Modified Causation in Workers Compensation

Causation in law is a legal fiction. The philosophical underpinnings behind compensation in tort law require some finding of fault, in order to restore the party to their original position. The but-for test used to evaluate these claims is the compromise the law has developed to hold someone accountable for harm suffered by another party.

However, not all forms of compensation in law are administrated by tort law. Injuries suffered by workers as part of the workforce, in particular, have been carved out into a no-fault regime, specifically for the purpose of resolving these issues more efficiently, more effectively, and with less acrimony between the parties. Employers pay into this system for compensation into a shared pool, and workers forgo their rights to sue in tort as a result of the relevant scheme in place.

The statutory basis for doing so, which varies slightly between the different provinces, necessarily employs a different basis for determining eligibility. For example, British Columbia’s Workers Compensation Act provides that a list of defined occupational diseases, listed in Schedule B, are deemed to be due to the nature of the employment, but this presumption can be rebutted by the employer with evidence. The inclusion of these conditions in part is based on a known prevalence of certain conditions when exposed to occupational hazards in the workplace, but also the logistical difficulties in proving causation

Exclusion from this list is not definitive. Part 4 of the Act provides for an adjudicative tribunal to resolve additional and disputed claims. Section 250(4) of the Act states,

If the appeal tribunal is hearing an appeal respecting the compensation of a worker and the evidence supporting different findings on an issue is evenly weighted in that case, the appeal tribunal must resolve that issue in a manner that favours the worker.

This test is a lower standard than the but-for test employed in civil actions, and because a tie goes to the claimant, it is even a lower standard than the balance of probabilities (50 per cent plus one). A reduced is justifiable as a matter of public policy given the compensatory purposes stated under Division 2 of the Act.

But how low does this standard go? The Supreme Court of Canada weighed in on this issue this past summer in British Columbia (Workers’ Compensation Appeal Tribunal) v. Fraser Health Authority. A modified version of this case was discussed this weekend at the 2016 Paralegal Cup in Toronto.

In this case, 7 lab technicians in a hospital contracted breast cancer, which they claimed as arising out of and in the course of that employment. An investigation of their workplace revealed a lack of any scientific link between their employment and and the higher incident of cancer, but the law does not work in scientific certainties, and certainly not in the context of administrative law.

Although the Workers’ Compensation Board officer denied each of the workers’ claims, they appealed to the Workers’ Compensation Appeal Tribunal, which found that these were indeed occupational diseases. In applying s. 250(4), the Tribunal used the presence of a statistically significant cluster of incidence for breast cancer, eight times higher than the standard incident ratio, and a historic exposure to low levels of a known carcinogen. The Tribunal also employed the Court’s decision in Snell v. Farrell to create a reasonable inference of “causative significance,” though the tribunal is notably “not bound by legal precedent” under s. 250(1).

The different threshold is obtained not from the statute itself, but from volume 2 of the Board’s Rehabilitation Services & Claims Manual. The statute does state in s. 250(2) that although the Tribunal must make decisions on the individual merits and justice of the case, they must apply the relevant policies of the Board.

This manual provides guidelines for the conditions of payment of benefits through consideration of any physiological associations the extent of the employment activity and any physiological association with harm, any temporal relationship present, and any medical conditions from outside of the workplace that were a factor in the harm. Chapter 3, policy item #14.00 of the Manual indicates that causative significant requires “more than a trivial or insignificant aspect of the injury or death, but Chapter 4, policy item #26.22 of the Manual also indicates that if there is no evidence or insufficient “positive evidence,” the Board must deny the claim.

Maintaining a less stringent burden of proof in this administrative context is one that has been upheld by the Supreme Court of Canada in Pasiechnyk v. Saskatchewan. At para 27, the Court emphasized that the policy principles behind workers compensation are deeply interconnected. Quick compensation is only possible with quick and no-fault resolution of claims, and security of payment is only possible because of a mandatory system maintained by employer contributions. It is the exclusion of these claims from the tort system, and the tort definition of causation, which makes the system feasible to begin with.

On judicial review of BCWCAT v Fraser Health at the Supreme Court of British Columbia, the decision was set aside and sent back to the Tribunal, on the basis that the Tribunal ignored the expert evidence that there was no evidence of workplace factors contributing to the incidence of breast cancer. The court cited Moore v. Castlegar & District Hospital at para 11 that the common sense inference used in Snell should not be applied when confronted with contradictory medical evidence.

The matter was then appealed to the Court of Appeal, where the court spent an inordinate amount of time discussing the post-Dunsmuir use of a patently unreasonable test, and then to the Supreme Court of Canada. The majority of the Court held for the workers, and the sole dissent by Justice Côté finding for the Board in denying the claim.

In upholding the Tribunal’s decision, the majority relied heavily on the distinction for causation under s. 250(4). The expert reports used by the Board were still attempting to hold the claim to a level of scientific certainty, which is wholly inapplicable in determining causation for workers’ claims. Instead, it is open to a trier of fact to consider other information outside of expert evidence to make an inference of causation.

The dissent expressed deep concern over the type of evidence introduced at the Tribunal to support causation. At para 60, Justice Côté noted that elevated rates of clusters always occur as a statistical phenomenon, even when there are no risk factors present at all. The workers’ exposure to chemical substances were minimized due to safety controls, and it was highly unlikely that they were ever exposed to levels which were harmful. No synergistic or additive effect was found, and there was no basis for inferring causative significance, absent mere speculation. The inability of the experts to definitively rule out any environmental work factors as a causal risk was not in of itself the basis for positive evidence.

The majority stated at para 38 that the inference of causation can be made through merely circumstantial evidence, meaning the use of “positive evidence” from the Board’s policy can include inconclusive or even contrary expert evidence. In fact this inference can be made in the civil context as well. At para 32, the majority cited Clements, which itself explained at paras 10-12 how inferences can also be made in the civil context when there is no evidence available. As I pointed out in the Western Journal of Legal Studies, Clements also appeared to have created a more restrictive approach towards the material contribution test, focused on impossibility of identifying a specific tortfeasor, rather than simply a complicated factual record. The material contribution test may be a rare beast which has never actually been sighted at the Supreme Court of Canada, but its distant cousin, the causative significance test, appears to have paid a visit, albeit outside the context of tort law.

The majority of the Court’s decision can perhaps be best understood as a a high level of deference afforded to a specialized tribunal. The weighing of a statistically significant cluster of cases may suffice, but only if properly weighed as such by the trier of fact. In particular, the Tribunal used a 1965 article by Bradford Hill, “The Environment and Disease: Association or Causation?” This article explained how to properly weigh epidemiological evidence of causation using 9 different factors:

  1. Strength
  2. Consistency
  3. Specificity
  4. Temporality
  5. Biological gradient
  6. Plausibility
  7. Coherence
  8. Experimental evidence

The Tribunal decided it was not necessary to point to any specific causal agent, only for the evidence to point to causal link between a disease and an occupation. Although the majority of the criteria in the Hill article were not met, the criteria of strength of association was met through the incidence of the disease, and the temporal relationship was met with the time of the onset of the disease.

The flexibility and deference afforded to the Tribunal is ultimately what makes the system workable, as workers compensation systems across the country are already under considerable strain. Limited budgets have pushed the Board in Ontario to apply cuts and increasingly deny claims. Frustrated workers and unions have urged the province’s ombudsman to investigate.

Utilizing a lower standard than the but-for test allows for workers compensation to provide some relief without relying on exhaustive scientific evidence. But the approval of claims of a more tenuous nature than the civil standard also means there is less money in the pool for other workers who do have a clear scientific or evidentiary claim for their injury being related to their job. The Board has a unique challenge ahead of it in continuing to shift and strike the appropriate balance between the two.



  1. After reading the case, I can appreciate why the Supreme Court of Canada would want to grant compensation to the workers. But, with all due respect to Justice Russell Brown, he appears to have conflated the concept of “burden of proof” with the concept of “standard of proof”. “Burden of proof” is expressed by identifying a party who, to be successful in the proceeding, has the onus of establishing that the evidence for or against a proposition meeting the applicable “standard of proof”. In criminal proceedings, where the evidentiary standard is “beyond a reasonable doubt”, the Crown is said to have “the burden of proof”. In a criminal proceeding the accused is not required to adduce any evidence. The Crown must establish the elements of the subject offence beyond a reasonable doubt, and if it fails to do this, the accused will succeed in the proceeding, without doing anything. In a civil proceeding, the “burden of proof” is on the plaintiff bringing the cause of action. In this particular administrative proceeding, there is no “burden of proof”. The WSIAT & the WSIB are investigative in nature, not adversarial. As such, it would be problematic to apply a “burden of proof”. The decision seems to use the “burden of proof” and the “standard of proof” interchangeably. They are two separately, and conceptually distinct issues. Section 124(2) of the WSIA does not apply to the “burden of proof”, but the “standard of proof”, creating a statutory “benefit of the doubt”, and not in the way this judgement suggests. This conflating of the two standards is an error of law, which ultimately renders the decision unreasonable.

    In F.H. v. McDougall, 2008 SCC 53 (CanLII), [2008] 3 S.C.R. 41 Rothstein J., writing for a unanimous Court, states at para. 54:

    Rowles J.A. was correct that failure by a trial judge to apply the correct standard of proof in assessing evidence would constitute an error of law. The question is how such failure may be apparent in the reasons of a trial judge. Obviously in the remote example of a trial judge expressly stating an incorrect standard of proof, it will be presumed that the incorrect standard was applied. Where the trial judge expressly states the correct standard of proof, it will be presumed that it was applied. Where the trial judge does not express a particular standard of proof, it will also be presumed that the correct standard was applied:

    Trial judges are presumed to know the law with which they work day in and day out.
    (R. v. Burns, 1994 CanLII 127 (SCC), [1994] 1 S.C.R. 656, at p. 664, per McLachlin J. (as she then was))

    Whether the correct standard was expressly stated or not, the presumption of correct application will apply unless it can be demonstrated by the analysis conducted that the incorrect standard was applied. However, in determining whether the correct standard has indeed been applied, an appellate court must take care not to substitute its own view of the facts for that of the trial judge.

  2. Causation is a little mental trick we humans came up with to prevent bad things from happening. When something often comes before something else we call it the cause – it’s a reliable predictor in fact. Philosophy has failed to provide us with a workable description of it beyond that. We don’t know what causation is, but we know what it isn’t : when something happens before something else but there’s another explanation: then it’s not the cause of the effect. Common sense reasoning is lazy like that and looks for the single smoking gun that allows one to easily throw a theory out and find a more plausible explanation.. so to see if something is the cause, we just look around for something else that might also have been the cause. if the exploration turns up null, we are then forced to accept the hypothesis of causation and move on. It’s how business people who know how to stay out of court make decisions every day. In medicine causation is easily disproved by the fact that every ailment has many possible causes, the human body being so complex a mechanism. What makes litigation expensive is that there is too much judicial notice of alternative theories about facts like causation. The parties have to overthink how their theory is going to be disproved in the absence of proper pleadings and by judicial notice. Causation has no definition – it’s just a label for how humans avoid the bad thing from happening by avoiding it’s cause. Causation as a legal requirement must then be understood as a desire to prevent compensation where the person wasn’t prudent or is shifting the blame somehow. So we make him prove causation, and I say that can only happen by disproving the alternative hypothesis (i.e. not caused by something the defendants shouldn’t be held responsible for) as a way of disproving that it really wasn’t somebody elses or his own fault. Imagine if you will a legal system in which each side must plead one or more complete story of events; case management happens and the stories get narrowed down to one for each side. At the trial, each party gets to lead one piece of evidence, which must disprove the other parties story somehow. The judge then decides whose evidence was most INconsistent with the other parties story. Smart business people who stay out of court make decisions exactly like that every day. But for this system to work, the defendant must be required to allege a full story otherwise the plaintiff could never win- there’s no story to disprove. ( Vexatious plaintiffs must be weeded out very early for this obvious reason). And – if a defendant was required to plead a source of causation, it would rid us of arid theories on the subject.