For over twenty years, I have exposed my students to the conceptual vacuity of the “substantial factor test” for factual (and, for that matter, proximate) causation. Most torts scholars I know do the same. I explain that although many jurisdictions still purport to use a substantial factor test, even such jurisdictions typically apply the test as a combination of the but-for test, the overdetermined cause rule, and a couple of evidentiary mechanisms that aid plaintiffs in cases in which the evidence to prove causation is thin or unavailable. I go so far as to exhort my students that: “The words ‘substantial factor’ are useless and should not appear on your exams!” And yet inevitably, each year several students claim that the misdeeds of my exam’s hapless defendants were a substantial factor in causing the plaintiff’s harm. Sigh.
On researching the current state of the law in this area, the reason for this teaching disconnect has become clear to me. Despite the valiant attempts of many professors to illuminate the “insubstantiality of the substantial factor test,” many treatises and study aids dutifully report the test’s continuing salience in the courts. Indeed, an overwhelming majority of jurisdictions still use the substantial factor language as a pillar of factual-causation reasoning. Moreover, students are intuitively drawn to the test like moths to a flame—the test invitingly glosses over the complex analysis demanded by factual causation problems and reduces the analysis to a mere gestalt.
Although the First and Second Restatements embrace versions of the substantial factor test, the Restatement Third of Torts has brought a welcome reprieve. The Third Restatement disclaims the substantial factor test, restating factual causation as a combination of the “but-for” and “multiple-sufficient-cause/causal sets” tests, adding to these tests the burden-shifting framework of the iconic case of Summers v. Tice. As Section 3 of this paper will elucidate, the Third Restatement is less definitive on a couple of thorny causation problems, addressing them in comments rather than black-letter sections and leaving some doubt as to their proper place in the causation architecture. Nevertheless, the Third Restatement’s approach offers significant promise towards conceptual clarity, the rule of law, and substantive justice in tort law.
Although the relevant portion of the Restatement was published in 2010, however, courts have not exactly jumped wholeheartedly on the bandwagon. As of this writing, for example, courts have only cited the Restatement’s basic but-for causation section, § 26, a relatively meager 63 times; § 27 on Multiple Sufficient Causes, 42 times; and § 28 on the Burden of Proof, 21 times. Moreover, as mentioned above, courts remain stubbornly wedded to versions of the substantial factor test.
In the midst of this landscape, the 2021 case of Doull v. Foster, out of the Massachusetts Supreme Judicial Court, appears as a breath of fresh air. Doull embraces the Restatement Third’s rejection of the substantial factor test and affirmatively traces a coherent, lucid approach to factual causation that is capable of being applied by juries. For this bold (if, to many law professors, obvious) advancement, this case deserves a place in the modern canon.
In this essay, I will detail the current legal landscape of the substantial factor test for factual causation, describe the accomplishments and modest shortcomings of the Restatement Third’s approach, and highlight the conceptual and pragmatic benefits of this approach, as adopted by Doull.
2 The Substantial Factor Test
The historical development of the substantial factor test, as well as its path toward widespread use and misuse, has already been well documented by legal scholars. A brief summary, with observations, is nonetheless warranted.
According to the Third Restatement, the “substantial factor” concept was conceived as an aspect of scope of liability (proximate cause)—not factual cause—by Jeremiah Smith, a Harvard Law Professor, in a 1911 law review article. In that article, Smith seemingly used the term as a proxy for a “superseding cause” analysis, explaining that an initial wrongdoing might not be a substantial factor in causing a plaintiff’s harm if there is a subsequent, intervening wrongdoing. Smith emphasized that this operation is not axiomatic—rather, whether liability attaches to the initial wrongdoing turns on its “substantiality,” a judgment to be determined in the context of “legal causation.”
The Third Restatement asserts that the first application of the substantial factor concept to factual causation was in Anderson v. St. Paul & Sault Ste. Marie. Ry. Co., a case involving two fires that joined and together burned the plaintiff’s property. In that case, the court affirmed the trial court’s jury instruction that: “If you find that bog fire was set by defendant’s engine, and that some greater fire swept over it before it reached plaintiff’s land, then it will be for you to determine whether the bog fire * * * was a material or substantial factor in causing plaintiff’s damage. If it was, defendant was liable. If it was not, defendant was not liable.” Because the court’s opinion did not address whether the “bog fire” was independently sufficient to burn plaintiff’s property, the court’s use of the phrase “substantial factor” served three possible functions: (1) to open the possibility for a finding of causation in an overdetermined-cause scenario, in which the but-for test would fail, (2) to impose a limitation on such a conclusion should the jury find that the defendant’s negligence was nevertheless insubstantial, or (3) to allow a jury to find that defendant’s -negligence was substantial, and therefore a factual cause, even if it was not independently sufficient. Each of these possibilities would later emerge in courts’ widespread use of versions of the substantial factor test.
The similarity between these two early uses of the substantial factor concept is striking. Both consider the problem posed by cases involving initial and subsequent tortfeasors, and both admit the possibility that the influence of the latter might be deemed to have “swept over” that of the former, so as to render the former “insubstantial.” While Smith might have considered this to be a scope of liability issue, however, the Minnesota Supreme Court felt it to be a matter of factual causation. As a preview to Part II of this paper, I propose that despite its forceful rejection of the substantial factor test, the Third Restatement retains a version of this core idea and sides with Smith’s conception that it is best treated as an aspect of scope of liability.
In the fourteen years between Anderson and the publication of the First Restatement, the substantial factor concept had not yet gained a pervasive hold in the case law. Nevertheless, § 431 of the First Restatement states, as the basic test for causation, that: “The actor’s negligent conduct is a legal cause of harm to another if (a) his conduct is a substantial factor in bringing about the harm.” As comment a explains, the First Restatement’s substantial factor test is a requirement in addition to but-for causation—that is, a plaintiff must prove not only that defendant’s tort was a but-for cause, but also a substantial factor of plaintiff’s injuries. Comment a further explains that the term “substantial factor” should be interpreted “in the popular sense in which there always lurks the idea of responsibility.” This statement echoes the gist of both Smith and Anderson in the sense that it conceives of substantial factor as a normative judgment rather than a factual finding, and one that is not susceptible to definition more specific than popular intuition.
Section 432 of the First Restatement states that tortious conduct is not a substantial factor if it fails the but-for test, and that multiple sufficient causes “may be held by the jury” to be substantial factors—presumably only if the jury determines that each is “substantial,” although the comments to the section do not address the issue.
Finally, § 433 of the First Restatement sets out “Considerations Important in Determining Whether Negligent Conduct is a Substantial Factor in Producing Harm.” This is where the first Restatement’s treatment of substantial factor becomes not only nebulous but affirmatively confusing, because the factors listed in § 433 are central to courts’ discussion of proximate causation, not factual causation. As I will discuss shortly, this confusion continues to reverberate throughout the case law.
The Second Restatement essentially repeats the First Restatement’s sections on causation, adding little further guidance. Its reaffirmation of substantial factor no doubt had an important consolidating effect, however, leading to the concept’s snowballing influence on the common law.
This brief summary of the early development of the substantial factor test yields three observations, none of which is particularly novel, but each of which sheds light on current case law and on the Third Restatement’s distinct departure from its predecessors. The first observation is that causation was viewed as a unified inquiry—not just in a rhetorical way, but as a concept and doctrine. “Legal causation” was not merely a combination of two distinct and unrelated elements of factual causation and scope of liability. Rather, these strands were seen to be inevitably related and intertwined, perspectives on a single question.
A second observation is that this unified inquiry was seen as normative—not a purely factual question with a yes or no answer, but a judgment about responsibility. Moreover, the judgment was an exercise in line-drawing: the “substantiality” of a wrongdoing’s causal relationship with an injury exists on a continuum, and the jury’s task is to draw a line at which liability will attach.
The third observation is fundamental to the first two, and it is that a causal element’s substantiality was central to instincts about justice and responsibility. Although the first and Second Restatements recognize that there can be more than one tortious cause of an injury, the idea that some tortious causes are simply not “enough” to trigger liability was seen as central to a claim’s merit. This is consistent with courts’ broader reluctance to hold a defendant liable if another tortfeasor either acted in a more culpable manner or was more “directly” responsible for the plaintiff’s harm. This general instinct not only supported the substantial factor approach to factual causation, it also animated the continuing dominance of contributory negligence, courts’ reluctance to allow recovery for enabling torts, and strong limitations on affirmative duties to protect plaintiffs from third-party tortious conduct.
Each of these aspects of early Twentieth-Century tort law has since evolved, of course. Most courts now describe factual causation and scope of liability as distinct elements. Most courts at least ostensibly recognize that factual causation is, in most cases, a factual inquiry into what happened, not a normative judgment about responsibility. And particularly with the move toward comparative fault, courts have become increasingly more comfortable with holding defendants liable even if they played non-central roles in a plaintiff’s misfortune. As Section 2 of this article illustrates, the Third Restatement reflects each of these developments in its sections on factual causation generally, and specifically with its loud (if, I will argue, incomplete) rejection of the substantial factor test.
2.2 The Current State of the Law
Notwithstanding the Third Restatement’s pronouncement that the substantial-factor test has “not withstood the test of time,” versions of the concept remain stubbornly pervasive in the case law. By my count, 46 jurisdictions still use the language of substantial factor in discussing factual causation. (No doubt, this is at the root of my students’ insistence on applying the test despite my exhortations.) There is less consistency, however, in the precise manner in which courts use the phrase. The primary approaches are as follows:
Eleven of these jurisdictions essentially follow the first and Second Restatement approach to causation, requiring plaintiffs to prove both that plaintiff’s injury would not have occurred but for defendant’s tort and also that defendant’s conduct was a substantial factor in causing the plaintiff’s harm. Some of these jurisdictions employ the latter requirement as the test for scope of liability.
Four jurisdictions apply the substantial factor test as the sole test for factual causation in most cases. In these jurisdictions, the plaintiff need not establish that the defendant’s tortious conduct was a but-for cause, but merely that it was a substantial factor in causing the plaintiff’s harm. In some of these jurisdictions, the but-for test is applied in some special subsets of claims—for example, cases involving professional malpractice or fraudulent misrepresentation.
In 2 jurisdictions, courts refer to their test for factual causation by the moniker “substantial factor,” but apply it as a but-for inquiry in all (or nearly all) cases.
In 26 jurisdictions, some variant of the substantial factor test steps in only where the but-for test fails—most commonly, cases involving multiple independently sufficient tortfeasors. In some of these jurisdictions, such as California, courts refer to this entire approach as the “substantial factor test.” Others describe it as an “alternative test” or a combination of the but-for and substantial factor tests.
Regardless of its name, as will be seen in Section 2 below, this approach most closely approximates that of the Third Restatement, with two important distinctions. First, such jurisdictions use the “substantial factor” language, rather than the language of “multiple sufficient cause/causal sets” used in the Third Restatement. Second, not all of these jurisdictions limit application of the substantial-factor test to multiple sufficient cause/causal set scenarios. Many apply the test to any case involving multiple potential causes, regardless of whether there is evidence that each was independently sufficient or a necessary part of a sufficient causal set—for example, in concurrent cause cases or loss-of-chance claims.
A set of cases that embody both of these points of departure is those employing the “substantial contribution to the risk test”—a test often applied to medical malpractice claims or cases involving exposure to toxic substances such as asbestos.
In another 3 jurisdictions, the basic standard for factual causation remains frustratingly unclear. In Illinois, for example, courts often state that a plaintiff may prove factual causation by meeting either the but-for or substantial factor standards. On the other hand, some courts explain the substantial factor test in but-for terms. Other jurisdictions in which the applicable test is less than clear include Mississippi and North Carolina.
Only 5 jurisdictions eschew substantial-factor language (almost) altogether. Among these, only 2—Massachusetts and Virginia—have expressly rejected the substantial-factor test, each in favor of the Third Restatement approach.
2.3 Service and Disutility
The continued widespread use of the substantial-factor test no doubt reflects the powerful general influence of the first two Restatements. It also suggests, however, some inherent merit to the concept. Courts’ instinct that an “insubstantial” causal element should not trigger liability is loosely appealing. And the many ways in which a causal element might be “insubstantial” gives credit to the Restatements’ decision to leave the concept without definition more precise than that it should be understood “in the popular sense in which there always lurks the idea of responsibility.” For example, inter alia, a causal element might be insubstantial:
Because it is superseded by intervening events—for instance, Henry Ford might be considered an “insubstantial” cause of every modern car accident.
Because it contributed only a small amount to an injury, and in a way in which it is difficult to tell whether it was a but-for cause—for example, a factory that dumps a small amount of a toxic substance into a stream might be considered an “insubstantial” cause of an illness in a person who consumed water in which much larger amounts of toxin had been dumped by another factory.
Because its contribution is overshadowed by other, more culpable wrongdoers—for example, a driver’s negligence in exceeding the speed limit by 5 mph in a 70 mph zone might be considered an “insubstantial” cause of a collision with a drunk driver who crossed the center line.
Sorting out the precise conceptual differences between these examples, and the most coherent doctrinal way to address them, is no simple matter. One might argue that just as courts instruct the jury on standards such as reasonableness, foreseeability, or directness/remoteness without more precise definition, justice is best served by leaving the causal complexities captured by the phrase “substantial factor” to instinct and community more.
Nevertheless, popular opinion among those who study tort law has, for some time, been that we can do better. As this Section’s survey of the law reveals, courts’ treatment of the standard is all over the map, often internally inconsistent, and sometimes downright Kafkaesque. The concept of “substantiality” provides considerably less guidance even than reasonableness, foreseeability, or directness/remoteness. Each of those concepts at least provides a particularized core idea to which a jury might tether its line-drawing. In judging reasonableness, for example, a jury might engage in cost-benefit analysis, or it might draw on its own experience in risk-related decisionmaking. In judging foreseeability, a jury has its own life experiences regarding the reasonable likelihood of an event (or analogs to the event), as well as exposure gained through stories and empirical data. Even directness/remoteness places a jury’s reasoning on a spectrum of time, space, and a series of intervening causes. As flawed as each of these inquiries may be, “substantiality” ties causality only to the vague idea of “enoughness.” That is a vat into which a jury might pour a whole variety of intuitions about the merit of a plaintiff’s claim.
Furthermore, many of the issues to which courts have applied the substantial-factor standard are indeed more coherently decided in doctrinal arenas other than factual causation. For instance, Henry Ford (even were his actions tortious) would not be held liable for a car accident in 2023 not due to factual causation, but because such an accident is not within the scope of liability. If a factory’s but-for factual contribution to a polluted stream is minimal, perhaps the jury ought to assign the factory a small percentage of damages. And if minimal speeding is of a lesser category of wrongdoing than drunk driving, then perhaps courts ought to apply the common rule barring a jury’s comparison of fault in such a case. Put differently, many of the justice intuitions captured by the substantial-factor test are more precisely, more coherently, and less confusingly dealt with in other areas of tort doctrine.
Finally, courts’ use of the substantial factor test mischaracterizes the fundamental nature of factual causation. Causation is, as indicated by its name, a factual inquiry. Its fundamental query is into what happened, not whether the defendant ought to be held responsible. In most cases, there is a Platonic answer to this query—either the defendant’s tortious conduct caused plaintiff’s injury, or it did not. In most cases, factual causation does not call for a normative judgment, or indeed any line-drawing at all. By contrast, “substantiality” is not a fact. It is, as the first and Second Restatements explain, a judgment having to do with responsibility. To be sure, in a narrow set of cases, factual causation demands a normative decision about responsibility. There is no factual answer to the two-fires problem, only a moral and/or policy decision about whether to green-light causation. And in other cases, although there might be a factual answer to causation “in the ether,” courts point to reasons of policy or insurmountable evidentiary hurdles as reasons to depart from the standard causation requirements. Such exceptions, however, are narrowly-tailored and are born of limited philosophical or practical necessity. By contrast, the substantial factor test in its strongest version altogether diverts causation from its essential purpose, robs it of its essential nature. Thus untethered, it is not surprising that courts’ application of the standard is diffuse and confused.
All of this said, is there no kernel of value lying within the substantial-factor concept? Or were Third Restatement Reporters Bill Powers and Mike Green right to say that “‘substantial factor’ provides nothing of use in determining whether factual cause exists”? I address this question in the following Section.
2.4 The Third Restatement
In more ways than I can address in this essay, the causation sections of the Third Restatement add precision, clarity, and guidance to courts that will go a long way toward enhancing substantive justice and the rule of law in torts cases.
The same might be said about the Third Restatement’s treatment of substantial factor, in particular. In this regard, however, the Restatement’s approach leaves room for incremental improvement. Before turning to this argument, which incorporates a discussion of Doull v. Foster, I offer a brief overview of the Restatement’s factual causation architecture.
Section 26 of the Third Restatement, entitled “Factual Cause,” restates the but-for test. In relevant part, it provides: “Conduct is a factual cause of harm when the harm would not have occurred absent the conduct.” As previously discussed, Comment j to this section expressly rejects the substantial factor test on the grounds that it has not “withstood the test of time, as it has proved confusing and been misused.”
Section 27, entitled “Multiple Sufficient Causes,” provides an alternate means of establishing causation to the but-for test. It states: “If multiple acts occur, each of which under § 26 alone would have been a factual cause of the physical harm at the same time in the absence of the other act(s), each act is regarded as a factual cause of the harm.” Comment f to this section extends this rule to cases in which a plaintiff can show that each defendant’s contribution—although not an independently sufficient causal factor—was necessary to at least one sufficient causal set. Comment b to this section reiterates § 26’s rejection of the substantial factor test, eliminating any discretion the jury might have had under the First and Second Restatements “on normative or evaluative grounds to pick and choose among tortious acts that are independently (with other background circumstances) sufficient to cause the harm. …” 
Sections 26 and 27, on their face, appear to put a definitive nail in the substantial factor coffin. On closer look, however, such a conclusion might be precipitous. Recall, from Part I.B above, that there are two ways in which a substantial factor test might operate: (1) it might allow a finding of causation where a plaintiff cannot meet the but-for standard (or perhaps even the overdetermined-cause standard); and (2) it might preclude a finding of causation where a defendant’s causal contribution was insubstantial. The comments of the Third Restatement arguably breathe a bit of life back into each of these applications of the substantial factor concept.
The first use for substantial factor is indicated by Comment g to § 27, entitled “Toxic substances and disease.” This comment initially recognizes that the most common modern instance of multiple sufficient causal sets is in toxic tort claims. About such claims, the comment states the following:
When a person contracts a disease such as cancer, and sues multiple actors claiming that each provided some dose of a toxic substance that caused the disease, the question of the causal role of each defendant’s toxic substance arises. Assuming that there is some threshold dose sufficient to cause the disease, the person may have been exposed to doses in excess of the threshold before contracting the disease. Thus, some or all of the person’s exposures may not have been but-for causes of the disease. Nevertheless, each of the exposures prior to the person’s contracting the disease… is a factual cause of the person’s disease under the rule in this Section.
This passage arguably extends the reach of Section 27 causation beyond cases in which each contributing factor is independently sufficient to cause the plaintiff’s harm, and even beyond cases in which each defendant’s contribution was necessary to at least one sufficient causal set. After citing a number of supporting cases—in many of which it was unknown whether some or all of the defendants contributed an amount necessary to at least one sufficient causal set—the Reporters’ Note tacitly confirms this understanding of the comment: “These cases thus reflect an application of the principles of this Section and Comment f to a situation in which none of the alternative causes is sufficient by itself, but together they are sufficient and perhaps necessary elements of multiple sufficient causal chains.”
Allowing a jury to find causation in such a case is at the core of many states’ application of the substantial factor test. Courts—typically citing Rutherford v. Owens-Illinois—often capture this approach by stating that causation may be found against a defendant whose tortious conduct contributed substantially—defined as “more than negligible or theoretical”—to the risk of the plaintiff’s harm. Thus, if a plaintiff can prove that a toxin is causally linked generally to a risk of disease, and that any particular defendant contributed to that risk in a “more than negligible” way, then a jury may find causation. A plaintiff might be successful in getting to the jury in such a case even without, for example, relative-incidence statistics that indicate a but-for or overdetermined causal relationship. The language of Comment g and its Reporters’ Note seems to endorse this version of the substantial factor concept in toxic tort cases—although this endorsement might have been more clearly stated.
The Third Restatement embraces the second use for substantial factor more resolutely, however—albeit with a twist. In § 36, entitled “Trivial Contributions to Multiple Sufficient Causes,” the Restatement unambiguously endorses the preclusion of causation where the defendant’s contribution to a causal set is deemed insubstantial. Section 36 states: “When an actor’s negligent conduct constitutes only a trivial contribution to a causal set that is a factual cause of harm under § 27, the harm is not within the scope of the actor’s liability.” This section thus embraces a Rutherford-like approach in excluding negligible causal contributions, but it does so not within the ambit of factual causation, but as a matter of scope of liability.
As the comments to § 36 make clear, the Restatement’s decision to cast this exclusionary rule as a matter of scope of liability is a matter of conceptual fit, “[t]o preserve the largely objective inquiry required by Chapter 5 on factual cause.” Put differently, according to the Restatement, causation is a factual inquiry; by contrast, scope of liability entails a judgment—and whether a causal element is substantial (a word § 36 studiously avoids) or “trivial” requires a judgment.
Although the Restatement is clear in its endorsement of this incarnation of the substantial factor construct, it is also clear that the rule is to be narrowly applied: only in overdetermined cause scenarios, only when “the tortious conduct at issue constitutes a trivial contribution to any sufficient causal set,” and only when the contribution of the various parties is similar in kind. It does not apply in cases in which “the actor who negligently provides the straw that breaks the camel’s back”; that is, when the actor is necessary to the outcome.
In sum, the Third Restatement recognizes that factual causation may be satisfied via either the gateways of but-for or multiple sufficient cause/causal sets (or by use of a Summers v. Tice burden-shifting mechanism). It also suggests—albeit in somewhat hesitant language—that causation might be established against all contributors of a toxin, whether or not they are necessary parts of any sufficient causal set. Only non-trivial contributors will do, however—and triviality is a matter of scope of liability, not factual cause.
3 Doull v. Foster
Despite the many advantages in clarity, guidance, and conceptual coherence offered by the architecture of the Third Restatement’s treatment of factual causation, there has not yet been a wave of cases embracing it. Most jurisdictions yet cling to the “substantial factor” language, and not many have become comfortable with the somewhat cerebral notion of “multiple sufficient causal sets.” Leave it to that valiant rampart of cerebralism, the Massachusetts Supreme Judicial Court, to be the first mover—as it has in the 2021 case of Doull v. Foster. For its decisive and early embrace of the Third Restatement’s provisions, I predict that Doull will hold a place in tort law’s emerging canon.
As is true of so many torts cases, the tale of Laura Doull is a tragic one. For nearly four years, Mrs. Doull was a patient of the defendants Anna Foster, a nurse practitioner, and Dr. Richard Miller, her supervising internist. In treatment of Doull’s perimenopause-related symptoms, Foster prescribed a naturally derived progesterone cream, to be topically applied. Foster admitted at trial that she did not discuss with Ms. Doull the risk that the cream might cause blood clots, not believing that such a risk existed. During the period of her use of the cream, Doull visited Dr. Miller’s practice three times complaining of shortness of breath. Foster assessed the condition as a combination of Doull’s pre-existing asthma and allergies.
In the fourth year of her treatment, Doull had a “seizure-like event” and at the hospital was diagnosed with pulmonary emboli—blood clots in her lungs—a condition that can cause shortness of breath. As it turns out, Doull was developing a rare disease called chronic Thromboembolic pulmonary hypertension (CTEPH). In that year, Doull had surgery to remove the blockage from her lungs, but it was unsuccessful. Nor did any medications help. Doull died from the disease four years later, at the age of forty-three.
Doull (and later her estate) and her family members sued her caregivers, claiming negligence in their failing to inform Doull of the risk of pulmonary emboli posed by the progesterone cream, failing to diagnose her condition, and, against Dr. Miller, failing to supervise Foster’s care of Doull.
The trial judge instructed the jury on causation by referencing only the but-for test. Plaintiffs objected, urging that because their claims were against multiple defendants, the jury must use the alternate, “substantial contributing factor” test. The jury found that the defendants were not negligent in failing to obtain informed consent, but were indeed negligent in failing to diagnose Doull and that Miller had been negligent in supervising Foster. However, the jury found that neither wrong was a but-for cause of Doull’s seizure-like event or her death.
The Massachusetts Supreme Judicial Court’s analysis of the plaintiffs’ causation argument is masterful. It begins by noting that all three Restatements and a majority of jurisdictions apply the but-for test in most cases. The Court then recognizes two categories of cases in which application of the but-for cause rule would be improper: the two-fires scenario and toxic tort cases in which “it can be difficult, if not impossible, for the plaintiff to identify which particular exposures were necessary to bring about the harm.”
Turning to the plaintiffs’ claim that Doull’s case requires a test other than but-for, the court correctly points out that not all cases involving multiple potential causes are alike. Plaintiffs’ claim involved neither an overdetermined cause nor a situation in which it is nearly impossible to determine which exposures were necessary to bring about the harm and which were not. Instead, according to the court, although the case involves multiple defendants, “[t]here is nothing preventing a jury from assessing the evidence and determining which of the causes alleged by the plaintiff were actually necessary to bring about the harm, and which had nothing to do with the harm.” The Court is careful to explain the fallacy of plaintiffs’ argument that because there are multiple defendants, there cannot be a sole but-for cause. (This is a mistake made not only by 1Ls, as it turns out.) Rather, the Court explains that any number of factors—innocent and even tortious—must exist for a tortfeasor’s conduct to injure a plaintiff.
The Court’s opinion might have stopped here, upholding the trial court’s instructions based on existing Massachusetts law. Instead, the Court shifts its sights to the substantial factor test itself—and this is where the Court’s opinion really shines. First, it explains the development of the test by the first and Second Restatements and its previous adoption by Massachusetts courts. It then expressly rejects the test, offering several reasons for its decision. In summary: (1) the substantial factor test is “confusing;” (2) the test is too lenient in that it “invite[s] jurors to skip the factual causation inquiry altogether; (3) it is also too strict, in that it requires a “substantiality” finding in all cases—a determination that is not properly a part of factual causation reasoning; (4) the substantial factor test improperly applies to cases beyond the two scenarios in which an alternative to but-for is needed; (5) the test “blur[s] the line between factual and legal causation”; and, relatedly, (6) it “inserts a high degree of subjectivity as to what is substantial and what is not.”
After a thorough and sophisticated thrashing of the substantial factor approach, the Court adopts the Third Restatement’s approach to causation nearly whole cloth. As the Court succinctly summarizes: “If there must be an exception to but-for causation in cases where the but-for standard fails, we should simply recognize such an exception rather than adopting an entirely different causation standard with confusing terminology and unexpected difficulties. The approach proposed by the Restatement (Third) does exactly that.” The Court then cites the Restatement approach in delineating jury instructions for cases involving multiple sufficient causes.
In two regards, however, the Doull opinion might have improved on the Restatement’s approach, but missed the opportunity to do so. First, the Court described in positive terms courts’ use of the substantial factor test to allow a finding of causation where “it may be clear that a toxic substance or asbestos caused the harm, and that the defendants exposed the plaintiffs to the toxic substance of the asbestos, but it may not be possible to determine which exposures were necessary to have caused the harm.” Note that this approval is not limited to cases involving multiple sufficient causal sets. In rejecting the substantial-factor test, the Court might have been more careful to preserve an alternative approach in toxic-tort cases—including in cases in which the plaintiff has insufficient evidence to show that each defendant was a necessary causal element in at least one sufficient causal set. It is true that such language would have been dictum, as Doull did not involve such a scenario. On the other hand, nor did the case involve overdetermined causes—but the court nonetheless carefully set out jury instructions for such cases.
Similarly, Doull implicitly follows the Third Restatement in suggesting that a “substantiality” determination involves “the types of policy considerations that animate our legal causation jurisprudence,” considerations that “should not be incorporated into the factual causation analysis as well.” Although I do not in this Article take a position on whether trivial causal factors ought to be excluded from liability, I disagree that such a consideration—narrowly defined as it is in the Third Restatement—fits better in the element of scope of liability than in factual causation. Although it is true that “substantiality” involves not a factual inquiry but a judgment about responsibility, so do both the multiple sufficient cause/causal sets rule of Restatement (Third) § 27 as well as the Summers v. Tice burden-shifting of § 28. Whether a substantiality inquiry fits better within the element of factual cause or scope of liability ought instead to be determined by asking “what is the purpose of the inquiry?” In my mind, the instinct to exclude trivial causes exists because proponents of such a rule do not see them as causes at all. By contrast, the focus of the scope-of-liability inquiry is not the allegedly causal element, but the resulting injury; and the purpose of inquiry is to decide whether that injury—although caused by defendant’s negligence—was too unforeseeable to justify liability (or, as § 29 of the Third Restatement explains it, a harm “that result[ed] from the risks that made the actor’s conduct tortious”).
Notwithstanding these relatively minor critiques, Doull is a tour de force on factual causation—its explanations particularly accessible for 1L Torts students. Both courts and casebook editors should take note.
My thanks to Jacob Winton for his dedicated and expert research assistance and to Sarah and Kathryn Higginbotham for their help crystalizing my thoughts during a writing retreat at the Allen Chapel AME Church in Staunton, VA. My thanks also to Mike Green for his helpful comments and for the feedback of participants in the Great Torts Cases conference held at the University of Arizona James E. Rogers College of Law.
© 2023 Walter de Gruyter GmbH, Berlin/Boston