A Small Test for Slaw’s Lawyer / Judge Readership

(And anybody else who cares to participate)

The assumption around here is that Slaw is the most widely read, or at least one of the most widely read law blogs, amongst lawyers (academic and practicing), judges, and others associated with the profession. I’m going to (try to) put that to the test.

As some of you know, my particular (legal) research interest is factual causation in tort, particularly negligence. I’ve described that interest, on occasion, as the search for the “f” in actual cause.

I pose this question for those of you involved in civil litigation.

Are you prepared to accept that sufficiency rather than necessity is the accurate explanation for factual causation?

That sufficiency doesn’t mean the negligence “all by itself” but means, in substance, “sufficient when added to the background conditions which, together with the negligence” produce a cause.

That the reason why the necessity test you’re accustomed to using in most cases – the but-for test – works is that the parameters of the case are defined, going in, so that necessity is equivalent to sufficiency?

I’m asking you to accept

  • 1. that negligence may be a cause of harm, on a more likely than not basis, even if the negligence is not necessaryfor the occurrence of the harm, so long as it is sufficient for the occurrence of the harm;

and

  • 2. that there is another configuration of facts, which does not include the negligence, which is also sufficient to cause the harm may mean, but does not necessarily mean, that the configuration including the negligence is not a factual cause.

Without giving examples, what this means is that

1. in some cases, a factual cause will be both necessary and sufficient
2. in some cases, a factual cause will be sufficient but not necessary
3. in some cases, a factual cause will be NEITHER necessary nor sufficient. (Trust me on this one, for now. If you accept 1 and 2, 3 necessarily follows – readily enough, I think – and in any event it’s clearly explained with an example. An example is on the next screen.)

I, of course, don’t expect responses from judges who read this blog, but perhaps someone will say something in reasons for judgment that will hint that the judge has considered the issue, maybe even because the judge read something here.

DC

The example of a cause which is neither necessary nor sufficient. (I’m not the originator of the example. It’s well known in more academic circles. One scholar has called it the “too little and too much”. The quotation comes from his text: Wright, Cases and Materials on Tort – it’s a U.S. text)

Assume three drops of poison are necessary and sufficient (in conjunction with other “background” conditions) for killing someone. Emily and three other people each put one drop into a cup which Paul drinks. Paul dies. … [E]each drop of poison was a cause of … Paul’s death, since each drop was a necessary element of a sufficient set of actual antecedent conditions that includes only two of the other drops. The extra (fourth) drop not in the set does not preempt that set’s causal effect by eliminating one of the “background” conditions that also is necessary for the actual sufficiency of the set, but rather duplicates or reinforces the causal effect of the sufficient set.

The underlying assumption is that the poison mixes with the liquid in the cup so it’s not possible to separate the individual does. The drops of poison simply combine into a larger amount of poison, amounting to more than 3 or more “drops” worth. There’s no chemical difference in the various drops.

Was the conduct of any one of the people necessary? No, because taking his or her drop out still leaves enough to poison. (Addendum: May 24 – this use of “necessary” is diferent from the but-for use of “necessary” because, there, the negligence is necessary for the harm to occur at all. Here, it’s necessary only for one configuration of of events sufficient to cause the harm.)

Was the conduct of any one of the people sufficient? No, because we need a total amount of poison equal to 3 or more drops. 

And it’s not a case of the straw breaking the camels back. We don’t have a case where P drank from the cup after the 3rd drop.

If you’ve come this far, and care to, tell whether you agree with the statements that follow after the hypothetical.

  • S Co. rents a car to D. S Co. hasn’t maintained the car properly, so the brakes will fail after a number of usages. D doesn’t know that. D drives away. The brakes seem to be working. D is driving within the speed limit. For the last 1/2 mile or so he hasn’t needed to apply the brakes. The last time D used them they worked properly. For whatever reason D stops paying proper attention to what’s ahead of him. As a result, he doesn’t see P’s car come out of the side road onto the highway. By the time D looks up and he sees P’s car, it is too late to brake. D’s car strikes P car. An investigation afterwards shows that the brakes on D’s car had failed about 1/2 mile earlier because the brake line broke, allowing the brake fluid to escape.

The questions are:

  • 1. Do you consider D’s failure to apply the brakes a factual cause of the accident?
  • 2. Do you consider S’s failure to maintain the brakes a factual cause of the accident?
  • 3. Do you consider S’s giving the defective vehicle to D a cause of the accident.

P’s conduct in pulling out in front of D was, obviously, a factual cause so you don’t need to mention that unless you somehow think it relevant.

Thanks.

David

P.S. May 23/10: For any one who is interested, west of here (in the physical world) and merely elsewhere in the datasphere (?) – blogsphere(?) – on the University of Alberta Facultyof Law, Faculty Blog – I have been and will be writing about the substantive law, and some of the cases, dealing with factual causation issues. Getting the law right is important for any number of good reasons, including that it does make a practical difference to the both the public and internally even if (formally) appeals are from the result and not the reasons for the result.

Comments

  1. I’m not sure I’m going to answer David’s question(s) but I will mention some thoughts that occur to me on reading this post.

    * The law of negligence arrives at liability on a balance of probabilities, which I expect allows (and is intended to allow) a certain amount of practical good sense to apply to questions of causation and responsibility.

    * The criminal law standard of proof beyond a reasonable doubt may require other degrees of causation.

    * Was Cook v Lewis a tort or criminal case? There two people might have fired the fatal bullet. The inability to determine which it was led to no one being held responsible in law (if memory serves me…)

    * The criminal law does allow for some sensible allocations of fault or at least sensible outcomes, though. I would think that the poisoners each had at least the intention of causing the victim harm, even if not killing him. They may have acted in reckless disregard of what other circumstances might have made the result of their single drop more serious. I would think they would have a hard time escaping a manslaughter charge.

    * I am reminded of the advice given (not by the RCMP or DHS) to the person afraid to fly because of a fear that there may be a bomb on board: take your own bomb, because the chances that there are two bombs on board the same flight are infinitesimal (or even more infinitesimal than the chances that there is one…). I accept David’s possible reply that it may be difficult to allocate responsibility if the plane then blows up in flight. Cook v Lewis?

    * In civil cases the rules of joint and several liability – in some expressions of the principle at least – presume that each defendant of several defendants is 100% responsible for the harm, and liability is apportioned among them only to prevent double recovery. In some fact situations, however, it is pretty clear that a combination of facts are required for any harm to have been produced – sometimes including the contributory negligence of the plaintiff. However, the analysis of fault and the legal rule that deals with the problem of compensating the victim when one or more defendants are insolvent or unavailable are not necessarily the same.

    * In David’s auto collision case, I don’t see a difference between Qs 2 and 3. Both had to be present for the rental agency to be liable, but both were. That kind of combination presents no problems for the law, or at least I do not understand David to have raised any on that kind of scenario.

    * As for Q 1, if the expert advice is completely certain that stepping on the brake would have made no difference, then David would not be liable for having failed to do so. I can imagine a court finding him partly responsible for the collision, though, for inattentiveness that might have prevented him from taking other evasive action, or at least taking his foot off the gas and reducing the severity of the crash.

    Are there examples of cases where what a sole defendant did was neither necessary or sufficient to cause the harm but liability ensues anyway (assuming no contributory negligence by the plaintiff)? I think the ‘last clear chance’ rule is in effect in some provinces, though not in most. Is that a possible example (and is David’s analysis the kind of reasoning that brought that rule to an end in most places)?

  2. The problem with substantive messages is that they often invoke substantive replies, which consume substantive time in crafting a substantive response which … Anyway, I’ll respond to JG’s points other than those relating to criminal law, except where the comment requires me to venture into those wilds.

    JG: The law of negligence arrives at liability on a balance of probabilities, which I expect allows (and is intended to allow) a certain amount of practical good sense to apply to questions of causation and responsibility

    Factual causation inquiries and decisions are (supposed to be) entirely separate from decisions of responsibility, at least once one has defined the parameters of what the system will ever take into account as relevant. That means responsibility (ought) isn’t supposed to sway factual causation (is). In the causation inquiry, responsibility issues rear in the proximate cause / remoteness inquiry. Having said that, amongst the jurisdictions about which I know something (Canada, US, UK, Australia and New Zealand) only the Aust HC has explicitly admitted that policy drives even the factual causation but-for decisions.

    The problems with appeals to common sense at the factual causation sense – leaving aside the issues of what the content is (process only? substance?) is that it’s an unregulated form of judicial notice.

    JG: The criminal law standard of proof beyond a reasonable doubt may require other degrees of causation.

    Yes: And there’s a wealth of scholarship about those different degrees and what that means, depending on what the charge is.

    My comment, though, does not extend to criminal law – even though Athey cited a Cdn criminal law decision as one of its authorities for the material contribution doctrine.

    * Was Cook v Lewis a tort or criminal case? …

    C v L was a tort case. Rumours of the plaintiff’s demise (as a result of the bird shot) are exaggerated. He was injured not killed. You’ve made the same mistake about the result that many make. Only the trial jury dismissed. Both the BCCA and the SCC held that the jury was obliged to decide which one – since both denied shooting in the victim’s direction and it had to be at least one of the two – and sent the action back for a new trial. The SCC reversed the onus. I’ve not been able to find out if the new trial was held. It turns out I know someone who knew the plaintiff’s lawyer, but he (the lawyer) died before I acquired that connection. I’m sure it’s possible to find out what eventually happened.

    JG: The criminal law does allow for some sensible allocations of fault or at least sensible outcomes, though. … I would think [the poisoners] would have a hard time escaping a manslaughter charge.

    No doubt, but we’re dealing with negligence where the conduct is intended but the consequences not. So, assume in my example that the P’s friends thought that the fluid was sweetener.

    JG: I am reminded of the advice given .. to the person afraid to fly because of a fear that there may be a bomb on board … I accept David’s possible reply that it may be difficult to allocate responsibility if the plane then blows up in flight. Cook v Lewis?

    Which Cook v Lewis explanation? The SCC’s in Cook v Lewis, itself, or the very different use of C v L by the SCC in Resurfice? First, the rule in C v L can apply only where there are just two persons, only one of whose conduct was the cause. (That, in theory, is likely why the SCC in Walker Estate held the ONCA was wrong in basing the Red Cross’s liability on C v L.) In the traditional C v L rule – the reverse-onus rule – it’s left to the defendant to show on the balance of probability that it wasn’t the defendant’s (conduct) bomb. If neither defendant does, C v L deems the conduct of both to have been a legal cause (even though we know that the conduct of both couldn’t have been.) This is known as legal fiction. (Literature calls that fantasy, but we won’t go there.) In the Resurfice restatement of C v L, there’s no longer a reverse onus. The possibility of factual causation added to fault is sufficient to satisfy the causation requirement of the the tort.

    JG In civil cases the rules of joint and several liability…

    Joint and several liability rules are responsibility rules, not factual causation rules.

    JG: In David’s auto collision case, I don’t see a difference between Qs 2 and 3. Both had to be present for the rental agency to be liable, but both were. That kind of combination presents no problems for the law, or at least I do not understand David to have raised any on that kind of scenario.

    On any orthodox application of the but-for test, the conduct of neither garage nor driver is a factual cause. The but-for test requires the judge or jury look a the facts as they actually occurred, substituting only proper conduct for the negligence of the defendant, in deciding if that defendant’s conduct is negligent. The driver is able to say: my failure did not make a difference. The brakes were not working so even if I had applied them properly, the car would not have stopped – the accident would still have happened.

    For the garage, the problem is that the brakes were never used. The but-for test does not allow the the court to change these facts to assume the brakes were The garage is able to say: the driver never applied the brakes, so the condition of the brakes is irrelevant as a factual cause. (It’s just as irrelevant as it would have been if the brakes were working, and not used. The situation becomes different if the driver learned the brakes were defective and that’s why he didn’t apply them. Then their condition becomes part of the process as it played out.)

    The double omission case create huge conceptual problems for the but-for test, both in practice and theory. It’s part of the larger problem of how to analyze omissions – a failure to act – as a cause, which becomes even more complicated where the failure to act is the failure to do something which, if it had been done, would have prevented the injury.

    The problem I’ve described comes from a “classic” and real US case: known Saunders System v Adams. Adams (the plaintiff) sued only the garage. Adams lost. More than a handful of reputable scholars think that was the correct decision. A leading US tort prof has done a study – the subjects were his tort students – posing questions 1 and 2, only. There was no majority position. There were some 1, some 2, some 1 and 2, some neither.

    Under more than one analysis of the case, 3 – leasing the defective vehicle – was the only aspect of the garage’s conduct which was a factual cause. The failure to repair was preempted by the failure to brake.

    JG: As for Q 1, if the expert advice is completely certain that stepping on the brake would have made no difference, then David would not be liable for having failed to do so….

    I’m not dealing with liability but only with factual causation. If there’s no factual causation at all – if D’s conduct is not a factual cause then there’s no basis for the imposition of liability when the governing test for factual cause is but-for.

    To hold that not stepping on the brakes is a sufficient defence, because “it would have made no difference” is a liability decision, not a factual causation decision. If “it would have made no difference” is relevant at all, D has to escape completely because, even if D had been completely attentive, the accident would still have happened.

    An the facts of the case, the only mistake D made was not braking, and there was nothing else he should (or could) have done that would have made a difference if he were paying attention.

    JG: Are there examples of … neither necessary or sufficient [cases] … [what’s the significance if any of the old] ‘last clear chance’ rule ….?

    The Viet Nam Agent Orange / Dioxin cases (in the US) would be an example if the Dioxin had been manufactured by more than one source or if there were more than one source of the dioxin to which any soldier was exposed. The DES cases – law students still study Sindell v Abbott – is another example if we assume it took more than one use of DES by the mother and the mother used DES from more than one mfg. There may well be better examples in US toxic tort jurisprudence.

    Every modern case of pollution alleged damage where the class alleges that that the pollution was caused by toxins from more than one source creates the problem, so long as we assume that the individual discharges, themselves, would not have been sufficient. That is quite often the case.

    Corey v Havener – two cars driving up beside a horse, one on each side, noise startling the horse, rider thrown and injured – is an example. We can’t assume that the noise from one car would have been enough.

    The last clear chance rule is supposed to dead everywhere in common law Canada, either by statute (for example Alberta) or by judicial fiat. The doctrine was born to provide the plaintiff with a way around the otherwise applicable complete defence of contributory fault. It never applied as between wrongdoers.

    I suppose the garage defective brakes – negligent driver example is a case in which, in some jurisdictions, some judge or jury might have thought of applying it. One problem, though, is that the driver didn’t have the last clear chance – the brakes didn’t work, recall? The last clear chance doctrine was a version of a proximate cause / remoteness analysis. It didn’t deny the historical causal status of the other causal candidates – just chooses amongst actual causal candidates for the one the court would hold was the legal factual cause. (The cause chosen, by the way, didn’t have to be the last in time.)

  3. Oh, boy, fun with thin skulls! Also, fun with medical causation.

    I did a master’s thesis on deconstructing the medical and legal languages of causation in trauma-triggered fibromyalgia cases. (SFU ’99)

    When doctors talk about causation, they are speaking another language. When I wrote the thesis, there were no cases discussing the differences. Perhaps there are now; I hope so.

    Classic medical causation absolutely required a necessary condition. For example, you can’t get tuberculosis unless you’re carrying a tubercle, but you can carry a tubercle around all your life without ever running into the other conditions that amount to sufficiency for actually having TB.

    However, in the past few decades, new models of medical causation have been proposed and worked with. The idea of disease now has looser strictures and medical causation looks more like a sufficiency model.

    This doesn’t make a judge’s life any easier, especially since the differences between medical and legal causation are seldom explained or explored.

    The problem with disease is that each one follows its own narrative before the true origin is determined. In the first stage of the narrative, the patient or victim commonly carries the blame (think poets with TB, poor blacks with pellagra, gay folk with HIV, women with FMS). Once the mechanism of the disease is finally discovered, the causation picture changes completely and it becomes easier to pinpoint the necessary and sufficient.

    In the case of illness, it’s never going to be easy. But it’s never going to be right, either, unless we know the medical language of causation as well as the legal.

  4. David Cheifetz

    When doctors talk about causation, they are speaking another language. When I wrote the thesis, there were no cases discussing the differences. Perhaps there are now; I hope so.

    There are recent enough cases discussing the reasons for the differences between the medical and legal causation concepts – I’m sure some in some detail, even if they’re not Canadian, though some probably are – but I don’t have them offhand. But, more often than not, most Cdn judges fall back on the Snell v Farrell (1990)declaration that the concepts are different because science requires a degree of precision that law doesn’t. “Causation need not be determined with scientific precision.”

    Reasonable people will differ on whether that statement has any valid meaning beyond this: “law’s answer doesn’t have to be science’s”.

    For what it’s worth, my questions ask about issues that precede thin-skull. If one says that the issue has a thin-skull problem, one has already identified some factors that are factual causes – the factors that predispose to whatever the ultimate harm is.

    I’m lazy. Is your thesis on line? If so, what’s the URL?

    DC