By: Michael L. Wells


In his contribution to Wake Forest Law School’s 2009 Symposium on the Restatement (Third) of Torts, Professor Kenneth Abraham starts with two propositions, one descriptive, the other normative.  The descriptive claim is that “tort law . . . is mature and largely stable,”[1] and  that “[o]ver time, the law of different states will converge.”[2] As  he points out, “The formation of the American Law Institute (“ALI”) itself, and the project of restating the law that the ALI . . . undertook” depends on these premises.[3]

I distrust the general notion that centralized rule making works better than law making dispersed over a large and culturally diverse nation like the United States.

The project of restating the law also depends on a normative premise, namely that  “mov[ing] that process along” is a worthy goal.[4] The idea that the private law should be the same in all fifty states certainly seems preferable to anyone who has coped with the difficulties of working with variations among states on issues of tort law.  Convergence not only avoids arbitrary differences in the way similarly situated litigants are treated, but also offers a solution to the many practical difficulties that arise when a given transaction crosses state lines.  In any event, adherence to this norm may be a matter of existential necessity for the ALI.  Otherwise, one is hard put to justify the resources that are expended in producing the Restatements.  The descriptive and normative claims are related in that the value of moving the process along is greater or lesser depending on how successful the project is likely to be.

Whether Professor Abraham subscribes to these propositions is not entirely clear from his article.  Since his project was to identify “stable divisions of authority” within the tort system, the premise of unity serves merely to highlight those divisions.[5] My aim here is not to find fault with anything in his article, but rather to examine the notion that tort law does and should move toward unity.  For convenience, I will call these descriptive and normative claims the “convergence thesis.”

Professor Abraham’s article addresses a problem for the descriptive side of this thesis.  He acknowledges that “[t]here are . . . areas of doctrine over which the states have what seems to be permanent or at least long-term disagreement.”[6] For the most part, he argues, these areas can be pigeonholed into three categories: “(1) developmental dead ends, (2) fundamental clashes of values, and (3) concerns about consistency of administration.”[7] If I understand Professor Abraham correctly, his point is that areas of disagreement do not necessarily pose a challenge to the general tendency toward convergence.  But Abraham seems to assure us that the divisions are discrete, limited, and themselves “stable.”  His point is well taken, for no general description of a system as large and complex as American private law can hope to account for its every feature.  Nonetheless, in this Article, I will raise questions about both the descriptive and normative claims of the convergence thesis.

I.  Transformations in Tort Law

As a description of tort doctrine, the notion that there is a basic unity across the states—except for a few stable divisions of authority on matters like market-share liability and the treatment of trespassers—seems to me to be correct over the very short term.  Over longer periods of time, however, tort law is more dynamic than Professor Abraham’s description suggests.  Rather than being “stable divisions of authority,” large areas of tort law are transformed, sometimes in subtle ways, and sometimes radically.  Compare, for example, the contents of the second edition of the Gregory and Kalven casebook, published in 1969,[8] with the ninth edition of the book, now edited by Richard Epstein and published in 2008.[9] Gregory and Kalven begin their casebook with causation, and the book contains no materials at all on either market-share liability or “increased risk,” theories of causation that have taken hold in some courts in the interim.[10] The 1969 edition of the Gregory and Kalven casebook was published a few years after theRestatement (Second) of Torts first addressed products liability in 1965.[11] Most of the cases deal with privity, the warranty-tort distinction, and the economic loss rule that obliges some litigants to sue in contract rather than tort.[12] Current casebooks contain extensive treatments of design defect, the duty to warn, the plaintiff’s conduct, federal preemption, and other matters.[13] These do not appear in the earlier edition of the book for the very good reason that none of these cases had been decided.  On the treatment of trespassers, the earlier book notes the “recent California case, Rowland v. Christian” for the proposition that the older rule “may now be in the process of change.”[14] The more recent edition includes Rowland as a main case and surveys the mixed reception the case received in the intervening four decades.[15] The Gregory and Kalven casebook includes a long chapter entitled “Ultimate Policy Issues” devoted mainly to no-fault schemes that would replace or supplement tort remedies with first-party insurance.[16] This idea was fashionable at the time, and New Zealand actually adopted such a scheme in 1972.[17] But no-fault schemes fell out of favor in the forty years between the publication of the two books.  Epstein’s chapter on policy focuses on specific insurance schemes, primarily workers’ compensation and no-fault car insurance.[18] The “tort reform” movement, which has generally involved state-by-state legislation, has produced another set of variations among jurisdictions.[19]

One could go on, but these examples will suffice for my purposes.  Indeed, noting the changes in tort law over the past forty years may seem not only to belabor the obvious, but also to miss Abraham’s point.  The ALI pays close attention to changes over time.  Those changes are the main reason for drafting the Restatement (Second) of Torts in the 1960s, and a third one today.  Defenders of the convergence thesis would no doubt charge that by noting changes in the law over time I have mixed apples and oranges.  It is clear that Professor Abraham means to refer to convergence at a given point in time and does not mean to deny that tort law changes significantly over time.  Given the extent of change over time, Abraham must mean to assert that change over time is different from, and compatible with, convergence at any given point in time.

But the distinction between change over time and variation at a given point in time is not as sharp as defenders of convergence may imagine it to be.  One reason is that in a system of private law that includes over fifty jurisdictions, each of them is sovereign over the matters of general tort law covered by the various Restatements.[20] In matters of common law, change comes to each state at a different pace. Change over time begins with one jurisdiction deciding to break with precedent, as the California Supreme Court did in Rowland with respect to landowner liability.  But other jurisdictions do not then consider the issue simultaneously or even in the same legislative session, and then either adopt it or not.  Rather, changes over time occur at an uneven pace from one jurisdiction to another.  Their courts take up the issue periodically over the ensuing decades, depending on whether and when an appropriate case raises it.  Even if the issue is one as to which the law ultimately does “work[] itself pure”[21]—as landowner liability does not—most of the time the law will differ from one place to another.

In addition, judicial opinions can be unclear on whether any change has taken place on a given matter and the substance of whatever change that may have occurred.  Consider the design-defect issue discussed by Abraham.[22] In one view, the Georgia Supreme Court has endorsed the “reasonable alternative design” test of the Restatement (Third) on products liability.[23] But the case also says that “[u]nder Georgia law, the proper analysis in a design defect case is to balance the risks inherent in a product design against the utility of the product so designed,”[24] and refers to the court’s previous “adoption of the risk-utility analysis” in its leading decision on the topic.[25] A prudent lawyer would conclude that exactly what Georgia has done remains an open question, though it surely has rejected the “consumer expectations” test that Abraham discusses.[26]

If all of the “changes in tort doctrine over time” had already been set in motion by, say, August 1, 2010, the process of change as it moves from one jurisdiction to another would be a temporary problem for the convergence thesis.  But no one would endorse that premise, as it is wholly unrealistic to suppose that the political, social, technological, and cultural factors that produce changes in the common law have now come to an end.  Perhaps Professor Abraham means to predict that the only political, social, technological, and cultural changes that will affect tort law going forward are discrete ones that will lead to more examples of the narrow and “stable divisions of authority” that he identifies.  Lacking a crystal ball, I cannot disprove that prediction, but I would suggest that it is ahistorical and implausible.  Someone making a similar prediction about products liability in 1970 would have been proven wrong by the exponential growth of products law that began in that decade.  Nor did anyone foresee the growth of mass tort litigation and the doctrinal innovations it would spawn.[27] Developments of similar scope surely await us, and with them will come less convergence, followed by sporadic, uncertain, and incomplete moves toward unity, followed by more innovations and renewed disparity. The ALI will continue to struggle to keep up.

II.  The Benefits of Convergence in Tort Law

The descriptive question raised by the plausible “at a given moment in time” version of the convergence thesis is not binary, but rather a matter of degree—not whether there is convergence at a given moment in time, but how much convergence we have attained. Answering this question is not just a matter of counting up doctrines and splits of authority.  It necessarily has a qualitative aspect: one must ask whether the areas in which courts disagree are especially significant or not.  Reasonable people can differ about the answer.

The more interesting question for me is the normative one: should we make an effort to achieve greater convergence, through Restatements and other means?  The first point that needs to be made about this question is that our system of private law is not a unitary one in which the Supreme Court of the United States exercises final authority over all tort law.  In such a system, the case for promoting convergence would be compelling.  Allowing variations within a system to persist indefinitely would be intolerable, as persons governed by the same set of legal rules would be treated differently.  At the time of the first Restatement of Torts in the 1930s,[28] it was possible to conceive of our system in this way, or to imagine it moving in this direction, for the federal courts were free, under Swift v. Tyson,[29] to follow common law rules of their own choosing.  But “the benefits expected to flow from the rule did not accrue,”[30] and the Court eventually overruled Swift in Erie R.R. Co. v. Tompkins.[31] Rather than opting for a unitary body of private law, Erie signaled that our federal system treats the great bulk of private law, including torts, as state law.[32] Erie not only put to rest Swift’s futile dream of a nationally uniform common law but also declared that “there is no federal general common law.”[33]

After Erie, arguments for and against convergence boil down to the costs and benefits of variety versus unity in the common law. Variations among states certainly add a layer of complexity and uncertainty to business planning and to litigation that crosses state lines, as issues may arise as to which state’s law applies to a given issue.  A whole body of law on “conflicts of law,” itself important enough to warrant a law school course on the topic, addresses those issues.  Litigants who seem to be similarly situated may be treated differently on account of the operation of conflicts principles.[34] Convergence would lessen the incidence of that kind of unfairness.

A quite distinct argument for convergence concentrates on the quality of the rules themselves.  It is plausible to believe that a centralized decision maker, like the ALI, will construct better rules than a myriad of state courts.  The ALI has access to all of the decisions of all of the state courts, as well as the expertise of leading judges, lawyers, and academics.  The Reporters and Advisory Committees can take their time and are obliged to shed their partisan concerns.  Under these conditions, issues can be addressed deliberately, disinterestedly, and free from the distractions of litigation.[35] The aim of improving the law through Restatements can only be achieved to the extent courts everywhere endorse convergence and follow the Restatement formulation of the doctrine. Consider, for example, Professor Abraham’s discussion of the diverse views among states concerning trespasser liability.[36] One implication of the “superior rule maker” argument for convergence is that the forty-one states that have rejected the Restatement (Second)’srule on trespassers should change their law, just because a superior decision maker has made a different choice.  As for Professor Abraham, his discussion of the topic carefully avoids taking sides.

On the other hand, an array of arguments can be marshaled on the opposing side:

•  One is entitled to doubt the superior wisdom of the ALI Reporters and Advisory Committees.  Some have charged that the drafting of modern Restatements may place an “extraordinary burden” on the participants “of reconciling the demands of law and politics.”[37]

•  Even if political compromises can be avoided, it is reasonable to think that calm and detachment are not necessarily virtues.  Better rules may be made in the heat of conflict, under pressure to reach conclusions that have significant impact on real litigants.

•  Variations among jurisdictions permit us to test the impact of one rule compared to another over time and better evaluate its quality.

•  In a big country, the social and cultural underpinnings of tort law may differ from one place to another.  The nonflagrant trespasser may be viewed as a more serious problem in a rural state than an urban one, such that a “no recovery” rule is appropriate in the former while the Restatement (Third)’s formulation is appropriate for the latter.

•  The federal structure of our system of government reflects a preference for localized over centralized decision making on matters that neither the Constitution nor Congress have addressed.  The point of Erie was not merely to do away with Swift.  Granting that there is a case for convergence, Erie reflects the value that our federal system places on the benefits of variety over those of unity.

•  In a centralized system, one cannot easily avoid a rule one finds repugnant.  When the issue is governed by state law, one can move away.

For my part, this latter set of arguments carries the day.  While the case for convergence has merit, I distrust the general notion that centralized rule making works better than law making dispersed over a large and culturally diverse nation like the United States.  With regards to tort law in particular, it is, as it was eighty years ago, “one of the most dynamic fields of government.”[38] It seems to me unwise to aim for convergence in such an area, even if that aim could be achieved.

[1].   Kenneth S. Abraham, Stable Divisions of Authority, 44 Wake Forest L. Rev. 963, 963 (2009).

[2].   Id. at 964.

[3].   Id. at 963.

[4].   Id.

[5].   Id. at 964.

[6].   Id.

[7].   Id.

[8].   Charles O. Gregory & Harry Kalven, Jr., Cases and Materials on Torts (2d ed. 1969).

[9].   Richard A. Epstein, Cases and Materials on Torts (9th ed. 2008).

[10].   See Gregory & Kalven, supra note 8, at 5–20.

[11].   See Restatement (Second) of Torts § 391 (1965).

[12].   See Gregory & Kalven, supra note 8, at 584–631.

[13].   See Epstein, supra note 9, at 723–852.

[14].   Gregory & Kalven, supra note 8, at 404 (discussing Rowland v. Christian, 443 P.2d 561 (Cal. 1968)).

[15].   See Epstein, supra note 9, at 593–606.

[16].   Gregory & Kalven, supra note 8, at 782–910.

[17].   For a description and critique of the plan, see generally James A. Henderson, The New Zealand Accident Compensation Reform, 48 U. Chi. L. Rev. 781 (1981).

[18].   See Epstein, supra note 9, at 961–1012.  The New Zealand plan receives just six pages at the end of the chapter.  See id. at 1012–18.

[19].   See Epstein, supra note 9, at 419–20 (discussing statutory modifications of joint and several liability).

[20].   See Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78 (1938).

[21].   Abraham, supra note 1, at 963 (quoting Omychund v. Barker, (1744) 26 Eng. Rep. 15 (Ch.) 23).

[22].   Id. at 967–70.

[23].   See James A. Henderson, Jr. & Aaron D. Twerski, Products Liability: Problems and Process 185(6th ed. 2008) (citing Jones v. NordicTrack, Inc.,274 Ga. 115, 118 (2001), for the proposition that under Georgia risk-utility standards “the ‘heart’ of a design defect case is the reasonableness of selecting from among alternative product designs and adopting the safest feasible one”).

[24].   Jones v. Nordictrack, Inc., 550 S.E.2d 101, 102 (Ga. 2001).

[25].   Id. at 103 (citing Banks v. ICI Ams., 450 S.E.2d 671, 674 (Ga. 1994)).

[26].   Abraham, supra note 1, at 969–70.

[27].   Many of these innovations are examined in Richard A. Nagareda, The Law of Class Actions and Other Aggregate Litigation (2009).

[28].   See 4 Restatement of Torts (1939); 3 Restatement of Torts (1938); 2 Restatement of Torts: Negligence (1934); 1 Restatement of Torts: Intentional Harms to Persons, Lands, and Chattels (1934).

[29].   41 U.S. 1, 19 (1842), overruled by Erie R.R. Co. v. Tompkins, 304 U.S. 64 (1938).

[30].   Erie, 304 U.S. at 74.

[31].   Id. at 79–80.

[32].   See Henry M. Hart, Jr., The Relations Between State and Federal Law, 54 Colum. L. Rev. 489, 504–05 (1954).

[33].   Erie, 304 U.S. at 78.

[34].   For example, suppose that a North Carolina resident and a Georgia resident are both driving while texting and collide on a roadway somewhere.  Each sues the other for negligence and each raises a contributory negligence defense.  Both are found negligent as defendants and contributorily negligent as plaintiffs.  Assume that under the conflicts rule of the jurisdiction where the suits are brought the plaintiff’s fault is governed by his residency.  The result is that the Georgian will have his recovery reduced under Georgia’s comparative negligence rule, while the North Carolinian will recover nothing under North Carolina’s absolute defense.  See Ga. Code Ann. § 51-12-33(a) (2010) (“Where an action is brought against one or more persons for injury to person or property and the plaintiff is to some degree responsible for the injury or damages claimed, the trier of fact, in its determination of the total amount of damages to be awarded, if any, shall determine the percentage of fault of the plaintiff and the judge shall reduce the amount of damages otherwise awarded to the plaintiff in proportion to his or her percentage of fault.”); Sorrells v. M.Y.B. Hospitality Ventures of Asheville, 332 N.C. 645, 648, 423 S.E.2d 72, 73–74 (1992) (holding that a finding of contributory negligence bars recovery from a defendant for acts of ordinary negligence).

[35].   See Catharine Pierce Wells, A Pragmatic Approach to Improving Tort Law, 54 Vand. L. Rev. 1447, 1448 (2001) (describing the aspirations of the ALI as it undertook the process of restating the law).

[36].   See Abraham, supra note 1, at 967–68.

[37].   Marshall S. Shapo, In Search of the Law of Products Liability: The ALI Restatement Project, 48 Vand. L. Rev. 631, 635 (1995).

[38].   Leon Green, The Torts Restatement, 29 Ill. L. Rev. 582, 585 (1935).