The History of Proximate Causation

Bacon's Rule—The Last Cause Controls

Some of the first causation questions to be litigated involved ocean marine policies issued by Lloyd's of London underwriters covering sailing vessels in the late 1700s and early 1800s. At that time, marine policies covered vessels and cargo against the "perils of the seas"—grounding, collision, allision, inrush of water, and so forth. But, in many instances, a peril of the sea was itself caused by some act of negligence on the part of the master or crew. For example, suppose the night watchman falls asleep in the crow's nest and consequently the ship runs aground on a sandbar. Is that a loss caused by a peril of the sea? True, the ship was damaged when it ran aground, but the underlying negligence of the watchman was what caused the grounding in the first place. Symbolically, the facts of the claim can be represented as follows.

• Cause 1. Negligence of the watchman in falling asleep
• Cause 2. Ship ran aground on the sandbar
• Loss. Damage to hull
• (Cause 1) › (Cause 2) › (Loss)

So which is the cause of the loss—the negligence of the watchman (not covered) or the grounding on the sandbar (a covered peril)?

To answer that question, some courts considering these early marine cases looked to an influential treatise called Maxims of the Law, written in 1596 by Sir Francis Bacon (1561–1626). Bacon read a lot of cases and tried to synthesize the underlying rules he thought courts were applying in various situations and on various issues, including causation. His first and foremost rule (Regula I) was "In jure non remota causa sed proxima spectator," which translates as, "In law, one looks to the near cause, not the remote one." He wrote:

It were infinite for the law to judge the cause of causes, and their impulsions one of another; therefore it contenteth itself with the immediate cause, and judgeth of the acts by that, without looking to any further degree.…

In other words, when asking what was the "cause" of an end result, courts should only look back one link in the chain of causation and settle on the event that immediately preceded the result. Otherwise, you could be looking backward along the chain of causation forever and never come to an answer. So, originally, "proximate" meant "close," as in Latin, and the proximate cause was the nearest cause or the cause immediately preceding the end result. Remote events further back along the chain should be ignored, in Bacon's view.

Bacon's theory made sense to courts considering causation questions under early marine policies. A great many losses can ultimately be traced to human negligence. If the courts were to follow the chain of causation to the root cause of every loss, not many marine claims would be covered, which would greatly diminish the economic value of taking out insurance to cover ocean commerce. Therefore, some of these courts adopted Bacon's view and initially held that the proximate cause of a loss under an ocean marine policy was the peril that was closest to the happening of the damage.

• Busk v. Royal Ins. Exch., 2 B & Ald. 73 (1818)
• Howard Fire Ins. Co. v. Norwich & N.Y. Transp. Co., 79 U.S. 194, 20 L. Ed. 378 (1870)
• Queen Ins. Co. of Am. v. Globe Rutgers Fire Ins. Co., 263 U.S. 487, 20 L. Ed. 738 (1924) (Holmes, J.) ("[T]he common understanding is that in construing these [ocean marine] policies we are not to take broad views but generally [we] are to stop our inquiries with the cause nearest to the loss.")

In the hypothetical above, where the watchman fell asleep and the ship ran aground on the sandbar, courts employing Bacon's rule would have held that (Cause 2) the grounding on the sandbar was the proximate cause of the damage to the hull, and that, because grounding was covered as a peril of the sea, the loss was covered. Under this view, the underlying negligence of the watchman would be considered a remote cause that has no legal significance for purposes of the causation analysis.

Court cases applying Bacon's version of the proximate cause rule establish a particular pattern. In chain-of-events-type situations, where an initial peril (Cause 1) directly caused a subsequent peril (Cause 2), courts employing Bacon's rule would always focus on the subsequent peril (Cause 2). If the subsequent peril (Cause 2) was covered, they would hold that the loss was covered. If the subsequent peril (Cause 2) was not covered or excluded, they would hold that the loss was not covered or excluded. Table 1 shows this pattern.

Newton's Rule—The Initial Cause Controls

Not all courts accepted Bacon's theory. There was a competing line of decisions based on the work of Sir Isaac Newton (1642–1727), who, in 1687, published Philosophiæ Naturalis Principia Mathematica, laying the groundwork for classical mechanics. Newton's first law of motion was that physical objects will remain at rest unless compelled to change their resting state by an outside force. Consider a cue ball and a nine ball resting on the surface of a pool table. Suppose a player takes a stick and strikes the cue ball, the cue ball strikes the nine ball, and the nine ball drops into the pocket. Newton would frame the chain of events as follows.

• Cause 1. Player strikes the cue ball
• Cause 2. Cue ball strikes the nine ball
• Loss. Nine ball drops into the pocket
• (Cause 1) › (Cause 2) › (Loss)

Newton would say that what caused the nine ball to drop was Cause 1, or the kinetic energy imparted to the cue ball by the player's strike. Cause 2, the subsequent collision between the cue ball and the nine ball, merely transmitted the kinetic energy of the player's strike from one to the other, making Cause 2 an instrumentality in the chain of events. But Cause 2 was not the underlying cause, in Newton's view.

Newton's theory of causation slowly permeated judicial thought. As early as the mid-19th century, American courts began rejecting Bacon's maxim "non remota causa sed proxima spectator" in favor of the more modern or scientific view that the proximate cause was the first event that set the chain of events in motion culminating in the end result.

• Montgomery v. Firemen's Ins. Co., 55 Ky. 427 (Ct. App. 1855) (explaining that "older cases" adhered to a rule that "a loss was to be attributed to the cause immediately operating at the time of its occurrence," but the "modern decisions" had "established the more reasonable doctrine" that a loss is covered if "the peril insured against puts the destructive cause in operation")
• Aetna Ins. Co. v. Boon, 95 U.S. 117, 24 L. Ed. 395 (1877) ("if two causes conspire, and one must be chosen, the more scientific inquiry seems to be, whether one is not the efficient cause, and the other merely instrumental or merely incidental, and not which is nearer in place or time to the consummation of the catastrophe")

Court cases applying Newton's version of the proximate cause rule establish a different pattern. Where an initial peril (Cause 1) directly causes a subsequent peril (Cause 2), courts employing Newton's rule would always focus on the initial peril (Cause 1). If the initial peril (Cause 1) was covered, they would hold that the loss was covered. If the initial peril (Cause 1) was not covered or excluded, they would hold that the loss was not covered or excluded. Table 2 demonstrates this phenomenon.

Differentiating Newton from Bacon

Adapted for use as a rule for interpreting causation language in a first-party property policy, Newton's theory of causation from classical mechanics was not completely satisfactory and required a few tweaks. For one thing, the word "proximate" means "close" in Latin. To differentiate Bacon's idea of the proximate cause as being the one nearest in time to the happening of the loss (i.e., Cause 2), courts employing Newton's idea of proximate cause as being the initial cause that set the chain of events in motion (i.e., Cause 1) had to come up with a new label.

Here, judges borrowed from the works of Aristotle (384–322 B.C.). In his Lectures on Physics, Book II, Aristotle wrote that the word "cause" has several meanings in the philosophic sense. One meaning of the word "cause" is "the primary source of the change or coming to rest." In his Lectures on Metaphysics, Book V.2, he wrote that in this sense, "causation" can be understood as "That from which the change or the resting from change first begins." In his Lectures on Meteorology, Book IV, Aristotle called the agent that brings about a change in the state of matter the "efficient cause." Aristotle's definition of the efficient cause as being "that from which the change first begins" aptly described Newton's idea, that the scientific cause of an end result was the first event setting the chain of events in motion. Hence, courts adopted the phrase "efficient proximate cause" or "efficient cause" to describe Newton's concept in their decisions.

Exception for Human Negligence

Another problem was that, under Newton's theory, the underlying negligence of the policyholder could be cited as the initial peril that set the chain of events in motion. If a named perils policy did not list negligence as a covered peril, insurers could argue that a good many property losses would not be covered. To combat this, courts simply carved out an exception to Newton's rule for human negligence. That is, courts decided that human negligence would be simply ignored when deciding causation questions under a named-perils policy.

For example, in Federal Ins. Co. v. Tamiami Trail Tours, Inc., 117 F.2d 794 (5th Cir. 1941), the negligence of the insured in failing to properly maintain a bus led to a fire, which caused damage to the vehicle. The named perils policy covered losses caused by fire but did not expressly cover losses caused by negligence. The insurer invoked Newton's rule that the initial negligence of the policyholder set the chain of events in motion and ultimately led to the loss. It argued that, since negligence was not covered, the loss was not covered. However, the court ruled that human negligence was not cognizable as a separate peril for purposes of the causation analysis. The court explained:

An overwhelming percentage of all insurable losses sustained because of fire can be directly traced to some act or acts of negligence. Were it not for the errant human element, the hazards insured against would be greatly diminished. It is in full appreciation of these conditions that the property owner seeks insurance, and it is after painstaking analysis of them that the insurer fixes his premiums and issues the policies. It is in recognition of this practice that the law requires the insurer to assume the risk of the negligence of the insured and permits recovery by an insured whose negligence proximately caused the loss. In the absence of fraud or gross negligence on the part of the insured, his negligence is no defense against his recovery. [Footnote omitted.]

Taking negligence out of the fact pattern leading up to the damage to the bus, the only peril left was fire, which was a covered peril.

Treatment of All Risks Policies

That took care of the initial negligence problem under named perils policies. But, as for all risks policies, most courts take the opposite view, reasoning that all risks policies are different from named perils policies. Under an all risks policy, losses caused by any cause of loss whatsoever are covered unless there is a specific exclusion. If an all risks policy does not specifically exclude negligence, most courts presume that the insurer meant to provide coverage where human negligence was the initial peril that initiated a chain of events leading up to the loss. See Associated Eng'rs, Inc. v. American Nat'l Fire Ins. Co., 175 F. Supp. 352 (N.D. Cal. 1959); General American Transp. Corp. v. Sun Ins. Office, 239 F. Supp. 844 (E.D. Tenn. 1965).

While that line of thinking has surface appeal, it is based on an incomplete understanding of causation theory and leads to a serious interpretation problem. Suppose human negligence precipitates an excluded peril. For example, consider what happened in Safeco Ins. Co. v. Guyton, 692 F.2d 551 (9th Cir. 1982). Officials with the California Water District were negligent in maintaining a levee. When rains from Hurricane Kathleen inundated the area, the levees broke, flooding parts of the City of Palm Desert, California. Framing the facts, we had:

• Cause 1. Negligence of the officials
• Cause 2. Flooding
• Loss. Damage to residences
• (Cause 1) › (Cause 2) › (Loss)

The court applied Newton's rule of causation and held that the first cause that set the chain of events in motion, the negligence of the officials, was the "proximate cause" and that, because the all risks homeowners policy did not specifically exclude negligence, the flooding losses were covered.

In practical effect, recognizing human negligence as a root cause under an all risks policy means that the policyholder can very easily argue around policy exclusions. A good many property losses are precipitated by human negligence. If negligence causes flood, and the courts accept negligence as the proximate cause, then there is no effective way an insurer can ever exclude flood. Insurers created so-called anti-concurrent causation (ACC) language in 1984 in an effort to combat this wrong-headed application of Newton's causation theory. But ACC language is difficult for policyholders to understand and for courts to apply.

A better, more reasoned approach is to treat named perils and all risks policies the same. Human negligence should be ignored under either kind of policy. Or, stated more precisely, the risk of human negligence should be considered subsumed in the subsequent peril, whatever that peril turns out to be. If negligence causes a fire, and the policy covers fire, the risk of negligence should be considered subsumed in the subsequent fire that the insurer chose to cover. If negligence causes a flood, and the policy excludes flood, the risk of negligence should be considered subsumed in the subsequent flood that the insurer chose to exclude. That way, human negligence could not be used by either side to wrongfully avoid policy coverages or wrongfully circumvent policy exclusions.

A minority of courts applies the same exception to all risks policies as they do to named perils policies. That is, they hold that human negligence that sets a chain of events in motion is not cognizable as a separate peril for purposes of causation analysis under an all risks policy (which is the better reasoned view).

• E.B. Metal & Rubber Indus., Inc. v. Federal Ins. Co., 84 A.D.2d 662, 444 N.Y.S.2d 321 (App. Div. 3d Dep't 1981)
• Chadwick v. Fire Ins. Exch., 17 Cal. App. 4th 1112, 21 Cal. Rptr. 2d 871 (1993)
• In re Katrina Canal Breaches Litig., 495 F.3d 191 (5th Cir. 2007)

Conclusion

For most of the 19th century, courts in America and England struggled to settle on which theory of causation they would employ in a first-party property case. By the early 20th century, the debate in the case law largely died out, with almost all courts adopting Newton's rule in the end. Overwhelmingly, courts now hold that, in chain-of-events-type cases, where an initial peril causes another peril, which causes the loss, the efficient proximate cause is (Cause 1) the initial peril that set the chain of events in motion. Sometimes, a stray court may apply a line of thinking reminiscent of Bacon's rule and will select the last event in the chain of events immediately preceding the loss as the proximate cause, but today those instances are exceedingly rare.