I Introduction

Gregory C Keating’s Reasonableness and Risk is an important and insightful contribution to tort theory. The book vindicates basic intuitions about fairness and interpersonal justice and the significance of harm within this area of law – and it does so in a compelling, rigorous, and sophisticated manner.

The account Keating offers is a distinctive view that rejects some of the central tenets of the economic analysis of tort law. At the same time, however, Keating rejects the formalism of Kantian theorists like Weinrib and Ripstein.^[1] As he argues against them, ‘we cannot understand or justify the law of torts without attending to the interests that it protects.’^[2] In this commentary, I want to focus specifically on the question of what separates Keating from the Kantian formalism of Weinrib and Ripstein. I will do so by concentrating, particularly, on Chapter 3 of Reasonableness and Risk, entitled ‘The Importance of Interests’.

As I will argue, the main distinction between Keating and Kantian formalists – despite their shared rejection of the welfarist interpretation of tort law defended or, more lately, simply assumed by legal economists – is that Keating is an instrumentalist about legal institutions. This difference, which has already been noted by Martin Stone,^[3] is important. Keating’s instrumentalism makes his theory more plausible, to my mind, than the interpretation of tort law in terms of corrective justice, Kantian right, or mutual independence. But it also makes the central features of tort law, and of the common law more broadly, subject to contingency: once we adopt an instrumentalist perspective, it is always an open factual question whether the rules, doctrines, and institutions we have are justified.

II Reasonableness and risk and Kantian formalism

Keating makes two central claims in Chapter 3 of Reasonableness and Risk. The first claim is that rights protect interests. Thus, one cannot understand or rationally vindicate the rights protected by tort law unless one can ground them in some important human interests that the legal institution is trying to protect. And from this first philosophical point follows a second one: tort law is just one of the many legal institutions (both public and private) that protect these important human interests. In this sense, Keating is an instrumentalist.^[4] From his perspective, and as he puts it, ‘legal institutions are instruments for protecting certain interests and realizing certain val22ues.’^[5] Tort law is just one more tool in a larger toolbox, a specific legal institution situated within a larger ‘web of institutions,’^[6] some of which are part of what we would call private law, and some of which are part of what we would call public law.

In both respects, Keating is at odds with the views of Kantian formalists. For Arthur Ripstein, for example, the notion that grounds the rights recognized and enforced in tort law is the simple idea of mutual independence: no person is in charge of any other.^[7] Interests play no role in grounding rights. Of course, Ripstein – or any Kantian theorist, for that matter – would not deny that tort law protects valuable human interests. But these functions performed by tort law are not the fundamental idea that underlies tort law and that explains the particular rights it recognizes. The rights of tort law are simply the rights that derive from the idea of mutual independence – interests play no independent role in explaining those rights. From this starting point, it follows that the rights protected by tort law are not merely instruments to protect interests. They are, instead, the institutional instantiation of an abstract idea that gives them their constitutive explanation.

Keating has quite plausible worries about whether this type of view can make sense – as an interpretive matter – of the common law of torts.^[8] But he also has deeper philosophical concerns. For Ripstein, the law of torts concerns itself with the particular interaction between tortfeasor and victim. Its role is determining the conditions under which purposive beings can coexist with each other as mutually independent.^[9] This means that there is a strong division of labour between private law institutions like the law of torts and public law regimes focused on questions of distributive or ‘background’ justice.^[10] In response, Keating argues that tort law and the institutional domains of distributive justice have complementary roles: tort law is concerned with harmful interpersonal violations of rights, while distributive justice is concerned with what those rights are.^[11] There is still a distinction between tort law and the institutions concerned with distributive justice more directly. But the distinction is not premised on a strong notion that tort law is only the institutional instantiation of the abstract idea of mutual independence and does not concern itself with distributive justice. The distinction is merely that tort law deals with a particular upshot of the rights determined by distributive justice – namely, ‘what 23those rights require in the way of coercively enforceable reciprocal obligations.’^[12] And, again, what those rights are turns, for Keating, on which urgent human interests the law ought to recognize.

On Keating’s view, thus, legal rights specify and determine the interpersonal demands that arise from specific human interests. They are ‘institutional instruments’ that serve those interests.^[13] The upshot is that understanding tort law requires understanding which interests this body of law selects for protection – and morally evaluating and justifying tort law would similarly require a moral theory about which interests ought to be selected. One way in which we can interpret the project of Reasonableness and Risk is to read it as a historically sensitive account that answers these two questions with regards to contemporary tort law. Once we engage in this project, Keating argues, it is easy to see that only some torts – paradigmatically, what Keating calls sovereignty-based torts^[14] – are directly concerned with mutual independence as a distinct moral idea. For many others (in fact, for most torts), in contrast, we need to appeal to interests beyond mutual independence in order to offer both an interpretation of contemporary tort law and to articulate a justificatory theory.

I agree with much of Keating’s position, and I think it offers a compelling conception of tort law. In the remainder of this commentary, though, I will explore how the view’s articulation of the relationship between moral interests, moral rights, and legal rights, as well as its instrumental conception of legal institutions, perhaps have more radical implications than those that Keating would want to accept.

III Moral interests and legal rights

In an important section of his book, Keating argues that, on his view, the legal rights that courts are typically concerned with when they decide tort disputes aim at protecting a set of underlying moral interests. More specifically, as he writes:

1. ‘A claim of moral right asserts that some interest is important enough to warrant legal protection.’^[15]

2. ‘Legal recognition of a right imposes constraints on others in service of that interest.’^[16]

3. 24‘The protection that legal rights confer on interests creates some gap between the interest and the right.’^[17]

This third point is crucial. It is a recognition of the fact that the institutional protection of moral interests through legal rights is not always going to produce the exact outcome that would be produced if we were reasoning from the interest directly.

Nevertheless, Keating argues that we should not ‘make too much’ of the gap between legal rights and the underlying interests they attempt to protect. There is usually – though not always – a harmonious relationship between rights and interests.^[18] When there is no harmony, and we feel the conflicting pull of legal rights and the interests they serve – that is, in hard cases – ‘reasons rooted in the interests that justify legal rights guide the application and articulation of the rights that they justify.’^[19] Thus, there is a potential gap here. But this potential gap can be navigated by conscientious and capable judges, who can be aware of potential mismatches between interests and legal rights, and can therefore articulate and apply the relevant legal right in hard cases with due consideration to the underlying interest.

I am less certain about judges’ ability to successfully navigate hard cases in this way. In order to see why, allow me to reconstruct the rational structure that goes from the basic interests that – on Keating’s view – ground the whole edifice of tort law to the decision of particular tort disputes. That structure is something like this:

Interests → Moral rights → Legal rights → Decisions

The first transition, from interests to moral rights, strikes me as simple – at least in the following sense. We can have substantive moral uncertainty and disagreement over which interests qualify to be singled out for special recognition and protection. But once we have agreed on an answer to that question, then the relevant moral right flows relatively naturally from the moral argument.^[20] And, at least at a certain level, we can see that, in fact, different moral theories and traditions evince some 25level of convergence about the most important human interests and the moral rights that they generate.^[21]

Yet the two later transitions – from moral to legal rights, and from legal rights to decisions in particular cases – become a little more complicated.

Consider first the transition from moral to legal rights. Even if ultimately grounded in moral rights and in the interests they recognise, in a more immediate sense the content, structure, and scope of legal rights is determined at least in part by legal materials – constitutions, statutes, regulations, and precedents. There are two potential problems that arise here. First, the relevant decision-makers – legislators, regulators, judges, etc – might not structure the legal right adequately in light of the underlying moral right. The legal rights recognized by our legal institutions might track very imperfectly the underlying moral right given the epistemic limitations of actual legal officials. This is always a real possibility, and it is at least thinkable (and I would add, highly likely) that whatever legal materials we happen to have will fail to perfectly track the demands of justice. Second, and more importantly, the adequate design of these legal materials is supposed to take into account a host of considerations beyond the recognized moral right. To name a few obvious considerations: reasons pertaining to the effective guidance of conduct; reasons pertaining to sociological and psychological facts about the agents whose conduct will be guided; reasons about the epistemic, institutional, and economic constraints faced by those supposed to enforce the relevant legal rights; reasons about spillover effects regarding other legal regimes; questions about distribution, cost-effectiveness, and optimal incentives; etc. This means that the all things considered optimal design of the legal right might in fact need to diverge from the underlying moral right. In other words, even a perfectly rational designer with the appropriate aims must design legal rights that will diverge from the underlying moral right. The underlying moral right is only one of the morally relevant considerations that bear on the design of morally adequate legal rights, which means that the gap might, in fact, be much more systematic than what Keating seems inclined to allow – and for perfectly legitimate reasons that ought to be taken into account by a morally legitimate institutional design.

Consider now the transition from legal rights to specific case outcomes. Legal rights are set up, generally, by general legal rules derived from cases – or, more typically, lines of cases – as well as statutes and administrative regulatory schemes. In contemporary legal systems, and particularly common law systems, these general rules do not really behave according to what Rawls called the practice conception – ie, they are not followed in every single instance, even when their results are 26thought to be evidently suboptimal.^[22] Legal rules are, in most legal systems, defeasible. But crucially, rules also do not behave according to the summary conception – they are not treated as mere rules of thumb that summarize the underlying values or goals they aim to achieve, and that therefore are discarded in every instance they produce less than optimal results. Instead, rules are resilient and treated as decision criteria even when they can sometimes be over- or under-inclusive and produce outcomes that are not optimal in light of the underlying goals.^[23] To put the point in terms of Keating’s theory: legal rights sometimes lead to outcomes that are, from the perspective of the very interests the legal rights are supposed to track, suboptimal.^[24] This comes with the territory when regulating behaviour and deciding disputes on the basis of general rules. Unless we completely do without the rules or deliberately ignore them, sometimes the conscientious and rational decision-maker will reach decisions that are not optimal from the perspective of the moral right the relevant legal rule is attempting to vindicate.

I should note this is not, to my mind, necessarily morally problematic. The relevant moral question to ask about legal institutions, in my view, is not whether they perfectly replicate an ideal moral template or achieve optimal results in every single case, but whether as institutions they produce, in the aggregate, a good set of moral results. In the terms of Keating’s tort theory, the question is whether the law of torts, in the aggregate, does a good job at protecting fundamental human interests – even if, by necessity, it might in many specific cases not do so. But that means, too, that the connection between the underlying interests and values and the content and structure of tort law might not be as neat and tidy as Keating’s argument at times seems to suggest.

IV Instrumentalism

And here we can come back to Kantian formalism. On the Kantian view, legal rights and judicial decisions are determinations of a more abstract idea of right – in Ripstein’s terms, the idea of mutual independence.^[25] This starting point entails two things. First, because the point of legal institutions and legal rights is not to act as 27the instruments of any underlying moral aims or values, questions of instrumental legal design like the ones I mentioned in the previous section are at most of secondary importance. Second, because legal institutions determine and specify the abstract demands of right, there is in fact significant leeway for them to do so in different ways. There is no determinate extra-legal template that laws ought to approximate as much as possible.^[26]

Both aspects of the Kantian view – the fact that legal institutions are not instruments and that the abstract demands of right are consistent with a variety of institutional arrangements – suggest that questions of optimal legal design and worries about the gap between moral rights and legal outcomes are not a problem for that view. More specifically, because a legal institution is not evaluated as an instrument for realising certain goals but rather as an instantiation of an abstract idea, multiple specific versions of that institution can all be equally justified as contingent embodiments of that idea.

Keating’s theory, however, is more susceptible to these worries. Not as much, certainly, as other theories that have a one-size-fits-all optimal template. For Keating does recognize that moral rights can be less determinate than legal rights, and that therefore moral demands set an agenda but not a perfectly determinate model. Still, Keating is an instrumentalist about legal institutions. The potential divergence between case outcomes, legal rights, and the moral interests they serve do matter as long as we recognize that, while indeterminate, moral interests and rights can be served by legal institutions in better and worse ways, that there are different ways in which legal institutions can be arranged, and that there will be trade-offs between securing the relevant moral interests and other morally legitimate considerations. Questions of optimal legal design, on a view like his, ought to loom large.

In my view, this is in fact a strength of Keating’s theory. The fact that his view lends itself to instrumentalist analysis leads me to think it offers a better and more attractive account of contemporary tort law than the corrective justice alternative. In substantive terms, I also think Keating’s Rawlsian theory about the values that underly this institution is more persuasive than other alternatives on the table, such as the civil recourse theory of Goldberg and Zipursky (even though the latter has many strengths, particularly as an interpretive matter).^[27] And Keating’s theory also has a crucial advantage when compared to both Kantian corrective justice accounts and civil recourse theory – an advantage that flows directly out of Keating’s instrumentalist view. That advantage is its ability to incorporate the economic analysis of law 28and other social scientific perspectives within the framework of its own theory. Any instrumentalist theory requires a theory of legal design, models that predict human behaviour in response to legal rules and institutions, and empirical information about changes brought about by changes in said rules and institutions. This means that the tools of economic and social scientific analysis have a role to play in Keating’s view in a way that they do not under other non-welfarist frameworks. In other words, Keating’s framework can incorporate in a direct and straightforward way the insights of legal economists and other social scientists even if his account of the substantive ends of tort institutions is quite different from welfare-maximisation or some other economic or utilitarian perspective. Indeed, the work of social scientists, and particularly of economists, becomes fundamental for a project that sees legal institutions as instruments. From this perspective, I would expect legal economists to be quite sympathetic, or at least open, to a view like Keating’s. Whatever its demerits as a normative theory about private law’s foundations, the economic analysis of law is the most complete, thorough, and rigorous tool we have for predicting the impact of legal rules and institutions on human behaviour.^[28] And Keating’s theory, as an instrumentalist one, requires taking account of the impact of legal rules on behaviour. This, I think, also suggests that Keating’s tort theory and law and economics might be much more consistent than what first impressions might suggest.^[29]

But, again, Keating’s instrumentalism has two implications that should at least make us less optimistic than he is about tort law. First, it would be surprising, to say the least, if the law of torts we happen to have – warts and all, subject to the contingency of history and politics, and so on – were the optimal legal regime for securing the protection of the moral interests that underlie it. The second implication, as I mentioned above, is that there might be a morally legitimate (instrumentally justified) gap, that might be quite large, between the legitimate structure and content of tort law and the underlying moral interests and rights it is serving. This introduces a possibility for a certain opacity in legal reasoning that Keating, whose admiration of the common law method is quite evident throughout the book, might not be that willing to take on board.

V 29Open empirical questions

There is, however, a response Keating (or, perhaps, someone who might want to interpret Keating in a less instrumentalist way than me) might offer here. I might be making too much of the idea of legal rights and institutions as instruments. While Keating does think of tort law as an institutional instrument, he still thinks that the rights structured by the institution simply ‘put flesh on the bones of the interests that they serve by specifying constraints on others.’^[30] And, if that is the case, then the design of tort law’s rights and the actual deliberation about how to decide particular cases should transparently involve questions about the underlying interests. As Keating puts it in the case of the right to privacy, the link between rights and the underlying interests is tight – so tight that usually we name the right by reference to the underlying interest.^[31] Perhaps Keating’s instrumentalism is compatible with the claim that the reasoning of common law courts ought to track the underlying moral interests as much as possible. There is no need for opacity or for a deliberate gap between what courts actually do and what they are, in theory, aiming at doing.

The issue here, though, is that it is an open empirical question whether a transparent form of legal reasoning will in fact be the best way to ensure we protect the underlying moral interests we want to. The facts in any particular legal system at any given point in time might be such that, in reality, the best way to ensure courts protect the underlying moral interests is by not engaging questions about them directly. In other words, it might be instrumentally justified to preserve a gap between legal rights and moral rights by design. As Kornhauser puts it (from a welfarist perspective):

It may be that social welfare is best promoted by the establishment of institutions in which government, or some set of public officials within the government, pursues a different aim. One might, for example, argue that a formalist system of adjudication would better promote social welfare than an adjudicatory system in which judges sought to maximize social welfare. Phrased differently, one ought to consider government as instrumental in the pursuit of social welfare; consequently, it may be wise to design institutions in which policy makers do not directly pursue the overall goal.^[32]

There is a long tradition of such arguments in, for example, contract theory. Schwartz and Scott, for example, argue that a set of formalist interpretive defaults might maximize the parties’ welfare better than a set of more expansive (and therefore more 30accurate) rules about contract interpretation.^[33] And I have made a similar argument about contract law adjudication in general: given the plural values that contract law ought to serve, a formalist approach to contract doctrine might be warranted.^[34]

Once one sees a legal institution as an instrument, this type of approach might always be warranted – whatever the underlying values or goals one wants to realize are. It all turns on what the facts happen to be at any given time and place. From this perspective, it would be a very happy coincidence if pursuing the vindication of moral interests directly and transparently in common law adjudication were the best way to vindicate them across the board. And I think Keating is ultimately committed to precisely that view, given his obvious commitment to the idea that different legal regimes might be adequate in different contexts. In this regard, at least, his account – though more complex and open to contingency – strikes me as preferrable to the Kantian’s non-instrumental account of the law of torts.