I am enormously grateful to the distinguished judges and scholars who have read and engaged with my book Constitutional Rights and Constitutional Design, and appreciate both the praise and criticism. In this reply I respond to some points they raise. As I write, news from the US brings word that a district judge in Texas has declared invalid the 2010 Patient Protection Affordable Care Act—popularly known as Obamacare—on the ground that its mandate to purchase health insurance is beyond Congress’s jurisdiction. Although the judge has stayed his order pending appeal—in contrast to judges who recently have issued nationwide injunctions against laws and executive orders on immigration—the ruling casts a cloud of uncertainty over the law and the insurance policies of millions of citizens. Earlier this year the same Texas judge struck down a federal law on adoption of Native American Indian children on the ground that it violated the principle of equality.

These cases dramatically illustrate the problem my book addresses. Does it make sense to have a constitutional system in which a single judge—or for that matter, a panel of judges—has the power to decide questions about health insurance, adoption law, immigration policy, and other matters of political controversy that have momentous consequences for the entire nation? The aim of my book is to go beyond traditional debates about whether judicial review is undemocratic and re-think the problem from the ground up, from the perspective of constitutional design and comparative analysis of institutional capacity. While I am favourable to a system of legislative supremacy such as that of the United Kingdom—as some of the commenters have noted—I acknowledge that there are reasons why constitutional framers might want to create an institution to check the power of an elected legislature against the standards of a written constitution. This could be in in the aftermath of a totalitarian regime or other circumstances where electoral politics cannot be trusted. However—and this is the central point of my argument—constitutional framers must do what they can to ensure, in any system, that a reviewing institution has the basic capacities necessary to carry out the task that it has been assigned.

Professor Delaney, while seeming to accept much of the book’s argument about comparative institutional capacity of courts and legislatures, raises several points for further development of the argument. While they are all valid considerations, there is a sense in which they sidestep the above-outlined structure of the book’s argument. Delaney suggests that an overall assessment of the merits of judicial review of legislation in the US must consider the comparative error rate of courts and legislatures in producing outcomes, the propensity of the US Supreme Court to follow the political majority in the long run, and related matters. She concludes that ‘[i]f the outcome-based comparative analysis is a toss-up, one might come down on the side of judicial review, if only to preserve the possibility for those rare but serious occasions when the political process breaks down, or legislative pathologies excessively burden rights.’

Nothing in my book is inconsistent with reaching this conclusion; but the aim of the book was not to assess a comprehensive balance sheet of the pros or cons of having judicial review. If I were to conduct such an inquiry, I would add items to Delaney’s useful catalog. These would include the knock-on effects for the political process of empowering nonelected judges to decide basic questions of political morality. For example, what is the cost, in the US, of the capture of the presidential election process by the judicial politics? Judicial review creates a feedback loop in which many voters make choices in presidential elections solely or primarily on the basis of a candidate’s prospective judicial appointments.

Instead of conducting such a comprehensive balancing exercise and applying it to the US or any other country, my book is aimed to inform constitutional framers about matters they should take into account in deciding whether to adopt a system of judicial review (and if so in what form) or to reform an existing system. The argument is equally intended for constitutional theorists and others responsible for influencing the structural frame of constitutional law and its ongoing development. I argue that, according to the criteria in the core chapters of the book, the approach in some common law countries of assigning to all ordinary courts the power to annul legislation on the basis of constitutional rights is one that fails badly. Even supreme courts in common law systems, which are designed as courts of appeal for points of law, fundamentally lack the capacities needed to engage in the kind of moral and empirical reasoning commonly used in today’s constitutional rights cases. While the book devotes relatively more time to discussing examples from the US, since it has the longest experience of judicial review, it is equally critical of the Canadian system and argues that courts in many other jurisdictions exhibit similar problems. Indeed, the basic structure of techniques of constitutional rights adjudication used throughout the world, as I show in chapter 2, invites courts to engage in moral and empirical reasoning beyond their competence. I argue, however, that the constitutional courts of European continental systems, typified by Germany, are relatively better designed than common law systems in a number of significant respects. These include having a centralized jurisdiction for constitutional review in a single court created for that purpose, a system of abstract rather than concrete review, and remedies more fit for purpose.

Thus, the argument of the book does not consist, as Delaney suggests, in moving from ‘an attack on one of the strongest systems of judicial review [i.e., the US, which Delaney calls an outlier] to generalizable lessons for constitutional designers’. The book is sensitive to many nuanced differences between systems and the argument picks out certain features as commendable and others as the reverse. For example, I argue in chapter 7 that the German constitutional court’s practice of engaging directly with experts—by calling them to open hearings or occasionally commissioning research—is better than the practice in common law courts of accessing empirical research through some combination of party submissions (Brandeis briefs) and the judges’ own independent research. I am grateful to Justice Wallis for noting that South Africa has a similar facility to that of the German court.

Delaney characterises my argument as asserting that legislatures have the ability to access a greater amount of information than courts, and that the book ‘elides the information-gathering ability of Congress (through special entities, such as the Government Accountability Office, the Congressional Budget Office, and the Congressional Research Service) with legislators’ capacity to review the data’. However, the ability to access relative amounts of information is not the point of my argument, and I warn against judges doing their own research on the Internet—a portal to an unlimited amount of information. The focus of my argument is on qualitative differences in the way that courts and legislatures engage with information. At pages 98-104, accessible through the sample chapter online, I discuss the dynamics through which legislators inform themselves and employ empirical data in lawmaking through specialised committees, interacting with constituents, pooling their own knowledge from diverse backgrounds, and accessing empirical research and other expert information. On the latter point a key factor is that legislatures can actively engage with experts, by holding hearings, examining witnesses, commissioning investigations, and asking follow-up questions. Moreover, legislators can utilise employees and agents who are themselves trained in research methods to interface with experts. Lastly, because legislatures act on their own motion and do not have to wait for a case to be brought, they can correct mistakes and respond to changing understandings of facts.

Delaney says that my claim that ‘legislators are better placed to review empirical work is less persuasive, in part because [Yowell] compares actual judges (and their mistakes) with an idealized version of the U.S. Congress.’ Delaney’s focus is on the wrong point of comparison. It is true that the book dwells at length on judicial mistakes in using empirical research—something I deemed necessary to highlight a real but often overlooked problem. Another reason I offer case studies is to show that there is a widespread problem in the basic method by which courts engage with and think about empirical research. This analysis is not simply critical of courts or judges; it follows a detailed exposition of why courts are well structured to assess and review adjudicative facts. I contrast the basic method by which courts engage with general empirical research—in particular, the legislative facts about the need for a law or its anticipated consequences—with the method entailed, or at least made possible by, legislative structure. Professor Vermeule, focusing on this level of comparison, has a different assessment from Delaney; he says, ‘There is no trace here of a standard professional deformation in conversations about judicial review — Harold Demsetz’ “nirvana fallacy,” which takes the form of comparing idealized courts with a realistically jaundiced view of legislatures, or for that matter the reverse (although the latter case is far less common among the knowledge class).’ Nothing in my argument is blind to the possibility of legislatures making mistakes, and nothing entails a prediction that courts will make more mistakes than legislatures. I acknowledge at p.126 that in a series of important cases in which the Supreme Court upheld laws based on mistakes, the legislature shared the greater part of the blame. My argument boils down to the simple point that courts, especially common law courts, have deeply inadequate structures for acquiring and assessing general empirical research. Both Justice Wallis and Justice Sarmas make the same point in their contributions to this symposium. If we want to assign courts a task of judicial review that involves empirical research, we should take steps to make courts more like legislatures in this respect.

The particular reform I suggest, building on the work of Kenneth Davis, is to provide supreme and constitutional courts with a dedicated research service. Professor Ekins is not convinced of this proposal for reform, stating, ‘I fear this would simply empower courts; their dependence on the research of the parties, and especially on the government, may be (and should be) an important limit on their freedom of action’. It is true that a court backed by a dedicated research service might feel more emboldened to strike down legislation than one without such a facility. But I think that such a risk must be assumed if constitutional framers make the decision to give courts the power to decide the constitutional validity of law on grounds that include judicial finding of legislative facts. The book provides reasons to the constitutional designer for not giving courts such power in the first place. But if the power is thought to be needed, then courts would not be properly equipped if they were required to rely solely on party submissions (Brandeis briefs). As Justice Wallis notes, judges will feel compelled to go beyond party submissions and conduct their own research. In my estimation, that poses a greater risk of activism than a research service, which would have the effect of making judicial fact-finding more transparent and accountable than when judges conduct their own research.

Vermeule notes different kinds of trade-offs implicit in the arguments in the book, making the general point that they could have been improved by more dynamic analysis of relative risks or models of legal change. The trade-off labeled ‘sentinel effects vs. epistemic moral hazard’, considered in light of the point raised by Ekins in relation to a judicial research service, shows how complex an issue of constitutional design might be in a particular situation. It is possible that a judicial research service would lead a legislature to exercise less diligence in its own research, trusting the supreme court to catch mistakes (epistemic moral hazard). But it is also possible that anticipation of an enhanced form of review would induce higher-quality research by the legislature into the factual grounding for laws (sentinel effect). All of this, including Ekins’s fear that a research service might embolden a court too much, is relevant; but applying this trade-off requires the analyst to predict, among other things, the misuse of appropriate tools (e.g,. an overly active court vs a legislature under-utilising its fact-finding capacity). My book focusses on analysing the fit between the responsibilities assigned to courts and the tools fit for the task. I do not foreclose the anticipation of risks of misuse of those tools, and constitutional designers applying the book’s arguments in a given situation would do well to anticipate the dynamics Vermeule identifies. Using a hammer to drive in nails is better than a screwdriver, but one should not ignore the relatively greater risk of injury to thumbs.

Paul Yowell is a fellow of Oriel College and associate professor in the faculty of law. Previously, he was a lecturer at New College. He did his postgraduate studies at Oxford (DPhil, MPhil, BCL), having previously practiced law and studied in the US (JD, BA, Baylor University). He is also on the adjunct faculty of the University of Notre Dame. He is the author of Constitutional Rights and Constitutional Design: Moral and Empirical Reasoning in Judicial Review, co-author of Legislated Rights: Securing Human Rights Through Legislation and one of the editors of Lord Sumption and the Limits of the Law. He researches broadly in public law and legal theory, with particular interest in the separation of powers, Constitutional theory, comparative constitutional law, and human rights.