Employees are often left unprotected from discrimination because they are unable to satisfy the requirement of causation. Courts have made clear that to obtain legal redress for discrimination, it is generally insufficient to show that a protected characteristic such as race or sex was a “motivating factor” of an adverse employment decision. Rather, under Supreme Court precedent—including the Court’s Comcast and Babb decisions in the 2020 term—the antidiscrimination statutes generally require a showing of “but-for” causation. This means that employees can rarely prevail because it is often easy for an employer to rebut allegations of discrimination by asserting a legitimate purpose—true or not—for the adverse decision. Therefore, although there is good reason to reject the motivating-factor test, the but-for requirement undermines the objectives of antidiscrimination law. In this article, I draw on notions of cause and effect in the sciences and in tort law to propose a new standard of causation for antidiscrimination law. In particular, I formulate a simple test—which I call the “fortified NESS” test, or “FNESS”—for courts and legislatures to apply as a uniform and effective standard of causation in all disparate-treatment cases. I then employ this formulation to propose concrete amendments to the civil rights statutes, and I demonstrate why these amendments are necessary and how they allow courts to uphold the critical aims of antidiscrimination law.
The Supreme Court has said that general jurisdiction provides at least one clear and certain forum to sue defendants, and that assumption has begun to shape the Court’s understanding of specific jurisdiction. But that assumption is wrong. General jurisdiction does not provide a guaranteed U.S. forum for foreign defendants or in cases involving multiple defendants. And even when defendants can be sued “at home,” such cases may be (and not infrequently are) dismissed for forum non conveniens, sometimes even when no alternative forum is available. Nor is a regime reliant on a general jurisdiction backstop desirable. The Court’s narrowed version of general jurisdiction creates incentives for states to favor local defendants—as Michigan has done through choice-of-law rules that give preference to lex fori and substantive laws that favor car manufacturers. Overreliance on general jurisdiction also channels litigation to states that may not want it—a concern already voiced by Delaware courts. This essay warns against developing the law of personal jurisdiction on the assumption that general jurisdiction will guarantee an available forum in which to sue defendants. Instead, we argue that the primary engine of personal jurisdiction must remain a flexible doctrine of specific jurisdiction. Rather than hunting for new formalisms in specific jurisdiction’s relatedness requirement, the Justices should embrace specific jurisdiction’s reasonableness factors as a ready-made tool for answering their recent worries.
From IBM Watson's success in Jeopardy! to Google DeepMind's victories in Go, the past decade has seen artificial intelligence advancing in leaps and bounds. Such advances have captured the attention of not only computer experts and academic commentators but also policymakers, the mass media and the public at large. In recent years, legal scholars have also actively explored how artificial intelligence will impact the law. Such exploration has resulted in a fast-growing body of scholarship. One area that has not received sufficient policy and scholarly attention concerns the law-machine interface in a hybrid environment in which both humans and intelligent machines will make legal decisions at the same time. To fill this void, the present article utilizes the case study of fair use automation to explore how legal standards can be automated and what this specific case study can teach us about the law-machine interface. Although this article focuses on an example generated from a specialized area of the law—namely, copyright or intellectual property law—its insights will apply to other situations involving the interplay of artificial intelligence and the law. The article begins by outlining the case study of fair use automation and examining three dominant arguments against such automation. Taking seriously the benefits provided by artificial intelligence, machine learning and big data analytics, this article then identifies three distinct pathways for legal automation: translation, approximation and self-determination. The second half of the article turns to key questions concerning the law-machine interface, the understanding of which will be important when automated systems are being designed to implement legal standards. Specifically, these questions focus on the allocation of decision-making power, the hierarchy of decisions and the legal effects of machine-made decisions. The article concludes by highlighting the wide-ranging ramifications of artificial intelligence for the law, the legislature, the bench, the bar and academe.
An arrest puts a halt to one’s free life and may act as prelude to a new process. That new process—prosecution—may culminate in a finding of guilt. But arrest and guilt—concepts that are factually and legally distinct—frequently seem to be fused together. This fusion appears in many of the consequences of arrest, including the use of arrests in assessing “risk,” in calculating “recidivism,” and in identifying “offenders.” An examination of this fusion elucidates obstacles to key aspects of criminal justice reform. Efforts at reform, whether focused on prosecution or defense, police or bail, require a robust understanding of the differences between arrest and guilt; if they run counter to an implicit fusion of the two, they will inevitably falter.
In recent years, shareholders have started to launch information lawsuits with new vigor. These demands — which often yield valuable corporate data without an explicit contractual right or public disclosure obligation — are typically brought in connection with a shareholder grievance about some other matter. Firm managers often resist the intrusion, leading to some blockbuster disputes. Indeed, it is becoming increasingly clear that information litigation is starting to play a much greater gatekeeping role for corporate governance problems. There are often serial litigation battles where a fight over access to firm information is followed by a fight over the primary governance concern. Delays associated with the initial information lawsuit can even stymie resolution of the primary case. Notwithstanding this increase in information litigation, lawmakers lack a comprehensive theory for evaluating the private right to corporate information. Many courts and commentators offer only short statements about a need for balance: shareholders should be able to obtain information necessary to exercise their rights, but they should not be able to harass managers or expose crucial corporate secrets. Legal standards for adjudicating private, ex-post information demands are vague. Most disputes are decided by asking whether the requested information is “necessary and essential” for pursuing a “proper shareholder purpose.” But how should we award and scope this right? This article argues that inspection rights are best justified in connection with efforts to mitigate managerial agency costs through forensic review. Much of corporate law seeks to maximize the benefits of centralized economic activity while minimizing the costs of suboptimal agent decision-making. Endless varieties of agency problems exist, but the distortions are all grounded in information asymmetry. Shareholder inspection rights are thus well-suited to addressing agency lapses, and they are warranted in connection with this fundamental aim. An important corollary to this claim, however, is that shareholder inspection rights may not be a fundamental entitlement when deployed more broadly. If so, then corporations should enjoy some latitude to modify inspection rights via private ordering.
Nationwide injunctions have become a focus of heated judicial, academic, and even public debate. Much of this analysis treats nationwide injunctions as a unitary concept, referring to a particular type of court order. In fact, the term may apply to five different categories of orders of national applicability, each of which raises very different constitutional, fairness, rule-based, structural, prudential, and other concerns. This Article presents a taxonomy of the five types of nationwide injunctions and the proper judicial treatment of each. Rather than focusing on the geographic applicability and scope of a court order, in-junctions should instead be categorized based on the entities whose rights they seek to enforce and whether the case is a class action. Based on these considerations, the proposed taxonomy distinguishes among “nationwide plaintiff-oriented injunctions,” “nationwide plaintiff-class injunctions,” “nation-wide associational injunctions,” “nationwide defendant-oriented injunctions,” and “nationwide private enforcement injunctions.” After presenting this new framework for determining the validity of nationwide injunctions, this Article goes on to demonstrate that stare decisis, rather than nationwide defendant-oriented injunctions or even class certification under Federal Rule of Civil Procedure Rule 23(b)(2), is the most appropriate means of protecting the rights of third parties who are not personally involved in litigation. Affording district- or circuit-wide stare decisis effect to district court rulings allows members of the public to benefit from them and reduces the need for wasteful relitigation. At the same time, this approach recognizes the limited authority of lower court judges in our decentralized, hierarchical judiciary; mitigates the effects of extreme forum shopping; and ensures some degree of percolation of important constitutional issues.
From Catalonia to Kurdistan, Crimea to California, and Scotland to St. Kitts, secession has become a hotly debated political issue. While secessionist disputes appear to hinge on high power politics, they also have an important constitutional dimension. This Article offers a comprehensive exploration of how the world’s constitutions treat secession, and how constitutional secession clauses can affect real-world secessionist disputes. Drawing on an original dataset that captures how the world’s constitutions have dealt with secession from 1789 to 2015, we make a number of contributions. First, we document the various ways in which constitutions have dealt with secession. While the bulk of the literature has been focused on a right to secession, such clauses are exceedingly rare in constitutions today, and most of the countries that constitutionalized such a right have broken up. Instead, where constitutions deal with secession, they tend to prohibit it, either by banning it explicitly or through an implicit prohibition that emphasizes territorial integrity. Other constitutions remain silent on the matter, even when there are live secessionist disputes. We provide a taxonomy of these various approaches and show how they have changed over time. Second, we provide an analysis of how constitutional design choices about secession can have important real-world consequences. We develop a theory of how different constitutional design choices can affect downstream mobilization for secession. We test this theory empirically by replicating a number of existing studies, to which we add our own new data. Granting a right to secession, we find, is associated with a higher chance of a nation breaking up, and makes such break-ups less violent. By contrast, a clear and explicit prohibition of secession decreases popular support for secessionist movements and reduces associated violence, thus reducing the odds of a nation breaking up. Perhaps the worst option is for the constitution to remain silent on secession. Where the constitution is silent on secession, we find that secessionist movements are motivated to seize on this ambiguity and are able to garner popular support for the movement. Yet, these movements are surrounded by relatively high levels of violence, as central governments end up opposing the claim or the country’s highest court rules that it is without constitutional basis. We find similar results for implicit prohibitions: they increase both popular support and secessionist violence. Overall, these findings suggest that constitutional ambiguity can be harmful: it allows secessionist movements to garner support, but produces higher level of violence without improving the chances of success. We further probe these findings by studying two recent high-profile secessionist disputes: Catalonia in Spain and Kurdistan in Iraq, which confirm our findings. Our findings have important real-world implications for constitutional drafting. As a number of countries are debating various degrees of autonomy for regionally clustered minority groups, our findings suggest that it is important to address the question of secession head-on and that careful constitutional drafting can avoid violence and instability.
For well over a century, legislators, courts, lawyers, and scholars have spent significant time and energy debating the optimal duration of copyright protection. While there is general consensus that copyright’s term is of legal and economic significance, arguments both for and against a lengthy term are often impressionistic. Utilizing music industry sales data not previously available for academic analysis, this article fills an important evidentiary gap in the literature. Using recorded music as a case study, we determine that most copyrighted music earns the majority of its lifetime revenue in the first 5-10 years following its initial release (and in many cases, far sooner than that). Our analysis suggests at least two results of interest to legislators, lawyers, and scholars alike: First, it contributes to the normative debate around copyright’s incentive-access paradigm by proposing a more efficient conception of copyright’s term for information goods; namely, one that replaces the conventional “life plus” durational standard with one based on the commercial viability of the average work. Second, it demonstrates that advocates’ and legislators’ tendency to focus on atypical works leads to overprotection of the average work, suggesting that copyright’s term is not nearly as significant for copyright owners as conventional wisdom submits.
The digital economy is changing everything, including how we borrow money. In the wake of the 2008 crisis, banks pulled back in their lending and, as a result, many consumers and small businesses found themselves unable to access credit. A wave of online firms called fintech lenders have filled the space left vacant by traditional financial institutions. These platforms are fast making antiques out of many mainstream lending practices, such as long paper applications and face-to-face meetings. Instead, through underwriting by automation — utilizing big data (including social media data) and machine learning — loan processing that once took days for mainstream lenders can now be done in minutes by fintech firms. The result of these fintech advances has been quicker access to capital, more economic efficiencies, and even greater prospects for access to credit for the unbanked and underbanked. “Click here” is the new “sign on the dotted line.” But there is a lot still to learn about fintech lending. How do these firms work and what kinds of products do they offer? Moreover, what role will they play in the future of American debt markets, particularly when it comes to the role of machine learning in assessing a borrower’s creditworthiness? This Article explores these questions and assesses current government responses to the nascent industry. It also surveys the current consumer protection landscape for fintech lenders and analyzes a multi-year dataset of complaints submitted to the CFPB relative to products offered by these firms. The Article concludes by offering several policy recommendations for how to regulate this new world of bitcredit.
States prosecute and incarcerate thousands of fathers every year for failing to pay their child support obligations. Ostensibly, these prosecutions aim to foster the health and well-being of children, without requiring the child’s mother to bear the costs of raising the child. What may appear on the surface to be a system that balances out inequities is actually a deeply flawed government program, one that promotes criminal recidivism and reinforces the poverty of indigent fathers. Contrary to the common image of a “deadbeat dad” raking in money and staying on the lam to avoid helping a mother raise their child, the vast majority of fathers who owe large amounts of child support make little to no income. These fathers do not pay their child support obligations because they are unable to. Once a state prosecutes and imprisons an indigent father, his odds of being able to pay that debt diminish even more. The criminalization of failing to pay child support is unconstitutional and revives prohibited debtors’ prisons. Fathers of lesser means are being incarcerated for failing to pay a private debt owed to their child’s mother or for not reimbursing the government for costs provided to the mother in the form of state assistance. In other words, an indigent father may be criminally sanctioned for not subsidizing the government’s welfare programs. Relying on antiquated ideas about a father’s role in the family, the child support system punishes poor fathers for their reproductive choices, and morally condemns them for bringing a child into the world without being able to financially provide for her. This Article calls for an end to the criminalization of failing to pay child support, and proposes several fiscally responsible changes to the current system that will improve the welfare of children whose parents are no longer together.
Centralized oversight of agency policymaking and spending by the President’s Office of Management and Budget is a hallmark of the modern administrative state. But tax regulations have almost never been subject to centralized review. The Trump administration recently proposed to require centralized review of tax regulations, but it is unclear what regulations would be subject to such review or how it would be conducted. This Article examines the normative desirability of the longstanding approach of exempting tax regulations from centralized review, and the alternative of imposing such review. Scholars and policymakers have provided various incomplete justifications for excepting tax policy from centralized review, including concerns about politicizing tax administration, analytical challenges, and ossification. I conclude that none of the reasons offered in the past for a default rule of no review is sufficient in light of the potential normative benefits of centralized review. The analysis here brings to the fore multiple functions of tax regulations: some rules are focused on shaping private behavior, whereas others focus on raising revenue or implementing precise congressional directives. I make the case that centralized review can facilitate productive coordination with other parts of government, increase political accountability, and introduce analytical rigor through quantified analysis. This Article outlines the limitations of current centralized review conventions, and proposes some specific adjustments for tax regulations, including setting a threshold for review based on revenue estimates, and producing revenue estimates and distributional analysis as part of the regulation-drafting process. The major tax legislation Congress enacted at the end of 2017 included numerous broad delegations. Thus, tax regulations will reshape the tax system significantly, and the Trump administration is poised to begin some version of centralized review of tax regulations, although many important issues remain unresolved. Recognizing the strengths and weaknesses of centralized review as applied to tax policy will help to establish consistent and productive oversight of the tax regulatory process.
This essay explores one central part of Tom Robinson’s trial in Harper Lee’s To Kill a Mockingbird: Atticus Finch’s cross-examination of Mayella Ewell. The eight- year-old Scout cannot fully understand the strategy and meaning of Atticus’ questions, but the trial supplies enough clues to understand more of Mayella’s life than is generally understood.
The now-conventional account of patent law holds that infringement is a strict liability offense, meaning that intent is not an element of an infringement claim. This account heightens the apparent injustice of patent law’s special knowledge problem, that as ambiguous descriptions of intangible resources, patent claims do not sufficiently make potential infringers aware of a patentee’s right to exclude. Particularly in the age of so-called “patent thickets,” clusters of patents of variable merit which are indistinguishable from each other and from prior art, strict liability for infringement seems rather hard. These problems reflect a conceptual misunderstanding. When infringement is understood as a species of trespass, as it was long described in American law, the various aspects of infringement doctrine fall into place. Common law traditionally recognizes three forms of trespass. Together, those three forms explain all of infringement doctrine as a coherent whole and resolve the apparent injustices that seem problematic on the conventional account of infringement. The only aspect of infringement doctrine that does not fit the trespass picture is the four-factor interpretation of eBay, which is contrary to the Justices’ insistence in eBay that they were neither overturning patent infringement doctrine nor disturbing traditional equitable maxims.
This paper analyzes the survival of organizations in which decision agents do not bear a major share of the wealth effects of their decisions. This is what the literature on large corporations calls separation of ownership and control. Such separation of decision and risk bearing functions is also common to organizations like large professional partnerships, financial mutuals and nonprofits. We contend that separation of decision and risk bearing functions survives in these organizations in part because of the benefits of specialization of management and risk bearing but also because of an effective common approach to controlling the implied agency problems. In particular, the contract structures of all these organizations separate the ratification and monitoring of decisions from the initiation and implementation of the decisions.
More than a third of our states have given judges a little-known power to dismiss prosecutions, not because of legal or factual insufficiency, but for the sake of justice. Whether phrased as dismissals “in furtherance of justice” or dismissals of de minimis prosecutions, these exercises of judicial power teach two important lessons. First, judges exercising these dismissals are rebutting the common notion that in the face of over-criminalization and over-incarceration they are powerless to do more than rubber-stamp prosecutorial decision making. In individual cases, they push back against some of the most problematic aspects of our criminal justice system: its size, harshness, and bias. Second, these cases converge on shared principles of justice. These principles conjure a vision of a very different criminal justice system: one in which an alleged criminal act is viewed not in isolation, but within a broader context that includes the apparent motivations for it, the state’s role in and response to it, and possible responses other than the criminal law. There is no logical reason to confine these principles to this procedural context, and the Article urges their broader consideration.
Does Congress have power to deny copyright protection for specific content? The Copyright Clause grants Congress power to “promote the Progress of Science” by legislating copyright laws. Certainly some content may reasonably be viewed as failing to promote the progress of science. Violent video games or pornography, for instance, may reasonably be viewed as not promoting progress in science, even though they receive protection as free speech under the First Amendment. So even if the Free Speech Clause bars Congress from banning content, does the Copyright Clause provide Congress a permissible means to discourage production of that content?This Article considers whether such content-based copyright denial is permissible under Congress’s copyright power. Neither courts nor scholars have considered this question, despite the fact that lawmakers are presently seeking to control negative effects of specific content. This Article posits that the copyright power provides Congress that means. The Copyright Clause’s mandate to promote the progress of science suggests a power to exercise content discrimination. At the same time, denying copyright to content would not prevent content creators from engaging in, and even profiting from, any speech protected by the First Amendment. The Article concludes that the Copyright Clause provides a constitutional tool for fixing content-based problems.
