Physical Sciences and Mathematics Commons^™

Regulating Machine Learning: The Challenge Of Heterogeneity, Cary Coglianese

All Faculty Scholarship

Machine learning, or artificial intelligence, refers to a vast array of different algorithms that are being put to highly varied uses, including in transportation, medicine, social media, marketing, and many other settings. Not only do machine-learning algorithms vary widely across their types and uses, but they are evolving constantly. Even the same algorithm can perform quite differently over time as it is fed new data. Due to the staggering heterogeneity of these algorithms, multiple regulatory agencies will be needed to regulate the use of machine learning, each within their own discrete area of specialization. Even these specialized expert agencies, though, …

Assessing Automated Administration, Cary Coglianese, Alicia Lai

All Faculty Scholarship

To fulfill their responsibilities, governments rely on administrators and employees who, simply because they are human, are prone to individual and group decision-making errors. These errors have at times produced both major tragedies and minor inefficiencies. One potential strategy for overcoming cognitive limitations and group fallibilities is to invest in artificial intelligence (AI) tools that allow for the automation of governmental tasks, thereby reducing reliance on human decision-making. Yet as much as AI tools show promise for improving public administration, automation itself can fail or can generate controversy. Public administrators face the question of when exactly they should use automation. …

Moving Toward Personalized Law, Cary Coglianese

All Faculty Scholarship

Rules operate as a tool of governance by making generalizations, thereby cutting down on government officials’ need to make individual determinations. But because they are generalizations, rules can result in inefficient or perverse outcomes due to their over- and under-inclusiveness. With the aid of advances in machine-learning algorithms, however, it is becoming increasingly possible to imagine governments shifting away from a predominant reliance on general rules and instead moving toward increased reliance on precise individual determinations—or on “personalized law,” to use the term Omri Ben-Shahar and Ariel Porat use in the title of their 2021 book. Among the various technological, …

Algorithm Vs. Algorithm, Cary Coglianese, Alicia Lai

All Faculty Scholarship

Critics raise alarm bells about governmental use of digital algorithms, charging that they are too complex, inscrutable, and prone to bias. A realistic assessment of digital algorithms, though, must acknowledge that government is already driven by algorithms of arguably greater complexity and potential for abuse: the algorithms implicit in human decision-making. The human brain operates algorithmically through complex neural networks. And when humans make collective decisions, they operate via algorithms too—those reflected in legislative, judicial, and administrative processes. Yet these human algorithms undeniably fail and are far from transparent. On an individual level, human decision-making suffers from memory limitations, fatigue, …

From Negative To Positive Algorithm Rights, Cary Coglianese, Kat Hefter

All Faculty Scholarship

Artificial intelligence, or “AI,” is raising alarm bells. Advocates and scholars propose policies to constrain or even prohibit certain AI uses by governmental entities. These efforts to establish a negative right to be free from AI stem from an understandable motivation to protect the public from arbitrary, biased, or unjust applications of algorithms. This movement to enshrine protective rights follows a familiar pattern of suspicion that has accompanied the introduction of other technologies into governmental processes. Sometimes this initial suspicion of a new technology later transforms into widespread acceptance and even a demand for its use. In this paper, we …

Antitrust By Algorithm, Cary Coglianese, Alicia Lai

All Faculty Scholarship

Technological innovation is changing private markets around the world. New advances in digital technology have created new opportunities for subtle and evasive forms of anticompetitive behavior by private firms. But some of these same technological advances could also help antitrust regulators improve their performance in detecting and responding to unlawful private conduct. We foresee that the growing digital complexity of the marketplace will necessitate that antitrust authorities increasingly rely on machine-learning algorithms to oversee market behavior. In making this transition, authorities will need to meet several key institutional challenges—building organizational capacity, avoiding legal pitfalls, and establishing public trust—to ensure successful …

Regulating New Tech: Problems, Pathways, And People, Cary Coglianese

All Faculty Scholarship

New technologies bring with them many promises, but also a series of new problems. Even though these problems are new, they are not unlike the types of problems that regulators have long addressed in other contexts. The lessons from regulation in the past can thus guide regulatory efforts today. Regulators must focus on understanding the problems they seek to address and the causal pathways that lead to these problems. Then they must undertake efforts to shape the behavior of those in industry so that private sector managers focus on their technologies’ problems and take actions to interrupt the causal pathways. …

What Is The Relationship Between Language And Thought?: Linguistic Relativity And Its Implications For Copyright, Christopher S. Yoo

All Faculty Scholarship

To date, copyright scholarship has almost completely overlooked the linguistics and cognitive psychology literature exploring the connection between language and thought. An exploration of the two major strains of this literature, known as universal grammar (associated with Noam Chomsky) and linguistic relativity (centered around the Sapir-Whorf hypothesis), offers insights into the copyrightability of constructed languages and of the type of software packages at issue in Google v. Oracle recently decided by the Supreme Court. It turns to modularity theory as the key idea unifying the analysis of both languages and software in ways that suggest that the information filtering associated …

Power And Statistical Significance In Securities Fraud Litigation, Jill E. Fisch, Jonah B. Gelbach

All Faculty Scholarship

Event studies, a half-century-old approach to measuring the effect of events on stock prices, are now ubiquitous in securities fraud litigation. In determining whether the event study demonstrates a price effect, expert witnesses typically base their conclusion on whether the results are statistically significant at the 95% confidence level, a threshold that is drawn from the academic literature. As a positive matter, this represents a disconnect with legal standards of proof. As a normative matter, it may reduce enforcement of fraud claims because litigation event studies typically involve quite low statistical power even for large-scale frauds.

This paper, written for …

Administrative Law In The Automated State, Cary Coglianese

All Faculty Scholarship

In the future, administrative agencies will rely increasingly on digital automation powered by machine learning algorithms. Can U.S. administrative law accommodate such a future? Not only might a highly automated state readily meet longstanding administrative law principles, but the responsible use of machine learning algorithms might perform even better than the status quo in terms of fulfilling administrative law’s core values of expert decision-making and democratic accountability. Algorithmic governance clearly promises more accurate, data-driven decisions. Moreover, due to their mathematical properties, algorithms might well prove to be more faithful agents of democratic institutions. Yet even if an automated state were …

Contracting For Algorithmic Accountability, Cary Coglianese, Erik Lampmann

All Faculty Scholarship

As local, state, and federal governments increase their reliance on artificial intelligence (AI) decision-making tools designed and operated by private contractors, so too do public concerns increase over the accountability and transparency of such AI tools. But current calls to respond to these concerns by banning governments from using AI will only deny society the benefits that prudent use of such technology can provide. In this Article, we argue that government agencies should pursue a more nuanced and effective approach to governing the governmental use of AI by structuring their procurement contracts for AI tools and services in ways that …

Environmental Soft Law As A Governance Strategy, Cary Coglianese

All Faculty Scholarship

Soft law governance relies on nongovernmental institutions that establish and implement voluntary standards. Compared with traditional hard law solutions to societal and economic problems, soft law alternatives promise to be more politically feasible to establish and then easier to adapt in the face of changing circumstances. They may also seem more likely to be flexible in what they demand of targeted businesses and other entities. But can soft law actually work to solve major problems? This Article considers the value of soft law governance through the lens of three major voluntary, nongovernmental initiatives that address environmental concerns: (1) ISO 14001 …

On Environmental, Climate Change & National Security Law, Mark P. Nevitt

All Faculty Scholarship

This Article offers a new way to think about climate change. Two new climate change assessments — the 2018 Fourth National Climate Assessment (NCA) and the United Nations Intergovernmental Panel’s Special Report on Climate Change — prominently highlight climate change’s multifaceted national security risks. Indeed, not only is climate change a “super wicked” environmental problem, it also accelerates existing national security threats, acting as both a “threat accelerant” and “catalyst for conflict.” Further, climate change increases the intensity and frequency of extreme weather events while threatening nations’ territorial integrity and sovereignty through rising sea levels. It causes both internal displacement …

Deploying Machine Learning For A Sustainable Future, Cary Coglianese

All Faculty Scholarship

To meet the environmental challenges of a warming planet and an increasingly complex, high tech economy, government must become smarter about how it makes policies and deploys its limited resources. It specifically needs to build a robust capacity to analyze large volumes of environmental and economic data by using machine-learning algorithms to improve regulatory oversight, monitoring, and decision-making. Three challenges can be expected to drive the need for algorithmic environmental governance: more problems, less funding, and growing public demands. This paper explains why algorithmic governance will prove pivotal in meeting these challenges, but it also presents four likely obstacles that …

Regulation Of Algorithmic Tools In The United States, Christopher S. Yoo, Alicia Lai

All Faculty Scholarship

Policymakers in the United States have just begun to address regulation of artificial intelligence technologies in recent years, gaining momentum through calls for additional research funding, piece-meal guidance, proposals, and legislation at all levels of government. This Article provides an overview of high-level federal initiatives for general artificial intelligence (AI) applications set forth by the U.S. president and responding agencies, early indications from the incoming Biden Administration, targeted federal initiatives for sector-specific AI applications, pending federal legislative proposals, and state and local initiatives. The regulation of the algorithmic ecosystem will continue to evolve as the United States continues to search …

Decarbonization In Democracy, Shelley Welton

All Faculty Scholarship

Conventional wisdom holds that democracy is structurally ill-equipped to confront climate change. As the story goes, because each of us tends to dismiss consequences that befall people in other places and in future times, “the people” cannot be trusted to craft adequate decarbonization policies, designed to reduce present-day, domestic carbon emissions. Accordingly, U.S. climate change policy has focused on technocratic fixes that operate predominantly through executive action to escape democratic politics — with vanishingly little to show for it after a change in presidential administration. To help craft a more durable U.S. climate change strategy, this Article scrutinizes the purported …

Transparency And Algorithmic Governance, Cary Coglianese, David Lehr

All Faculty Scholarship

Machine-learning algorithms are improving and automating important functions in medicine, transportation, and business. Government officials have also started to take notice of the accuracy and speed that such algorithms provide, increasingly relying on them to aid with consequential public-sector functions, including tax administration, regulatory oversight, and benefits administration. Despite machine-learning algorithms’ superior predictive power over conventional analytic tools, algorithmic forecasts are difficult to understand and explain. Machine learning’s “black-box” nature has thus raised concern: Can algorithmic governance be squared with legal principles of governmental transparency? We analyze this question and conclude that machine-learning algorithms’ relative inscrutability does not pose a …

Lowering Legal Barriers To Rpki Adoption, Christopher S. Yoo, David A. Wishnick

All Faculty Scholarship

Across the Internet, mistaken and malicious routing announcements impose significant costs on users and network operators. To make routing announcements more reliable and secure, Internet coordination bodies have encouraged network operators to adopt the Resource Public Key Infrastructure (“RPKI”) framework. Despite this encouragement, RPKI’s adoption rates are low, especially in North America.

This report presents the results of a year-long investigation into the hypothesis—widespread within the network operator community—that legal issues pose barriers to RPKI adoption and are one cause of the disparities between North America and other regions of the world. On the basis of interviews and analysis of …

The Logic And Limits Of Event Studies In Securities Fraud Litigation, Jill E. Fisch, Jonah B. Gelbach, Jonathan Klick

All Faculty Scholarship

Event studies have become increasingly important in securities fraud litigation after the Supreme Court’s decision in Halliburton II. Litigants have used event study methodology, which empirically analyzes the relationship between the disclosure of corporate information and the issuer’s stock price, to provide evidence in the evaluation of key elements of federal securities fraud, including materiality, reliance, causation, and damages. As the use of event studies grows and they increasingly serve a gatekeeping function in determining whether litigation will proceed beyond a preliminary stage, it will be critical for courts to use them correctly.

This Article explores an array of …

Antitrust And Information Technologies, Herbert J. Hovenkamp

All Faculty Scholarship

Technological change strongly affects the use of information to facilitate anticompetitive practices. The effects result mainly from digitization and the many products and processes that it enables. These technologies of information also account for a significant portion of the difficulties that antitrust law encounters when its addresses intellectual property rights. In addition, changes in the technologies of information affect the structures of certain products, in the process either increasing or decreasing the potential for competitive harm.

For example, digital technology affects the way firms exercise market power, but it also imposes serious measurement difficulties. The digital revolution has occurred in …

Why Law Now Needs To Control Rather Than Follow Neo-Classical Economics, John William Draper

Librarian Scholarship at Penn Law

Selfish utilitarianism, neo-classical economics, the directive of short-term income maximization, and the decision tool of cost-benefit analysis fail to protect our species from the significant risks of too much consumption, pollution, or population. For a longer-term survival, humanity needs to employ more than cost-justified precaution.

This article argues that, at the global level, and by extension at all levels of government, we need to replace neo-classical economics with filters for safety and feasibility to regulate against significant risk. For significant risks, especially those that are irreversible, we need decision tools that will protect humanity at all scales. This article describes …

Wireless Network Neutrality: Technological Challenges And Policy Implications, Christopher S. Yoo

All Faculty Scholarship

One key aspect of the debate over network neutrality has been whether and how network neutrality should apply to wireless networks. The existing commentary has focused on the economics of wireless network neutrality, but to date a detailed analysis of how the technical aspects of wireless networks affect the implementation of network neutrality has yet to appear in the literature. As an initial matter, bad handoffs, local congestion, and the physics of wave propagation make wireless broadband networks significantly less reliable than fixed broadband networks. These technical differences require the network to manage dropped packets and congestion in a way …

Can We Learn Anything About Pleading Changes From Existing Data?, Jonah B. Gelbach

All Faculty Scholarship

In light of the gateway role that the pleading standard can play in our civil litigation system, measuring the empirical effects of pleading policy changes embodied in the Supreme Court's controversial Twombly and Iqbal cases is important. In my earlier paper, Locking the Doors to Discovery, I argued that in doing so, special care is required in formulating the object of empirical study. Taking party behavior seriously, as Locking the Doors does, leads to empirical results suggesting that Twombly and Iqbal have had substantial effects among cases that face Rule 12(b)(6) motions post-Iqbal. This paper responds to …

Moore’S Law, Metcalfe’S Law, And The Theory Of Optimal Interoperability, Christopher S. Yoo

All Faculty Scholarship

Many observers attribute the Internet’s success to two principles: Moore’s Law and Metcalfe’s Law. These precepts are often cited to support claims that larger networks are inevitably more valuable and that costs in a digital environment always decrease. This Article offers both a systematic description of both laws and then challenges the conventional wisdom by exploring their conceptual limitations. It also explores how alternative mechanisms, such as gateways and competition, can permit the realization benefits typically attributed to Moore’s Law and Metcalfe’s Law without requiring increases in network size.

Competition Policy And The Technologies Of Information, Herbert J. Hovenkamp

All Faculty Scholarship

When we speak about information and competition policy we are usually thinking about oral or written communications that have an anticompetitive potential, and mainly in the context of collusion of exclusionary threats. These are important topics. Indeed, among the most difficult problems that competition policy has had to confront over the years is understanding communications that can be construed as either threats to exclude or as offers to collude or facilitators of collusion.

My topic here, however, is the relationship between information technologies and competition policy. Technological change can both induce and undermine the use of information to facilitate anticompetitive …

Expert Mining And Required Disclosure, Jonah B. Gelbach

All Faculty Scholarship

No abstract provided.

An Empirical Analysis Of Cost Recovery In Superfund Cases: Implications For Brownfields And Joint And Several Liability, Howard F. Chang, Hilary Sigman

All Faculty Scholarship

Economic theory developed in the prior literature indicates that under the joint and several liability imposed by the federal Superfund statute, the government should recover more of its costs of cleaning up contaminated sites than it would under nonjoint liability, and the amount recovered should increase with the number of defendants and with the independence among defendants in trial outcomes. We test these predictions empirically using data on outcomes in federal Superfund cases. Theory also suggests that this increase in the amount recovered may discourage the sale and redevelopment of potentially contaminated sites (or “brownfields”). We find the increase to …

Competition In Information Technologies: Standards-Essential Patents, Non-Practicing Entities And Frand Bidding, Herbert J. Hovenkamp

All Faculty Scholarship

Standard Setting is omnipresent in networked information technologies. Virtually every cellular phone, computer, digital camera or similar device contains technologies governed by a collaboratively developed standard. If these technologies are to perform competitively, the processes by which standards are developed and implemented must be competitive. In this case attaining competitive results requires a mixture of antitrust and non-antitrust legal tools.

FRAND refers to a firm’s ex ante commitment to make its technology available at a “fair, reasonable and nondiscriminatory royalty.” The FRAND commitment results from bidding to have one’s own technology selected as a standard. Typically the FRAND commitment is …

Enhancing Public Access To Online Rulemaking Information, Cary Coglianese

All Faculty Scholarship

One of the most significant powers exercised by federal agencies is their power to make rules. Given the importance of agency rulemaking, the process by which agencies develop rules has long been subject to procedural requirements aiming to advance democratic values of openness and public participation. With the advent of the digital age, government agencies have engaged in increasing efforts to make rulemaking information available online as well as to elicit public participation via electronic means of communication. How successful are these efforts? How might they be improved? In this article, I investigate agencies’ efforts to make rulemaking information available …

Cloud Computing: Architectural And Policy Implications, Christopher S. Yoo

All Faculty Scholarship

Cloud computing has emerged as perhaps the hottest development in information technology. Despite all of the attention that it has garnered, existing analyses focus almost exclusively on the issues that surround data privacy without exploring cloud computing’s architectural and policy implications. This article offers an initial exploratory analysis in that direction. It begins by introducing key cloud computing concepts, such as service-oriented architectures, thin clients, and virtualization, and discusses the leading delivery models and deployment strategies that are being pursued by cloud computing providers. It next analyzes the economics of cloud computing in terms of reducing costs, transforming capital expenditures …