serious social problems. However, poorly designed policies exist. Often this happens when a well-intentioned policy generates unexpected and unintended consequences, and sometimes, these consequences leave policymakers farther away from their goal than when they started.

Consider just a few examples.

The first is the impact of an immigration law that was used in the United States ostensibly to control the flow of undocumented immigrants into the country. The controversial bill imposes extreme restrictions on undocumented immigrants in the state of Alabama and limits every aspect of immigrants’ lives.

By employing a synthetic control methodology, the bill proved to have a substantial and negative unintended effect – an increase in violent crimes. This could be linked back to the bill because while violent crime increased, property crime did not.

This may be because the passage of one of the country’s strictest anti-immigration laws signalled to the community that the system had more tolerance for discrimination against undocumented immigrants in Alabama, fuelling distrust and eventually violent conflict.

This is not a freak event either. Policymakers know that enacting laws doesn’t just change the wording of legislation. It shapes social norms, prescribes attitudes, and affects community behaviour. Of course, this is also why good policy-making can be so productive…(More)”.

Humans in the Loop

Curated on Posted on by Stefaan Verhulst

Paper by Rebecca Crootof, Margot E. Kaminski and W. Nicholson Price II: “From lethal drones to cancer diagnostics, complex and artificially intelligent algorithms are increasingly integrated into decisionmaking that affects human lives, raising challenging questions about the proper allocation of decisional authority between humans and machines. Regulators commonly respond to these concerns by putting a “human in the loop”: using law to require or encourage including an individual within an algorithmic decisionmaking process.

Drawing on our distinctive areas of expertise with algorithmic systems, we take a bird’s eye view to make three generalizable contributions to the discourse. First, contrary to the popular narrative, the law is already profoundly (and problematically) involved in governing algorithmic systems. Law may explicitly require or prohibit human involvement and law may indirectly encourage or discourage human involvement, all without regard to what we know about the strengths and weaknesses of human and algorithmic decisionmakers and the particular quirks of hybrid human-machine systems. Second, we identify “the MABA-MABA trap,” wherein regulators are tempted to address a panoply of concerns by “slapping a human in it” based on presumptions about what humans and algorithms are respectively better at doing, often without realizing that the new hybrid system needs its own distinct regulatory interventions. Instead, we suggest that regulators should focus on what they want the human to do—what role the human is meant to play—and design regulations to allow humans to play these roles successfully. Third, borrowing concepts from systems engineering and existing law regulating railroads, nuclear reactors, and medical devices, we highlight lessons for regulating humans in the loop as well as alternative means of regulating human-machine systems going forward….(More)”.

Policy Building Blocks, And How We Talk About The Law

Curated on Posted on by Stefaan Verhulst

Article by Cathy Gellis: “One of the fundamental difficulties in doing policy advocacy, including, and perhaps especially tech policy advocacy, is that we are not only speaking of technology, which can often seem inscrutable and scary to non-experts, but law, which itself is an intricate and often opaque system. This complicated nature of our legal system can present challenges, because policy involves an application of law to technology, and we can’t apply it well when we don’t understand how the law works. (It’s also hard to do well when we don’t understand how the technology works, either, but this post is about the law part so we’ll leave the issues with understanding technology aside for now.)

Even among lawyers, who should have some expertise in understanding the law, people can find themselves at different points along the learning curve in terms of understanding the intricacies and basic mechanics of our legal system. As explained before, law is often so complex that, even as practitioners, lawyers tend to become very specialized and may lose touch with some basic concepts if they do not often encounter them in the course of their careers.

Meanwhile it shouldn’t just be lawyers who understand law anyway. Certainly policymakers, charged with making the law, should have a solid understanding what they are working with. But regular people should too. After all, the point of a democracy is that the people get to decide what their laws should be (or at least be able to charge their representatives to make good ones on their behalf). And people can’t make good choices when they don’t understand how the choices they make fit into the system they are being made for.

Remember that none of these choices are being made in a vacuum; we do not find ourselves today with a completely blank canvas. Instead, we’ve all inherited a legal system that has chugged along for two centuries. We can, of course, choose to change any of it should we so require, but such an exercise would be best served by having a solid grasp on just what it is that we would be changing. Only with that insight can we be sure that any changes we might make would be needed, appropriate, and not themselves likely to cause even more problems than whatever we were trying to fix…(More)”.

Data Types, Data Doubts & Data Trusts

Curated on Posted on by Stefaan Verhulst

Paper by João Marinotti: “Data is not monolithic. Nonetheless, the word is frequently used indiscriminately, referring to a large number of different concepts. It may refer to information writ large, or specifically to personally identifiable information, discrete digital files, trade secrets, and even to sets of AI-generated content. Yet each of these types of “data” require different governance regimes in commerce, in life, and in law. Despite this diversity, the singular concept of data trusts is promulgated as a solution to our collective data governance problems. Data trusts—meant to cover all of these types of data—are said to promote personal privacy, increase corporate transparency, facilitate the sharing of data, and even pave the way for the next generation of artificial intelligence. These anticipated benefits, however, require the body and flexibility of equitable trust law and its inherent fiduciary relationships. Unfortunately, American trust law does not allow for the existence of such general data trusts. If anything, the judicial, academic, and legislative confusion regarding data rights—or its status as property—demonstrates that discussions of data trusts may be ignoring a key element. Without first determining whether (or what kind of) data can be recognized as a trust res (i.e., as trust property) under existing law, it may be premature to accept data trusts as the private law solution to our data governance ills. If, on the other hand, the implementation of data trusts requires legislative intervention, its purported benefits must be analyzed in contrast to the myriad other new and evolving data governance frameworks that would similarly require legislation. By analyzing existing trust law and the difficulties of defining data rights, this essay highlights the urgent need to pursue doctrinally, legislatively, and technologically viable data governance strategies….(More)”.

Data Literacy for the Public Sector: Lessons from Early Pioneers in the U.S.

Curated on Posted on by Stefaan Verhulst

Paper by Nick Hart, Adita Karkera, and Valerie Logan: “Advances in the access, collection, management, analysis, and use of data across public sector organizations substantially contributed to steady improvements in services, efficiency of operations, and effectiveness of government programs. The experience of citizens, beneficiaries, managers, and data experts is also evolving as data becomes pervasive and more seamlessly integrated within decision-making processes. In order for agencies to effectively engage in the ever-changing data landscape, organizational data literacy capacity and program models can help ensure individuals across the workforce can read, write, and communicate with data in the context of their role.

Data and analytics are no longer “just” for specialists, such as data engineers and data scientists; rather, data literacy is now increasingly recognized as a core workforce competency. Fortunately, in the United States several pioneers have emerged in strategically advancing data literacy programs and activities at the organizational level, providing benefits to individuals in the public sector workforce. Pioneering programs are those that recognize data literacy as more than training. They view data literacy as a holistic set of activities program to engage employees at all levels with data, develop employees with relevant skills, and enable scale of data literacy by augmenting employees’ skills with guided learning support and resources.

Agencies should begin by crafting the case for change. As is common with any emerging field, varying definitions and interpretations of “data literacy” are prevalent, which can affect program design. Being explicit in what problems are being solved for, as well as the needs and drivers to be addressed with a data literacy program or capacity, are vital to mitigate false starts…(More)”.

How Native Americans Are Trying to Debug A.I.’s Biases

Curated on Posted on by Stefaan Verhulst

Alex V. Cipolle in The New York Times: “In September 2021, Native American technology students in high school and college gathered at a conference in Phoenix and were asked to create photo tags — word associations, essentially — for a series of images.

One image showed ceremonial sage in a seashell; another, a black-and-white photograph circa 1884, showed hundreds of Native American children lined up in uniform outside the Carlisle Indian Industrial School, one of the most prominent boarding schools run by the American government during the 19th and 20th centuries.

For the ceremonial sage, the students chose the words “sweetgrass,” “sage,” “sacred,” “medicine,” “protection” and “prayers.” They gave the photo of the boarding school tags with a different tone: “genocide,” “tragedy,” “cultural elimination,” “resiliency” and “Native children.”

The exercise was for the workshop Teaching Heritage to Artificial Intelligence Through Storytelling at the annual conference for the American Indian Science and Engineering Society. The students were creating metadata that could train a photo recognition algorithm to understand the cultural meaning of an image.

The workshop presenters — Chamisa Edmo, a technologist and citizen of the Navajo Nation, who is also Blackfeet and Shoshone-Bannock; Tracy Monteith, a senior Microsoft engineer and member of the Eastern Band of Cherokee Indians; and the journalist Davar Ardalan — then compared these answers with those produced by a major image recognition app.

For the ceremonial sage, the app’s top tag was “plant,” but other tags included “ice cream” and “dessert.” The app tagged the school image with “human,” “crowd,” “audience” and “smile” — the last a particularly odd descriptor, given that few of the children are smiling.

The image recognition app botched its task, Mr. Monteith said, because it didn’t have proper training data. Ms. Edmo explained that tagging results are often “outlandish” and “offensive,” recalling how one app identified a Native American person wearing regalia as a bird. And yet similar image recognition apps have identified with ease a St. Patrick’s Day celebration, Ms. Ardalan noted as an example, because of the abundance of data on the topic….(More)”.

The first answer for food insecurity: data sovereignty

Curated on Posted on by Stefaan Verhulst

Interview by Brian Oaster: “For two years now, the COVID-19 pandemic has exacerbated almost every structural inequity in Indian Country. Food insecurity is high on that list.

Like other inequities, it’s an intergenerational product of dispossession and congressional underfunding — nothing new for Native communities. What is new, however, is the ability of Native organizations and sovereign nations to collectively study and understand the needs of the many communities facing the issue. The age of data sovereignty has (finally) arrived.

To that end, the Native American Agriculture Fund (NAAF) partnered with the Indigenous Food and Agricultural Initiative (INAI) and the Food Research and Action Center (FRAC) to produce a special report, Reimagining Hunger Responses in Times of Crisis, which was released in January.

According to the report, 48% of the more than 500 Native respondents surveyed across the country agreed that “sometimes or often during the pandemic the food their household bought just didn’t last, and they didn’t have money to get more.” Food security and access were especially low among Natives with young children or elders at home, people in fair to poor health and those whose employment was disrupted by the pandemic. “Native households experience food insecurity at shockingly higher rates than the general public and white households,” the report noted.

It also detailed how, throughout the pandemic, Natives overwhelmingly turned to their tribal governments and communities — as opposed to state or federal programs — for help. State and federal programs, like the Supplement Nutrition Assistance Program, or SNAP, don’t always mesh with the needs of rural reservations. A benefits card is useless if there’s no food store in your community. In response, tribes and communities came together and worked to get their people fed.

Understanding how and why will help pave the way for legislation that empowers tribes to provide for their own people, by using federal funding to build local agricultural infrastructure, for instance, instead of relying on assistance programs that don’t always work. HCN spoke with the Native American Agriculture Fund’s CEO, Toni Stanger-McLaughlin (Colville), to find out more…(More)”.

Towards a Standard for Identifying and Managing Bias in Artificial Intelligence

Curated on Posted on by Stefaan Verhulst

NIST Report: “As individuals and communities interact in and with an environment that is increasingly virtual they are often vulnerable to the commodification of their digital exhaust. Concepts and behavior that are ambiguous in nature are captured in this environment, quantified, and used to categorize, sort, recommend, or make decisions about people’s lives. While many organizations seek to utilize this information in a responsible manner, biases remain endemic across technology processes and can lead to harmful impacts regardless of intent. These harmful outcomes, even if inadvertent, create significant challenges for cultivating public trust in artificial intelligence (AI)….(More)”

Publicizing Corporate Secrets for Public Good

Curated on Posted on by Stefaan Verhulst

Paper by Christopher Morten: “Federal regulatory agencies in the United States hold a treasure trove of valuable information essential to a functional society. Yet little of this immense and nominally “public” resource is accessible to the public. That worrying phenomenon is particularly true for the valuable information that agencies hold on powerful private actors. Corporations regularly shield vast swaths of the information they share with federal regulatory agencies from public view, claiming that the information contains legally protected trade secrets (or other proprietary “confidential commercial information”). Federal agencies themselves have largely acceded to these claims and even fueled them, by construing restrictively various doctrines of law, including trade secrecy law, freedom of information law, and constitutional law. Today, these laws—and fear of these laws—have reduced to a trickle the flow of information that the public can access. This should not and need not be the case.

This article challenges the conventional wisdom that trade secrecy law restricts public agencies’ power to publicize private businesses’ secrets. In fact, federal agencies, and regulatory agencies especially, have long held and still hold statutory and constitutional authority to obtain and divulge otherwise secret information on private actors, when doing so serves the public interest. For many regulatory agencies, that authority extends even to bona fide trade secrets. In an age of “informational capitalism,” this disclosure authority makes U.S. federal regulatory agencies uniquely valuable—and perhaps uniquely dangerous. Building on recent work that explores this right in the context of drugs and vaccines, and drawing heavily from scholarship in privacy and information law, the article proposes a practical framework that regulators can use to publicize secret information in a way that maximizes public benefit and minimizes private harm. Rather than endorse unconstrained information disclosure—transparency for transparency’s sake—this article instead proposes controlled “information publicity,” in which regulators cultivate carefully bounded “gardens” of secret information. Within these gardens, agencies admit only certain users and certain uses of information. Drawing on existing but largely overlooked real-world examples, the article shows that regulators can effectively and selectively publicize trade secret information to noncommercial users while thwarting commercial uses. Regulators can protect trade secrets’ integrity vis-à-vis competitors while simultaneously unlocking new, socially valuable uses…(More)”.

Holding Out for Something Better

Curated on Posted on by Stefaan Verhulst

Essay by Rebecca Williams on the “Limits of Customer Service and Administrative Burden Frameworks” : “On December 13th, the Biden Administration published an Executive Order on Transforming Federal Customer Experience and Service Delivery to Rebuild Trust in Government. The EO promises to improve a slew of government services with the help of technology and rests on a theory of change that these “customer service” improvements will “engender trust,” but does not speak to changing the substance of these public goods, which may be the primary cause of the public’s trust issues, only their delivery. While the EO harkens to democratic principles, it makes no mention of how public input informed why they were prioritizing the delivery of the services mentioned versus other services.

Words are imbued with meaning and connotation and using “customer service” to describe the delivery of public goods has a dark side. It’s not just that the analogy doesn’t logically work — everything that makes “customer service” high quality in the business context is missing from government, there is no competition forcing the government to attract and retain customers — this phrase will not get us there. It’s that this mismatch of power dynamics makes it a dangerous phrase to substitute in. Calling the public “customers” implicitly reduces their participatory power to mere consumers and doesn’t fully embody the government’s duty to serve all its people well.

Michèle Champagne @michhhamI “love how “service design” and “design thinking” consultants have slowly invaded public policy circles, where public servants and policymakers are taught that “design skills“ are mandatory positive thinking, rapid prototyping, and problem solving. Thing is, that‘s solutionism.April 21st 2021“

It’s important in these times of diminished voter rights, rising police surveillance, and prosecution of protestors to protect our democratic rights and be wary of anyone co-opting democratic language for lesser rights. As illustrated by Michèle Champagne’s brilliant tweet (above), asking for feedback after the bulk of the substance has been decided isn’t democratic, it’s providing a very small set of choices and dressing it up as democratic.

Let’s move away from consumer language for public goods to participatory and rights-based language; let’s lead delivery improvement initiatives with public input and place these improvements in the service of larger debates about what collective goods we want to have as a community. For example, if 63% of the population is supportive of healthcare for all, let’s be sure related public service improvements contemplate and serve that substantive expansion; investing in more application infrastructure might make less sense than considering how technology can support the issuance of universal medicare cards or uniform reporting standards. This is a job the Executive Branch could spearhead (the Federal Government takes on pilots projects routinely with input from the public), but it is also one the larger civic tech community should hold in their minds as a possibility…(More)”.