Causation on trial – necessity, sufficiency and the causal questions

I write this addendum to my paper with the objective of better understanding how we compose the counterfactual questions in cases involving multiple factors and the causal relationship to a phenomenon. The concepts of “necessity” and “sufficiency” make the counterfactual question complex and challenging to compose.

As a preliminary comment, I think it is important to point out that the causal question follows only if there has been a demonstrated breach of the duty of care. In this way, we ensure that the causal question does not include a presumption of a breach of duty. I also want to emphasize that issues of what have been called “legal causation” do not arise until the causation question has been asked and answered. Legal causation is concerned with foreseeability, proximate cause, remoteness, and other policy considerations that might lead to no finding of liability despite proof of causation. I suggest it would be better to call legal “causation” by another name to ensure it is not confused with factual causation.

My problem with how courts have dealt with causal questions is focused on the manner in which the causal questions are posed in order to recognize whether it is fair to attribute harm to wrongful conduct.

When most of us think about the but-for test we think about wrongful conduct as being “necessary” to bring about the harm. As I have previously asserted, this is unduly simplistic and creates confusion in the application of the causal test. Necessity is important, but only insofar as it relates to a set of events, taken together, that bring about the harm.

In its simplest terms, a factor is “necessary” if the harm could not have occurred in the absence of the factor. In a scenario where we consider the involvement of a single factor in creating harm, we merely compose a counterfactual that removes that factor, in an imagined world where it did not occur, and ask whether the absence of the factor avoids the harm. If the answer is “yes”, then the factor is “necessary” for the infliction of the harm. For Stapleton, this simple form of necessity merely illustrates one “role” that the factor played. The role here is that the factor was “necessary”. ² Stapleton uses the notion of “involvement” to determine whether there is a causal connection between a factor and a phenomenon. ³ For her, necessity is merely “one form of involvement” – the others being duplicative necessity and contribution. ⁴

Two Factors, Only One is the Cause (both necessary and sufficient)

Let’s use Cook v. Lewis to illustrate. The case involved two hunters firing in the direction of Mr. Lewis, where only one shot struck Mr. Lewis. The breach of duty questions must be asked first. Having found a breach of duty, the causation question is then posed. With regard to Hunter A we might ask the following question:

Has Mr. Lewis proven, on a balance of probabilities, that but for the breach of the standard of care Mr. Lewis would not have been shot?

This is essentially how the question was asked of the jury in the Sacks case. In the two-hunter case, the answer is “no”. This is due to the fact that we do not know if the shot from Hunter A’s shotgun was the one that struck Mr. Lewis. By posing the question this way, we fail to identify the conduct that was “necessary” to bring about the harm. If we pose the same question about Hunter B’s conduct, we get the same answer. Suppose, however, that we ask a different question, such as:

To prevent the injury suffered by Mr. Lewis, would the breach of Hunter A and the breach of Hunter B have to be absent?

This must be the first question posed when one or the other of two wrongful acts caused the loss, but the evidence falls short of establishing which wrongful act is to blame. I have called it the threshold question. One could go further and break this down into two questions:

To prevent the injury suffered by Mr. Lewis, would the breach of Hunter A have to be absent?

To prevent the injury suffered by Mr. Lewis, would the breach of Hunter B have to be absent?

This demonstrates, in the circumstances of the case, that, on the given facts, both breaches of Hunter A and Hunter B would need to be absent to avoid the injury. This does not help us with the dilemma in that case of establishing which shot struck Mr. Lewis.

Two Factors, Each Enough to be the Cause (either necessary, both sufficient)

Another form of involvement in the creation of a phenomenon is that of duplicate necessity, where two factors, either one on its own, is enough to cause the harm. ⁵ This is also referred to as overdetermination. The two merging fires, Fire X and Fire Y, each sufficient to cause the loss, is a simple example of overdetermination. Fire X was not necessary to cause the loss, because its effect was duplicated by Fire Y. Neither is a but-for cause of the loss in the form of simple necessity. Therefore, using simple necessity we might ask:

Has the plaintiff proven that but for Fire X the loss is avoided?

The answer is “no”. This leads us to the absurd conclusion that neither Fire X nor Fire Y were necessary to cause the plaintiff’s loss. As an alternative, the question might be:

Had Fire X not occurred, would Fire Y need to be absent in order to avoid the plaintiff’s loss?

The answer is clearly “yes”. In posing the causal question this way, we have removed the duplicative factor, Fire X, which then allows us to tease out the involvement of Fire Y in the creation of the harm. This illustrates that in overdetermination cases, the involvement of a single factor cannot be fairly determined without the context or role of the other duplicative factor being considered. The same question would then be put in regard to Fire X’s involvement, with a finding that both fires caused the harm.

Another way to ask the question, and avoid a double negative is as follows:

Would Fire Y have caused the plaintiff’s loss if Fire X did not occur?

Once again, the answer is “yes”. In asking the question this way we prove causation where we know causation exists.

To read more, download the full whitepaper. 

¹ Ver 2.1. January 28, 2021

² See Stapleton, Jane. Choosing What We Mean by Causation in the Law, Missouri Law Review, volume 73, Issue 2, Spring 2008, at page 436.

³ Ibid page 436.

⁴ Ibid page 441.

⁵ Ibid page 441.