Elliott v. Insurance Crime Prevention Bureau

Elliott v. Insurance Crime Prevention Bureau Court Court of Appeal Date 2005-08-10 Citation 2005 NSCA 115 Docket CA 221995 Judge/Registrar/Adjudicator Bateman, Nancy J. (Honourable Justice) (CA); Cromwell, Thomas A. (Honourable Justice); Freeman, Gerald B. (Honourable Justice) (CA) Document Type Decision Relations Library Sheet - Elliott v. Insurance Crime Prevention Bureau - 2005 NSCA 115 - 2005-08-10 - Library Sheet Decision Content NOVA SCOTIA COURT OF APPEAL Citation: Elliott v. Insurance Crime Prevention Bureau, 2005 NSCA 115 Date: 20050810 Docket: CA 221995 Registry: Halifax Between: Hugh Laird Elliott, Helen Geraldine Elliott, Michael David Elliott and Thomas Augustus Elliott Appellants v. Insurance Crime Prevention Bureau, William A. Wilson, Donald T. Matheson Engineering Limited, Donald T. Matheson, MDS Environmental Services Limited, Dominique Levesque, The Attorney General of Nova Scotia representing Her Majesty the Queen in Right of the Province of Nova Scotia, Francis J. Savage and Marsh Adjustment Bureau Limited Respondents Judges: Bateman, Freeman and Cromwell, JJ.A Appeal Heard: January 11 and 18, 2005, in Halifax, Nova Scotia Additional written submissions May 24, 2005 Held: Appeal dismissed per reasons for judgment of Cromwell, J.A.; Freeman and Bateman, JJ.A. concurring. Counsel: Keith MacKay, for the appellants Michael E. Dunphy, Q.C., for the respondents, Insurance Crime Prevention Bureau and William A. Wilson Douglas Skinner, for the respondents, Donald T. Matheson Engineering Limited and Donald T. Matheson Matthew Williams, for the respondents, MDS Environmental Services Limited and Dominique Levesque Catherine Lunn, for the respondents, Attorney General of Nova Scotia and Francis J. Savage John MacDonald, for the respondent, Marsh Adjustment Bureau Limited Reasons for judgment: I. INTRODUCTION: [1] In this appeal, a novel claim is met by a novel defence. The appellants, who claimed on their fire insurance, say they are owed a duty of care by the respondents who investigated the fire. To my knowledge, no claim like it has ever succeeded in Canada. The respondents, while denying that any such duty exists, also advance a novel defence based on witness immunity. They say that anyone who investigates anything which might result in the investigator becoming a witness in possible future litigation is absolutely immune from liability. If accepted, this would be a considerable extension of witness immunity. So the appeal raises two novel points of law, one relating to the investigator’s duty of care and the other to the scope of witness immunity. [2] These novel points arise out of a straightforward set of facts. The appellants’ home was destroyed by fire. They claimed on their fire insurance policy. The insurer, relying on investigations and reports by the respondents, accused the appellants of arson and denied coverage. The appellants sued the insurer on the policy. (I will refer to this as the insurance action.) The insurer persisted in its arson defence, calling as witnesses, the respondents Savage, Wilson, Levesque and Matheson. The trial judge in the insurance action found that the insurer had not proved that the appellants had set the fire, although he also found that the insurer had not acted in bad faith by denying the claim. The appellants accordingly recovered their loss, interest and costs. However, they did not succeed in their claims for inconvenience, mental distress, aggravated, punitive or exemplary damages. [3] The appellants then sued the respondents who had investigated and reported on the cause of the fire, claiming that they had failed to investigate carefully, knowing that the insurer would rely on their work. (I will refer to this as the negligence action.) Essentially, the appellants claimed the sorts of damages which had been denied to them in the insurance action. [4] Once the pleadings in the negligence action were completed, the respondents applied to Coughlan, J. for an order dismissing the action on either of two grounds: that they did not owe the appellants a duty of care or that the doctrine of witness immunity afforded a complete defence. The application proceeded on an agreed statement of facts. Coughlan, J. found for all respondents on both grounds and dismissed the action. [5] The appellants appeal, raising three main issues: first, did the chambers judge err by applying the wrong test and making unjustified findings of fact; second, did the chambers judge err in finding that none of the respondents owe the appellants a duty of care; and, third, did the judge err in finding that all the respondents are immune from suit under the witness immunity doctrine? [6] I agree with the judge’s overall conclusion that the appellants’ action should be dismissed. The judge was entitled to draw reasonable inferences from the agreed facts and did so. I agree, although for somewhat different reasons than those given by the chambers judge, that none of the respondents owe the appellants a duty of care. While I would not uphold the claims for witness immunity advanced by the respondents, Insurance Crime Prevention Bureau (“ICPB”), Wilson, Marsh, MDS Environmental Services Limited (“MDS”) and Levesque, I agree with the judge that the other respondents are entitled to this protection. In the result, I would dismiss the appeal. II. FACTS AND PROCEEDINGS: 1. The insurance action: [7] On August 30, 1992, the appellants’ home was destroyed by fire. They had fire insurance through the Royal Insurance Company of Canada (“Royal”) and in October filed a proof of loss. Royal rejected the proof of loss on the grounds that they had claimed excessive amounts. In March of 1993, the appellants sued Royal on the policy. Royal defended the action on the basis of arson– that the appellants or someone acting on their behalf had deliberately set the fire. [8] Marsh Adjustment Bureau, one of the respondents, is an independent adjusting firm which was retained by Royal to investigate and adjust the loss. Its involvement started immediately after the fire, and it provided a series of written reports containing its findings and opinions to Royal between late August of 1992 and mid February of 1993. So far as the record discloses, no one from the Marsh firm testified at the trial. The record contains no further particulars concerning Marsh’s mandate or about the substance of its reports to Royal. We know simply that the reports were concerned with the fire and the surrounding circumstances, that Marsh retained various experts to investigate and report as detailed below, and that Royal relied on the reports from all of the respondents in alleging that the fire had been set deliberately. [9] Marsh asked ICPB to investigate, and it started to do so two days after the fire. All we know about ICPB’s mandate is that it was to investigate the fire and the surrounding circumstances relating to the Elliotts’ claim under the policy and that it became involved at the request of the adjusters. [10] William A. Wilson was the senior special agent in charge of the ICPB Halifax office. He personally carried out the investigation and ultimately was called by Royal at the insurance action as an expert witness in relation to the causes, origin and spread of fires. There is no express evidence about whether the ICPB mandate included or contemplated testimony in future litigation or for that matter about the relationship, if any, between the substance of the written reports to Royal and Mr. Wilson’s trial evidence. There is no indication, either in the agreed statement of facts or in the reasons of the trial judge following the trial of the insurance action, as to whether any of the reports provided by ICPB to Royal were placed in evidence at that trial. We do know that ICPB ’s investigation extended until July 27, 1994, that is, until about a month before the trial of the insurance action started in September of 1994. We also know that when Royal rejected the appellants’ proof of loss on December 21, 1992, the investigation with respect to the cause of the fire by police and other agencies, including the ICPB, was still ongoing. [11] On September 2, 1992, two days after the fire, Donald T. Matheson Engineering Limited was engaged by Marsh to investigate whether the fire had an electrical cause. Donald T. Matheson was the president of the firm and personally carried out the assigned investigation and reported to Royal on September 18, 1992. He was called as an expert witness by Royal at the trial of the insurance action. The report he prepared for Royal dated September 18, 1992, as well as a subsequent report prepared in September of 1994, were adduced into evidence. [12] MDS was retained by Marsh in October of 1992 to analyse two samples of material taken from the appellants’ property. Dominique Levesque was the manager of laboratory services for the firm in Halifax. MDS personnel carried out the analysis and Mr. Levesque reviewed and interpreted the data and reported his findings to Marsh on October 21, 1992. Apart from this, there is no indication of Mr. Levesque’s mandate. Mr. Levesque testified on behalf of Royal at the trial of the insurance action as to the presence of a petroleum product in the carpet and soil sample which MDS had analysed. There is no specific indication in the record of what the link is between the report to Marsh and the reports and evidence given at trial, but it is a reasonable inference that they all related to the analysis of the samples. There is no indication in the record as to whether or not the report to Marsh was placed in evidence at the trial. [13] Francis J. Savage, a deputy fire marshall, prepared a report dated September 3, 1992 at the request of the local deputy fire chief. That report, as well as a supplementary one prepared in September of 1994, were placed in evidence at the trial of the insurance action and he testified concerning his opinion as to the cause and origin of the fire. [14] The insurance action was tried before Boudreau, J. over several days between September of 1994 and January of 1995. The judge found that Royal had not proved arson and was therefore liable on the insurance policy for the appellants’ loss. He also found, however, that Royal had not acted in bad faith in denying the claim. In light of that finding, he refused to award the appellants general damages for inconvenience and mental distress, aggravated damages, punitive or exemplary damages or solicitor and client costs. 2. The negligence action: [15] Having failed to recover these damages against Royal, the appellants tried to do so by suing the investigators for failing to exercise proper care in investigating and reporting in respect of the fire loss. They alleged that Royal relied on these investigations and reports, as the investigators knew it would, in deciding not to pay the appellants’ claim. That refusal, it was alleged, caused the appellants inconvenience, embarrassment and mental distress and aggravated the damages which they suffered. [16] The respondents filed defences denying liability. 3. The application under Rule 25: [17] In April of 2002, the respondents applied in Supreme Court chambers under Rules 25.01 and 37.10(a) and (g) for answers to two questions: 1. Did the defendants, or any of them, owe a duty of care to the plaintiffs and, if not, should the plaintiffs’ action against the defendants, or any of them, be dismissed? 2. Are the defendants, or any of them, immune from liability to the plaintiffs with respect to the claims made in this proceeding by reason of the application of the doctrine of witness immunity and, if so, should the plaintiffs’ action against the defendants, or any of them, be dismissed? [18] The relevant text of the Rules is as follows: 25.01. (1) The court may, on the application of any party or on its own motion, at any time prior to a trial or hearing, (a) determine any relevant question or issue of law or fact, or both; . . . 37.10. On a hearing of an application, the court may on such terms as it thinks just, (a) summarily dispose of any cause of action or question arising on the application, and make such order as is just; . . . (g) exercise such jurisdiction and grant any other order as it deems just. [19] The application was supported by a statement of facts agreed upon by all counsel. It recited that the parties wished to submit the two questions to the court for determination and that the agreed statement of facts was provided to allow the court to determine those questions. I have already summarized the content of the agreed statement of facts in the two preceding sections of my reasons. 4. The decision under appeal: [20] Coughlan, J. held that the appellants’ action should be dismissed because the respondents owed them no duty of care. After referring, among other authorities, to Anns v. Merton London Borough Council, [1978] A.C. 728 (H.L.(Eng.)) and Cooper v. Hobart, [2001] 3 S.C.R. 537, the judge found that there was not the required proximity between the respondents and the appellants to give rise to a duty of care. [21] The judge also found that the appellants’ action must fail on account of the doctrine of witness immunity. He said that an expert (whether ultimately a witness or not) is immune from suit by any person if two conditions are present: the expert’s only relationship derives from the judicial proceeding and the expert was retained to investigate a matter which might give rise to possible litigation in which the expert would be a possible witness. (¶ 12 - 14) The judge also concluded that if the individual experts have the protection of witness immunity, their employers or firms could not be vicariously liable. III. ISSUES: [22] The appellants attack the judge’s conclusions on both procedural and substantive grounds. The issues raised may be summarized as follows: Procedural Issues: 1. The judge applied the wrong test on the application. He ought to have treated the application as if it were one seeking summary judgment and ought to have determined that there were genuine issues requiring a trial. 2. The judge erred by proceeding to answer the questions posed to him on the basis of the agreed statement of facts because that statement of facts was inadequate. Specifically, as regards witness immunity, the agreed facts did not resolve the critical factual issues of the status of the respondents as investigators or witnesses and, with respect to the duty of care issue, the agreed facts did not deal with many of the recognized indicators of proximity. Duty of Care Issue: 1. The judge erred in his analysis of the duty of care issue. Specifically, he neglected to consider the issue of foreseeability of harm, ignored numerous indicia of proximity and put undue emphasis on the absence of reliance by the appellants on the respondents. Witness Immunity Issues: 1. The judge’s finding that the individual respondents are entitled to witness immunity is unreasonable. Specifically, he erred (a) in finding that the respondents’ “... only relationship with the appellants derived from the first action ... ” and that therefore the witness immunity rule applied; and (b) in finding that witness immunity applied to the investigations and reports which were completed when they were only “possible” witnesses in possible future litigation. Instead, he ought to have found that witness immunity would only apply if the “dominant purpose” of the investigations and reports was to prepare evidence for litigation. IV. ANALYSIS: A. Procedural Issues: 1. The test under Rule 25: [23] The appellants say that the judge applied the wrong test in resolving this application under Rule 25. They say that this was, in substance, an application for summary judgment. The judge, therefore, ought to have determined whether there were genuine issues of material fact requiring a trial. Had he done so, he would have concluded that there were and dismissed the respondents’ application. [24] I do not accept the appellants’ contention on this point. They do not cite any authority for the proposition that the judge hearing an application under Rule 25 ought to treat it as an application for summary judgment. Respectfully, the appellants’ contention is based on confusion between the test to be applied on a Rule 25 application with the question of when resort to Rule 25 is appropriate. Rule 25 authorizes the court to determine “ ... any relevant question or issue of law or fact, or both ...”. It is not a procedure to determine whether there are issues to be tried, but to resolve questions of law and/or fact. The test for summary judgment is neither appropriate for nor relevant to that exercise. I, therefore, reject the contention that the judge erred by applying an incorrect test on the respondents’ Rule 25 application. 2. The adequacy of the record and drawing inferences from the agreed facts: [25] The appellants’ second procedural point is this: The statement of agreed facts was not adequate to permit determination under Rule 25 of the questions posed. As part of this submission, the appellants say that the judge ought not to have drawn inferences from the agreed facts and that he was limited strictly to the specific facts set out in the agreed statement. To consider these submissions, it will be helpful first to do a brief review of the principles governing applications under Rule 25. [26] Generally, all issues in a case should be heard and determined in one continuous trial. Piecemeal determination of the issues – sometimes called “trial by instalment” – is discouraged. But every rule has its exceptions and sometimes it will be advantageous to decide an issue before trial. Rule 25.01(a) and 25.02 provide a mechanism to allow the court to do this and, where that determination substantially disposes of all or part of the case, to give judgment accordingly. But that authority is discretionary. To determine whether the discretion ought to be exercised in a particular case, two basic issues must be addressed. [27] The first relates to the utility of making the determination otherwise than at a trial. This turns primarily on a cost-benefit exercise to determine whether the likely savings in cost and time justify relaxing the usual rule against fragmenting litigation. The question, as well expressed by Green, J.A. (as he then was) in Miawpukek Band v. Ind-Rec Highway Services Ltd. (1999), 172 Nlfd. & P.E.I.R. 245; N.J. No. 74 (Q.L.)(C.A.) at ¶ 14, is whether there is “some discernable advantage to [making preliminary determinations] rather than dealing with them as part of an overall trial or hearing.” The second main consideration is whether the issue is suitable for determination under Rule 25. The important question is whether the court has an adequate record to make the requested determination. [28] While the Rule is silent on the point, the authorities hold that there generally (although not invariably) ought to be an agreed statement of facts where parties seek a determination of a point of law under Rule 25.01(a): McCallum v. Pepsi-Cola Canada Ltd. et al. (1974), 15 N.S.R. (2d) 27 (T.D.) at ¶ 14 - 15; Curry v. Dargie (1984), 62 N.S.R. (2d) 416 (A.D.) at p. 430; Seacoast Towers Services Ltd. v. MacLean (1986), 75 N.S.R. (2d) 70 (A.D.) at ¶ 18 - 23; Brown v. Dalhousie University (1995), 142 N.S.R. (2d) 98 (C.A.) at ¶ 22 ff ; Fraser v. Westminer Canada Limited (1996), 155 N.S.R. (2d) 347 (C.A.) at ¶ 31 - 36; Binder v. Royal Bank of Canada (1996), 150 N.S.R. (2d) 234 (C.A.) at ¶ 5 - 10; Fortune v. Reynolds (2003, 212 N.S.R. (2d) 94 (C.A.); Knock v. Fouillard, 2004 NSCA 70 (C.A.). The agreement of the parties provides assurance that all the necessary facts are before the court and that they are the real facts. This, in turn, assures that the court has an adequate record to make the determination and that making it will be useful. If the factual underpinnings of the determination of law are not firmly anchored in the real facts of the parties’ situation, the resolution of a legal issue based on them will be not likely be helpful in resolving the parties’ dispute. [29] As noted, the court has a discretion to make or not to make a determination requested under Rule 25. If the court is of the view that making the determination sought by the parties has no discernable advantage or that it cannot properly be made in the context of the application, it retains the discretion not to make it. So, for example, it has been held that a chambers judge erred by proceeding with a Rule 25.01(a) determination in the absence of an agreed statement of facts where one was necessary to resolve the issues: Binder, Knock and Fortune. Similarly, it has been held that the chambers judge erred by making a determination under Rule 25.01(a) when the agreed facts were incomplete: Brown v. Dalhousie. Of course, the adequacy of the agreed facts is a matter of degree. Even after a trial, the facts are sometimes vague and incomplete and so agreed facts of that nature should not necessarily prevent a determination under Rule 25: see e.g. Dawnstar Developments Inc. v. Ross (1989), 89 N.S.R. (2d) 265; N.S.J. No 79 (Q.L.) (C.A.). The court must exercise its discretion to make or not to make the determination in light of the particular circumstances of the case. [30] May the court on a Rule 25 application draw inferences from the agreed statement of facts or is it limited to considering the facts explicitly set out by the parties? In my view, the judge may draw inferences provided that they are reasonable and not inconsistent with the expressly agreed facts. This has been recognized in both civil and criminal cases involving agreed statements of fact: see, for example, Derksen v. 539938 Ontario Ltd., [2001] 3 S.C.R. 398 at ¶ 29; Salmo Investments Ltd. v. Canada, [1940] S.C.R. 263 at 269; Stettler(County No. 6) v. Conibear, [1975] A.J. No. 89 (Q.L.)(D.C.) at ¶ 15; Ponoka-Calmar Oils, Ltd. v. Earl F. Wakefield Co., [1960] A.C. 18 (P.C.); Bell v. Canada (Attorney General) (2002), 200 N.S.R. (2d) 154; N.S.J. No. 10 (Q.L.) (C.A.) ¶ 18, 33; Gordon v. Nova Scotia Teachers’ Union, [1983] N.S.J. No. 307 (Q.L.) (S.C.T.D.); R. v. Cooper, [1978] 1 S.C.R. 860 at 871; R. v. Collard (1987), 39 C.C.C. (3d) 471; M.J. No. 448 (Q.L.)(Man. C.A.). [31] The appellants say that there is an express power to draw inferences in the special case Rule (Rule 27.04) and that the omission of a similar provision in Rule 25 should be taken as excluding that power under the latter Rule. I do not agree. Unlike in Rule 27, there is no mention of agreed facts in Rule 25. It follows that the omission to specify how the agreed facts may be used cannot support a conclusion that the authority to draw inferences from them is excluded. [32] I conclude that the judge did not err in principle by drawing inferences from the agreed facts. The next question is whether he erred as a matter of fact in drawing the two main inferences he did. [33] The judge inferred that when the respondents carried out their investigations, they would have viewed litigation as a possibility and understood that if it ensued, they were possible witnesses. I cannot fault him for drawing this inference which is both reasonable and not inconsistent with the agreed facts. The deputy fire marshall reported within a very few days of the fire that it had been set. Almost immediately after the fire, the adjusters called on the services of the ICPB. It is not much of a stretch to infer that they thought, in the event erroneously, that they might be dealing with a case of arson. They must have known in such a case that civil and criminal proceedings were possible and that if proceedings occurred, they could possibly be witnesses. [34] The judge also inferred that the only relationship between the appellants and the respondents arose from the insurance action. This inference played a role in the judge’s decision both with respect to witness immunity and the duty of care. Respectfully, this inference is not consistent with the agreed facts, and it was not reasonable to draw it. The relationship between the parties arose out of the contract of insurance and the fire. The adjusters were involved to adjust a loss on behalf of the appellants’ insurer. The others, apart from Savage, were also involved by the insurer for the same purpose. Savage was involved to discharge his statutory public duty. There is no agreed fact and no reasonable inference that the investigators were retained to gather evidence for litigation. It is simply not factually accurate to say that the relationship derived solely from their roles as witnesses or from litigation. They all started long before litigation was underway. The most that could be said is that when they started, litigation was a possibility. I, therefore, accept the appellants’ submission that the judge erred when he concluded that the relationship between the parties arose solely from the insurance action. I will address the significance of this error later in my analysis. [35] The appellants also say that the agreed statement of facts was inadequate to permit the judge to consider the duty of care issue. I do not accept this submission. The parties collectively joined in asking the court to determine the duty of care question on the basis of the agreed facts. While the court has a discretion to decline to answer questions submitted if the record is inadequate, I do not think that discretion was engaged here. The record is, in my view, adequate to permit meaningful analysis of both the duty of care and witness immunity questions. The agreed facts (along with the reasons of the judge in the insurance action which are referred to in it) provide a reasonably detailed account of the respondents’ involvement with the appellants’ claim on their insurance policy. Moreover, the roles of adjusters and investigators in fire insurance claims are, in general terms, familiar to lawyers and judges. The appellants agreed to the facts and to the submission of the questions set out in the application. They must accept responsibility if they failed to place before the court other facts which strengthen their case. I do not accede to the submission that the judge ought not to have proceeded with the Rule 25 application on the basis that the agreed facts were inadequate. [36] To summarize: 1. The judge did not err by failing to use the summary judgment test on a Rule 25 application. 2. The court’s power to determine a point of law under Rule 25.01(a) is discretionary. The discretion should be exercised in light of two main considerations: first, whether there is some “discernable advantage” to doing so that outweighs the disadvantages of fragmenting the litigation; and second, whether the procedure available under Rule 25 is suitable for making the determination. 3. Generally, determinations of points of law under Rule 25.01(a) should be made on the basis of agreed statements of fact. 4. The agreed facts must be sufficiently complete to permit the court to make the determination. From counsel’s agreement to the stated facts, the court is entitled to assume that counsel think that the facts are complete for this purpose. If they are not, the court may exercise its discretion and refuse to determine the point. 5. The court may draw reasonable inferences from the agreed facts provided they are not inconsistent with the expressly agreed facts. 6. The judge did not err by inferring that, at the time they began their work, the respondents would have viewed litigation as a possibility and themselves as possible witnesses. He did err, however, by finding that the only relationship between the appellants and respondents arose from the insurance action. 7. The judge did not err in failing to conclude that the agreed statement of facts was inadequate to permit him to address the issues presented. B. The Duty of Care Issue: 1. Overview: [37] The law will impose a duty of care on a person if he or she ought to foresee a risk of harm, is in proximity to those at risk and it is not unwise, for policy reasons, to do so. The chambers judge found that there was no proximity between the appellants and the respondents and therefore that no duty of care should be imposed. The appellants challenge this conclusion. [38] I agree with the judge that there was no proximity between the appellants and the respondent, Mr. Savage but, in my view, he erred in finding that there was no proximity between the appellants and the insurance investigators. However, I conclude that, for policy reasons, no duty of care should be imposed on them. In my view, imposing that duty would be unwise for two reasons. First, the appellants have their remedy against their insurer. Second, the proposed duty would seriously distort the legal relationships among the insured, the insurer and insurance investigators. My reasons for these conclusions follow. 2. The duty of care asserted by the appellants: [39] I will begin the analysis of this issue by outlining the pleaded claims and the nature of the duty advanced by the appellants and then turn to the applicable legal principles. [40] The statement of claim alleges: (1) that the respondents were negligent and acted in a high handed and outrageous manner when they investigated and reported on the cause of the fire; (2) that Royal relied on their negligent work in denying coverage; and, (3) that the appellants suffered damage as a result. They claim damages for out of pocket expenses for legal services required to enforce the insurance policy, general damages for inconvenience, embarrassment and mental distress, aggravated damages, punitive damages, pre-judgment interest and costs. [41] The duty of care advocated by the appellants is this: “ Where property is damaged or destroyed by an insurable hazard, in respect of which a policy of insurance is in effect, persons investigating the loss event and the related claim for indemnity under the insurance policy for the purpose of ascertaining the cause of the event or the liability of the insurer under the policy, owe a duty of care to the person entitled to claim that indemnity.” [42] The proposed duty applies both to investigators who are investigating simply to determine the cause of the loss and those who are concerned with liability under the policy. Mr. Savage (and the Province) are in the former category while the other respondents are in the latter. Unlike the other respondents, Mr. Savage was not investigating on behalf of the insurer or for an insurance purpose. His investigation was pursuant to a discretion conferred by statute for a public purpose. I will therefore consider the question of duty of care in Mr. Savage’s case separately from the respondents who were acting on behalf of the insurer. 3. The appellants’ position - the insurance investigators: [43] The chambers judge found that there was no proximity between the appellants and the respondents. He relied on two factors to support this conclusion: that there was no contractual relationship between them and that the only relationship between them arose from the law suit on the policy. [44] The appellants say the judge erred in three respects, two of which I have addressed already. First, they submit that the agreed facts were not adequate to permit determination of the duty of care issue. However, for reasons which I set out earlier, I do not accept that position. The appellants’ second point is that the judge erred in concluding that the relationship between the appellants and the respondents derived entirely from the action on the policy. I have earlier set out my reasons for accepting this submission. The relationship predates litigation and arose out of the insurer’s administration of the contract of insurance with the appellants. [45] That brings me to the appellants’ third point: that the judge failed to take into account several factors relevant to the proximity analysis and that if he had undertaken a more complete analysis, he would have concluded the proximity issue in favour of the appellants. [46] I agree that the judge’s analysis of proximity was incomplete. He relied exclusively on two factors: that the relationship derived only from the litigation and that there was no contract between the appellants and the respondents. The first factor is, in my respectful view, wrong factually. The second - the absence of a contractual relationship between the appellants and the respondents - is not a telling point: In many, perhaps most negligence cases, there is no contractual relationship between the parties. I conclude that it is therefore our responsibility to consider the proximity issue afresh and make the decision that the chambers judge ought to have made. In my view, that analysis leads to the conclusion that there was proximity between the investigators retained on behalf of the insurer and the appellants. 4. Proximity: the insurance investigators: (i) Legal principles: [47] It will be helpful first to situate the proximity issue in the broader context of a duty of care analysis. In a novel case (that is, one in which a duty of care has not been authoritatively recognized before), the existence of a duty of care is tested by applying the analysis from Anns v. Merton London Borough Council and Kamloops (City) v. Nielsen, [1984] 2 S.C.R. 2. That test was described as follows in Cooper v. Hobart at ¶ 30: 30 In brief compass, we suggest that at this stage in the evolution of the law, both in Canada and abroad, the Anns analysis is best understood as follows. At the first stage of the Anns test, two questions arise: (1) was the harm that occurred the reasonably foreseeable consequence of the defendant's act? and (2) are there reasons, notwithstanding the proximity between the parties established in the first part of this test, that tort liability should not be recognized here? The proximity analysis involved at the first stage of the Anns test focuses on factors arising from the relationship between the plaintiff and the defendant. These factors include questions of policy, in the broad sense of that word. If foreseeability and proximity are established at the first stage, a prima facie duty of care arises. At the second stage of the Anns test, the question still remains whether there are residual policy considerations outside the relationship of the parties that may negative the imposition of a duty of care. ... [48] Foreseeability alone is not sufficient to establish a prima facie duty of care. In addition, the appellants must establish proximity – that the relationship between the respondents and the appellants is such that it is just and fair to impose a duty of care: Odhavji Estate v. Woodhouse, [2003] 3 S.C.R. 263 at ¶ 55; Donoghue v. Stevenson, [1932] A.C. 562 (H.L.) at 581; Cooper at ¶ 32. The nature of this inquiry was summed up by Iacobucci, J. in Odhavji at ¶ 50: [50] Consequently, the essential purpose of the inquiry is to evaluate the nature of that relationship in order to determine whether it is just and fair to impose a duty of care on the defendant. The factors that are relevant to this inquiry depend on the circumstances of the case. As stated by McLachlin J. (as she then was) in Norsk, supra, at p. 1151. "[p]roximity may be usefully viewed, not so much as a test in itself, but as a broad concept which is capable of subsuming different categories of cases involving different factors" (cited with approval in Hercules Managements, supra, at para. 23, and Cooper, supra, at para. 35). Examples of factors that might be relevant to the inquiry include the expectations of the parties, representations, reliance and the nature of the property or interest involved. [49] In Cooper, the Court indicated that sufficiently proximate relationships are identified through the use of categories. While the law may develop new ones, the starting point for the proximity analysis should be the existing categories in which proximity has been found: ¶ 31. If there is no closely analogous category, the question of whether there is a duty of care must be answered by applying the full Anns/Kamloops analysis. [50] The appellants do not suggest this case falls within any of the categories listed in Cooper or which have been authoritatively established elsewhere. Rather, they make two, alternative submissions. First, they say that their situation is sufficiently analogous to cases in which investigators have been found to owe a duty of care so that the duty which they assert here ought to be recognized. Second, they submit that a full Anns/Kamloops analysis leads to the conclusion that the respondents were in proximity to them. I will consider these two points in turn. (ii) Is this case analogous to an existing category? [51] The investigator cases relied on by the appellants do not, in my view, authoritatively establish the existence of a duty of care in the present case. The appellants cite one case involving a firm of adjusters, Halifax Insurance Co. v. Matheson (Donald T.) Engineering (1994), 143 N.S.R. (2d) 161 (S.C.), but it is not pertinent because it involved an action by the insurer against its own investigator. It therefore fell within the well-established category of negligent provision of a service to the party who contracts for it. [52] Two other authorities relied on by the appellants involve investigations, but by the police, leading to criminal charges, a situation quite different from this one. In Beckstead v. Ottawa (City) Chief of Police (1997), 37 O.R. (3d) 62 (C.A.), a police officer was found liable in negligence to a person whom he had investigated and recommended be charged. The focus of the case was on whether the officer was immune from suit and it contains virtually no analysis of the duty of care as we now understand that analysis. The appellants also rely on de Jong v. Midland Police Services Board, [2002] O.J. No. 1629 (Q.L.)(Sup. Ct.) in which it was held that although the officer had not been negligent, there was proximity between him and the person whose conduct he investigated and ultimately charged with careless driving. Again, there is no analysis of the proximity issue as we now understand it. Moreover, there is no attempt in either case to place those situations within one of the traditional categories of negligence law. In my view, these authorities do not show that the present case is sufficiently analogous to one of the categories so as to make unnecessary a detailed analysis of proximity from first principles. [53] The appellants also rely on Odhavji Estate v. Woodhouse, but I do not think that case establishes a category of duty of care pertinent here. The case (for immediate purposes) was concerned with allegedly negligent supervision by the chief of police. The plaintiffs alleged that members of the force’s special investigations unit (“SIU”) had failed to conduct a thorough investigation into the death of a member of the plaintiffs’ family. The inadequacy of the investigation resulted, it was claimed, from the failure of the chief to ensure that members of the force cooperated with the SIU. It was their statutory duty to do so and the statutory responsibility of the chief to see that they did. The case, therefore, concerned the allegedly negligent performance of a statutory duty to supervise the members of the police force. But this is not a category of duty of care which can be transplanted to the present fact situation. [54] I conclude that the duty of care asserted by the appellants is not authoritatively established and is not analogous to an existing category to which they have referred. As a result, I must turn to the appellants’ alternate submission that a duty of care ought to be recognized by applying the Anns/Kamloops analysis. [55] I would note, for clarity, that the appellants have not relied on any analogy to negligent misrepresentation and I have, therefore, not considered it. (iii) The Anns/Kamloops analysis: [56] The first step of the analysis has two branches, foreseeability and proximity. I do not think there can be any doubt that foreseeability is present in this case. To put the appellants’ case at its highest, these adjusters and investigators working for the insurer could foresee that they would create a risk of harm to the appellants if they investigated and reported carelessly concerning the cause of the fire. The “double foreseeability” test was met as both the risk of harm and the victims of it were foreseeable. [57] The second branch is concerned with proximity. As noted, a number of factors are relevant to the question of whether proximity exists: a close causal connection between the act and harm suffered, expectations, representations, reliance, assumed or imposed obligations, physical closeness (“propinquity”) and the property or other interests involved: see Odhavji at ¶ 50; Cooper at ¶ 34; Fraser v. Westminer Canada Ltd. (2003), 215 N.S.R. (2d) 377 at ¶ 79. This is not a checklist, but a collection of relevant considerations which must be weighed and assessed in the particular circumstances of the case. So, for example, in Odhavji, three of these factors combined to satisfy the proximity requirement: an “extremely close” causal connection between the negligent supervision and the resultant injury; a reasonable expectation that a chief of police will be mindful of the injuries that might arise as a consequence of police misconduct; and, a statutory duty imposed on the chief which was in line with the proposed duty of care: ¶ 56 - 58. [58] The most closely analogous case cited to us on the question of proximity in these circumstances is the decision of the New Zealand Court of Appeal in Mortensen v. Laing ( reported along with South Pacific Manufacturing Co. Ltd. v. New Zealand Security Consultants & Investigations Ltd.), [1992] 2 N.Z.L.R. 282 (C.A.). In Mortensen, the Laings claimed under a fire insurance policy on their business premises. Mr. Mortensen was appointed by the insurer to inquire into the circumstances surrounding the fire which he did for several months assisted by others. They advised the insurer that the fire had been deliberately set which, in turn, contributed to the insurer denying coverage. The Laings brought two actions. The first was against their insurer on the policy which the insurer defended on the basis of arson. The second was against Mortensen and another for negligent investigation. The Master refused to strike out the statement of claim against Mortensen but that decision was set aside in the Court of Appeal. [59] In five scholarly judgments, the Court thoroughly canvassed the duty of care issue. Four of the five judges found that there was sufficient proximity but that the duty was negated for policy reasons. I have found their analysis persuasive in the present case. [60] The four judges who found there was proximity relied on a number of factors. One of them was a statute giving a third party affected by the activities of a licensed private investigator such as Mr. Mortensen the right to file a disciplinary complaint based on negligence. Cooke, P. found this to be one of the strongest points. Richardson, J. relied on the licensing requirement for private investigators and the fact that the conduct of investigators toward the target of their investigations was subject to legislative oversight. We have not been referred to any comparable statutory oversight in the present case. However, the Court in Mortensen also relied on a number of other factors, all of which are present in this case and which, to my mind, combine to show that there is the sort of relationship between the insurance investigators and the appellants to make it fair and just to impose a prima facie duty of care. [61] One is the obvious risk of serious harm, a factor relied on by Cooke, P., Richardson, J. and Sir Gordon Bisson. The obvious risk of serious harm also goes, in my view, to the close causal connection between the negligent investigation and the resultant denial of coverage. [62] The same may be said in the present case. The agreed facts stipulate that Royal relied on the reports and investigations in deciding to allege that the fire had been deliberately set. Such reliance would be seen as likely by the insurance investigators because they knew that their reports were for the purpose of providing Royal with the information it needed to deal with the appellants’ insurance claim. In light of that, it would be obvious to the insurance investigators that carelessness on their part would subject the insured to risk of being wrongly accused of arson. The causal link between their allegedly negligent investigation and the insurer’s denial of coverage was, taking these factors into account, both close and direct. [63] Another consideration in Mortensen was the vulnerability of the insured to careless work by the investigators. Cooke, P. noted that by entering into the contract of insurance, the insured submits to the investigation and must rely on the investigators’ probity and carefulness. Richardson, J. pointed out that the insured has no obvious means of protecting him or herself from the investigators’ carelessness. This was also a key consideration for Hardie Boys, J. As he put it at p. 318: The insurer is of course under no obligation to accept the investigator’s conclusions. But the likelihood is that he will, so that an adverse report is likely to lead to declinature. And the insured is unlikely to call in an investigator of his own, certainly not until it is probably too late. The insured is therefore highly dependent on the investigator carrying out his task carefully and responsibly. (emphasis added) [64] This consideration is equally applicable to the present case. As the Elliotts stood in the ashes of their home, the contract of insurance was literally all that they had. It would be obvious to anyone involved in the insurance claim that any person in that situation would be highly vulnerable to the findings of the insurance investigators. [65] In Mortensen, Cooke, P. was of the view that the nature of the insurance contract and the insured’s obligation to cooperate with the investigation contributed to a close relationship between the investigators and the insured. Cooke, P. noted that by accepting the responsibility to investigate, the investigator brings him or herself into a close relationship with the person being investigated: at p. 300. As he put it, “To be investigated is an intrusion into one’s life in return for submitting to which it is not unreasonable to ask for reasonable care.” (at 298) [66] These considerations are also significant in the present case. Given the duty of good faith between an insured and an insurer, the contractual duty to cooperate with the investigation undertaken on behalf of the insurer and the vulnerability of the insured to the investigators, the insured would reasonably expect those retained to act on behalf of the insurer to do so with reasonable care. The circumstances virtually require the insured to rely, as Cooke, P. put it, on the probity and carefulness of the investigation. This virtually enforced reliance would be apparent to the investigator. [67] Finally, I note that there is a degree of physical closeness (or “propinquity”) between the investigators and the insured. The site of the investigation was the appellants’ home and property. It is evident from the reasons of Boudreau, J. at the trial of the insurance action that both Mr. Wilson of the ICPB and Mr. Connors of Marsh were physically present in the appellants’ home very shortly after the fire and Marsh retained MDS to provide an analysis of two samples of material taken from it. [68] The respondents, who are adjusters or investigators retained on behalf of the insurer, have referred to a number of cases supporting their position that no duty of care is owed by these respondents to the appellants. However, only one of them deals specifically with proximity. I will address it now and defer discussion of the others to the next section of my reasons. [69] In Hamilton v. Chris Marion Holdings Limited, [1981] I.L.R. 1-1398 (Ont. H.C.), the insured sued a bailee of her goods and the insurance adjusters. A gun collection was housed in premises heavily damaged as a result of burst pipes. The adjuster recommended the guns be removed and stored by the bailee, and the owner agreed. While at the bailee’s, half of the collection was stolen. The claim against the adjuster included negligence in failing to advise that the policy did not provide full coverage while the guns were stored outside the plaintiff’s premises. Reid, J. dismissed the action, holding that the adjuster was duty bound to give his undivided loyalty to the insuring company and that there was, therefore, no “special relationship” between the adjuster and the insured. [70] With respect, the existence of a contractual duty to one party does not necessarily exclude a tort duty of care in relation to others: see, for example, London Drugs Ltd. v. Kuehne & Nagel International Ltd., [1992] 3 S.C.R. 299. Moreover, there is a significant difference between the duty of care rejected in Hamilton and the one advanced by the appellants in this case. Hamilton rejected a duty of care on an adjuster to give advice concerning the terms of the policy, whereas this case concerns allegedly negligent investigation of the cause of the loss in relation to the insurer’s decision to deny coverage. It seems to me that the case for vulnerability, reasonable expectation of care and virtually enforced reliance is much stronger in a case like this one than it was in Hamilton. [71] To summarize, I conclude that there is sufficient proximity to make it just and fair to impose a prima facie duty of care owed to the insured by an investigator retained to investigate the cause of a loss on behalf of the insurer. There is a clear causal link between the alleged negligence and the damage allegedly sustained. The immediacy and severity of the risk of harm created by negligent investigation and reporting to the insured would be obvious to the investigators. The duty of good faith between the insurer and the insured, coupled with the vulnerability of the insured and the requirement to cooperate with the investigation, give rise to a close connection between the investigator and the insured as well as a reasonable expectation by the insured that the investigator will take reasonable care. Moreover, given that the focus of the investigation was the appellants’ family home, there is a degree of physical closeness (propinquity) between the insured and the investigators. In my respectful view, the judge erred in finding no proximity between the appellants and the insurance investigators. 5. Proximity: Mr. Savage: [72] The judge also found that there was no proximity between Mr. Savage and the appellants, a conclusion now challenged by them on appeal. Mr. Savage asserts that not only was there no proximity, but that he should not have foreseen the risk of harm to the appellants which could result from a negligent investigation. To consider the question of foreseeability and proximity, it will be helpful to review both the facts and the legislative framework in relation to Mr. Savage’s involvement. [73] As noted earlier, Mr. Savage was a deputy fire marshall who investigated the fire following a report from the local deputy fire chief. He was in the Elliotts’ house by mid-morning the day of the fire and completed his report four days later. [74] At the time, the fire marshall’s duties and responsibilities were set out in the Fire Prevention Act, R.S.N.S. 1989, c. 171 as amended. These included the duty to investigate and hold inquiries respecting the cause, origin and circumstances of such fires as he may deem advisable: s. 8(3)(a). As counsel for Mr. Savage points out, this section gave a discretion as to which fires would be investigated. An inquiry, as opposed to an investigation, was a more formal process spelled out in s. 12 and ff. It could involve summoning witnesses and result in a formal report stating an opinion as to the cause and origin of the fire and whether it appeared to have been of incendiary origin: s. 12(2). So far as we know from the record, no inquiry was held in this case. There were also provisions in the Act dealing with fire insurance companies and adjusters. Under s. 10, every licensed fire insurance company was to provide the fire marshall with a statement respecting every fire in which it was interested as insurer setting out the names and addresses of the owner and occupier, the amount of the loss, the probable cause of the fire and other specified information. When the fire appeared to be of suspicious origin, the company was required to make a preliminary report within three days of the fire: s. 10(2). Under section 11, every adjuster of a fire insurance claim was also obligated to report to the fire marshall within three days of the completion of the adjustment. [75] In my view, foreseeability of the risk of harm to the appellants was established here. It would be obvious to any experienced person that stating a fire was incendiary in origin could contribute to a denial of coverage on a fire insurance policy. The statute specifically required details of insurance coverage to be provided to the fire marshall and, where the fire was of suspicious origin, within three days of the fire: s. 10(1) and (2). The agreed facts make it clear that Royal had Mr. Savage’s report and relied on it in denying coverage. There is no suggestion that Mr. Savage was not aware that Royal would see his report. Of course, the decision to deny coverage is for the insurer. However, it is foreseeable that a report stating that the fire was incendiary in origin would likely elevate the risk of the insurer denying coverage. Moreover, if the report were wrong because it was negligently prepared, the risk of damage would be clear. The fire which he investigated was at a family home. Mr. Savage was in the home the morning of the fire and removed samples from it. So the risk to the appellants was foreseeable by Mr. Savage. [76] Turning to the issue of proximity, I agree with the chambers judge’s conclusion that proximity has not been established here. There is no relationship between Mr. Savage and the insurer. Unlike in the situation of the other investigators who were working for or on behalf of Royal, the contract of insurance does not contribute to any special vulnerability of the appellants in relation to Mr. Savage or to any heightened expectation on their part that he would take reasonable care. The causal link between Mr. Savage’s alleged negligence and the appellants’ loss is more indirect than in the case of the other investigators. Unlike them, Mr. Savage did not prepare his report for insurance purposes. While there is, as in the case of the other investigators, a degree of physical closeness (propinquity), the other factors I have noted, to my mind, combine to show that the relationship between the appellants and Mr. Savage is not such that it would be fair and just to impose on Mr. Savage a duty of care to them in relation to his investigation. [77] I conclude that the judge was right to find that there was no proximity between Mr. Savage and the appellants and that he, therefore, owed the appellants no duty of care. I would dismiss the appeal from that finding and uphold the judge’s order dismissing the action as against him as well as against the Province which is alleged to be vicariously liable for Mr. Savage’s acts. 6. Policy considerations in relation to duty of care – the insurance investigators: [78] In the case of the insurance investigators, I have concluded that there is a prima facie duty of care. I turn, therefore, to the second main step in the Anns/Kamloops analysis, that is whether that prima facie duty of care should be negated for policy reasons. [79] As summed up in Odhavji at ¶ 51, this stage of the analysis is not concerned with the relationship between the parties but with the effect of recognizing a duty of care on other legal obligations, the legal system and society more generally. The Supreme Court in Cooper made the point that most cases will be decided at the first step of the analysis, involving foreseeability and proximity rather than on the basis of overriding policy concerns: at ¶ 39. I would add, agreeing with Justice Linden, that to negate a prima facie duty of care, the policy concerns must be serious and overriding in nature given that the effect of invoking them will be to deny a party compensation for what would otherwise be the breach of a duty: Allen M. Linden, Canadian Tort Law, 7th ed. supp. (Toronto: Butterworths, 2004) at 19 - 20. [80] In the present case, two main policy considerations show that it would be unwise to recognize the duty of care proposed by the appellants. First, persons in the appellants’ situation have their contractual remedy on the policy, which remedy includes, in a proper case, a claim for aggravated and punitive damages. While this remedy will not always be complete for persons in the appellants’ situation, it is a substantial and meaningful remedy making the case for some new form of liability less compelling than it would absent such a remedy. Second, imposing the proposed duty would distort the legal relationships among the insurer, the insured and the investigators and could potentially undermine the ability of the insured and the insurer to properly deal with insurance claims. I will discuss each of these policy considerations in turn. [81] The availability of an alternate remedy has been recognized by the Supreme Court of Canada as one policy consideration which may negate an extension of liability in negligence: see, e.g. Cooper at ¶ 37; Odhavji at ¶ 60. This consideration figured prominently in the Mortensen case in which the court was unanimous that no duty of care should be imposed. Significantly, four of the five judges reached this result on the basis that policy considerations negated the prima facie duty they found to exist. [82] In the present case, the appellants’ success in their insurance action demonstrates that the alternate remedy is substantial. Moreover, as recognized by Whiten v. Pilot Insurance Co., [2002] 1 S.C.R. 595 a breach by the insurer of the contractual duty of good faith may, in an appropriate case, sound in punitive damages. Thus, this case is quite different than either Odhavji or Haskett v. Equifax Canada Inc. (2003), 63 O.R. (3d) 577 (C.A.) in which a complaints process or regulatory remedies, short of the right to sue, were found insufficient alternate remedies to negate the prima facie duties of care found to exist in those cases. [83] It is true that not all heads of damage may be available in an action on the contract, even taking into account a claim for punitive damages and remedies for bad faith. Cooke, P. recognized this in Mortensen, but nonetheless considered that the availability of the remedy on the contract was a significant policy reason not to expand tort liability: at 303 - 4. This view was shared by all the other judges: Casey, J. (p. 314), Hardie Boys, J. (at 319), Richardson, J. (at 308) and Sir Gordon Bisson (p. 326). [84] I conclude that the availability of a remedy in an action on the contract which, in Canadian law, includes the potential remedy of damages for breach of the insurer’s duty of good faith, is a substantial, if not always complete, alternative remedy which strongly works against the expanded liability in negligence proposed by the appellants. [85] I turn to the second policy consideration. Recognition of the proposed duty of care would distort the legal relationships among the insured, the insurer and the investigators and have serious negative effects for both insureds and insurers. This view is based on several considerations, most of which were noted and relied on by the judges in Mortensen. [86] The insured and the investigators all have defined their relationships with the insurer contractually. As Richardson, J. put it in Mortensen, sound public policy requires that where contracts cover the two relationships, those contracts should ordinarily control the allocation of risk unless special reasons are shown to warrant a direct suit in tort: at 308. Of course, one such important reason would be the absence of any remedy against the insurer, but that cannot be advanced by the appellants here. [87] Consider the situation as between the insurer and its investigators if the duty of care proposed by the appellants were recognized. Using this case as an example, the damages which the appellants now claim against the investigators were denied to them in their action against the insurer. The possibility that those retained by the insurer could have greater liability to the insured by virtue of simple carelessness than the insurer has under its contract and duty of good faith, to my mind, sets up an incoherent scheme of liability which is inconsistent with the primacy of the insurer’s obligation to the insured. [88] This incoherence is evident from the fact that, if permitted, a claim in negligence against the insurance investigators could be used by an insured as a means to avoid determination between the insured and the insurer of the real issue of what caused the loss. While this is not a concern in this case – the appellants successfully sued the insurer on the policy – the facts in the Mortensen case show the real potential for such indirectness, an indirectness which Cooke, P. rightly thought should be discouraged. When assessing the policy considerations, one must pay attention to the implications for other situations and not simply the single case of the appellants. [89] There is another way in which legal incoherence would result from recognizing the proposed duty of care. Permitting liability in negligence would interfere with settled principles in other areas of the law. For example, there is the potential for interfering with the law of defamation. A false allegation of arson, if published, is defamatory. But an investigator reporting to his principal would generally benefit from a qualified privilege making it necessary for the plaintiff in a defamation case to prove malice, not simply a breach of a duty to be careful: see e.g., Mortensen at 320. The same issues arise in the case of actions for malicious prosecution. Reports of arson not infrequently lead to prosecutions, but the tort of malicious prosecution requires proof of malice or improper purpose, not simple carelessness: see for example, Cooke, P. at 301 and 303; Richardson, J. at 309; Hardie Boys, J. at 319; G. H. L. Fridman, The Law of Torts in Canada, 2nd ed. (Toronto: Carswell, 2002) at 848. There is also some potential for interference with litigation privilege. As pointed out by Cooke, P. at 303, many investigators’ reports will be covered by litigation privilege and it would seem unfortunate to grant the privilege to the insurer in an action on the policy, but effectively remove it by permitting a negligence suit against the investigators. [90] I am also concerned about other difficulties that would arise if, as the appellants propose, insurance investigators were to be subject to both the contractual obligations to the insurer and a duty of care in negligence to the insured. The nature of the investigator’s engagement is defined by the contract and could be different from a duty imposed by the law of negligence: Casey, J. at 314. The insurer is entitled to contract with the investigator for a particular type of service to be performed within parameters of cost and time dictated by the circumstances and the parties’ choice. The express and implied duties arising from that contract of service may or may not be the same as the proposed duty of an investigator to take reasonable care for the interests of the insured. There is a real possibility of conflict between the contractual and the tortious duties which could undermine the ability of the insurer to administer insurance contracts in a cost effective and expeditious manner. [91] This concern also operates in the opposite direction. If an expert retained by an insurer for purposes of administering the contract owes a duty of care to the insured, it is hard to see why a similar duty would not apply to experts retained by the insured for the purpose of persuading the insurer to pay. In short, I would find it difficult to confine the proposed duty of care to situations even roughly analogous to the very sympathetic circumstances of this case. This concern was noted by Richardson, J. in Mortensen when he observed that the duty of care contended for could not reasonably be confined to insurance investigators and that its “... ambit would be inherently expansive and unacceptably indeterminate”: at 309. He said, and I agree, that it would be difficult to justify not extending the proposed duty of care to anyone who, in the course of a contractual engagement, carelessly investigates and reports on the conduct of a third party: at 309. I do not regard this concern as either speculative or trivial. For example, if this duty of care were recognized, why would it not apply to permit an insurer to sue the insured’s family doctor who, on behalf of the insured, prepared a doctor’s note in support of an application for insurance benefits? Such potential liability could well inhibit the preparation of routine medical letters and reports which are issued at nominal cost in the thousands every day for insurance and employment purposes. That such a result would be undesirable is, to me, self-evident. [92] Liability in negligence of investigators to both the insured and the insurer would create problems of divided loyalties on the part of investigators. When investigating a fire, the relationship between the investigator and the insured is at least somewhat adversarial and becomes, as Casey, J. says, “a direct confrontation if arson is suspected”: at 314. Citing the principle that one cannot serve two masters, he expressed the view that imposing a duty of care in relation to the insured might inhibit the investigators’ ability to discharge their primary duty to the insurer. I agree with this concern. [93] I noted earlier that the duty of care contended for by the appellants is not supported by authority. In fact, the weight of Canadian authority is firmly against imposing on insurance adjusters and investigators any duty of careful investigation or advice in favour of the insured: Sulzinger v. C.K. Alexander Ltd., [1972] 1 O.R. 720 (C.A.); Hamilton v. Chris Marion Holdings Ltd., [1981] I.L.R. 1-1398 (Ont. H. Ct.); Bullock v. Trafalgar Insurance Co. of Canada (1996), 9 O.T.C. 245 (Gen. Div.); Abbasi v. Portage La Prairie Mutual Insurance Co., [2004] 3 W.W.R. 665 (Alta. Q.B. Master); but see contra Bush v. Continental Insurance Co., [2002] I.L.R. I-4033. This appears to be the case in New Zealand and England as well: Mortensen; Graham v. Entec Europe Ltd. (t/a Exploration Associates), [2003] E.W.J. No. 4259 (Q.L.)(C.A.) at ¶ 33 - 34. Although opinion is divided in the United States, my impression is that the predominant view there too is that such a duty of care should not be recognized: see, for example, Sanchez v. Lindsey Morden Claims Services, Inc., 72 Cal. App. 4th 249 (C.A. Cal.) (1999); Charleston Dry Cleaners & Laundry Inc. v. Zurich American Insurance Co., 355 S.C. 614 (2003) (S.C.S.C.); Barry Lindahl, 3 Modern Tort Law, 2nd ed. 2005 cum. supp (Thomson, 2004) ¶ 26:59; Couch on Insurance, 3rd ed., part VIII c. 208, but contra see for example Morvay v. Hanover Ins. Co., 127 N.H. 723 (N.H.S.C.) (1986); Brown v. State Farm & Cas. Co., 2002 OK Civ App 107 (Okla. C.A.)(2002). [94] There are two roughly analogous situations in which liability has generally not been recognized. [95] The first is concerned with claims by parents (and/or children) against professionals who have been retained by opposite parties or statutory authorities to assess children for child welfare or custody proceedings. A leading example is X (Minors) v. Bedfordshire County Council, [1995] 2 A.C. 633 (H.L.(Eng.)). In rejecting a proposed duty of care on a social worker and psychiatrist who had been retained by a local authority discharging statutory child welfare responsibilities, Lord Browne-Wilkinson said (p. 752-54): In my judgment in the present cases, the social workers and the psychiatrist did not, by accepting the instructions of the local authority, assume any general professional duty of care to the plaintiff children. The professionals were employed or retained to advise the local authority in relation to the well being of the plaintiffs but not to advise or treat the plaintiffs. . . . Even if, contrary to my view, the social workers and psychiatrist would otherwise have come under a duty of care to the plaintiffs, the same considerations which have led me to the view that there is no direct duty of care owed by the local authorities apply with at least equal force to the question whether it would be just and reasonable to impose such a duty of care on the individual social workers and the psychiatrist. [96] The Bedfordshire case has been followed in Canada: H.V.K. v. Children’s Aid Society of Haldimand-Norfolk (2003), 37 R.F.L. (5th) 348; O.J. No. 1572 (Sup.Ct.); Branco v. Sunnybrook & Women’s College Health Sciences Centre, [2003] O.J. No. 3287 (Q.L.)(Sup.Ct.). [97] The second roughly analogous situation involves claims against medical assessors retained by insurers and employers to report to them concerning another person’s physical or psychiatric condition. Several of the cases were reviewed by Spence, J. in Branco v. Sunnybrook & Women’s College Health Sciences Centre. He concluded that “... the primary duty owed by the doctor was to the plaintiff’s employer and the only duty the doctor owed to the plaintiff was to do no harm to him in the course of conducting the IME”: at ¶ 13. See also Johnson v. State Farm Mutual Automobile Insurance Co., [1998] O.J. 3139 (Q.L.) (Gen. Div.); Lowe v. The Guarantee Company of North America (2003), 67 O.R. (3d) 124 (Sup.Ct.); but see Lynch v. Appell, [2001] O.J. No. 4740 (Q.L.) (Sup.Ct.). [98] While most of these cases do not contain a detailed Anns/Kamloops analysis, the results are not supportive of the duty of care asserted by the appellants in this case. [99] In saying this, I do not ignore the fact that courts have been more receptive to claims by insureds against adjusters based on alleged breaches of a duty of good faith, of a fiduciary duty or on intentional procurement of breach of contract: Spiers v. Zurich Insurance Co. (1999), 45 O.R. (3d) 726 (Sup.Ct.); Pilat v. Federation Insurance Co. of Canada, [2003] S.J. No. 471 (Q.L.)(Q.B.); Wigmore v. Canadian Surety Co., [1994] 9 W.W.R. 521 (Sask. Q.B.); Walsh v. Nicholls (2004). 241 D.L.R. (4th ) 643 (N.B.C.A.); but see, contra, Burke v. Buss, [2002] O.J. No. 2938 (Q.L.) (Sup. Ct.); Curtis v. State Farm Mutual Automobile Insurance Co., [2003] O.J. No. 3064 (Q.L.)(Sup.Ct.). I emphasize that the present case is concerned only with liability in negligent investigation and reporting. I express no view in relation to claims based on other causes of action. 7. Conclusion on Duty of Care: [100] Mr. Savage owed the appellants no prima facie duty of care in negligence law because he was not in a relationship of proximity with them. While the insurance investigators owed the appellants a prima facie duty of care, that duty should not be recognized for substantial reasons of policy. I would, therefore, uphold the judge’s conclusion that none of the respondents owed the appellants a duty of care and dismiss the appeal. C. Witness Immunity: 1. Overview: [101] My conclusion concerning the duty of care is sufficient to determine the result of the appeal. However, the judge also found that the appellants’ action should fail because all of the individual respondents had a complete defence based on witness immunity. This aspect of the case has been fully argued and I will address it in case I have erred in my conclusion concerning the duty of care. [102] Witnesses are immune from civil liability for what they say and do in a judicial or quasi-judicial proceeding. This is the core of witness immunity. Outside that core, the immunity may also extend to things witnesses (and even potential witnesses) say and do out-of-court, provided that the extension is necessary in order to make the protection of testimony effective. But how far the immunity extends to things said and done out-of-court is a grey area. This case falls within that grey area. It is concerned with whether witness immunity protects the respondents’ statements made out-of-court to the insurer while they were investigating the cause of the fire. [103] The respondents Matheson, Savage, Levesque and Wilson investigated and reported on the cause of the fire and then much later testified as expert witnesses at the trial of the insurance action. They claim that witness immunity protects them from potential liability in negligence because their only relationship with the appellants arises from the insurance action and they were, at the time they investigated and reported, possible witnesses in possible future litigation. Marsh also claims witness immunity even though no one from the firm actually testified. Witness immunity applies, Marsh submits, because it extends to investigators retained to investigate a matter which may possibly give rise to litigation. [104] The chambers judge accepted these broad claims of witness immunity. He held that an investigator investigating something that may possibly lead to litigation in which the investigator is a possible witness is absolutely immune from a suit based on the investigation brought by a party whose only relationship with the investigator arises out of the litigation. [105] The appellants attack this conclusion. They say that the judge’s finding is unreasonable and that he erred in finding that: (a) the respondents’ “... only relationship with the appellants derived from the first action ... ” and that therefore the witness immunity rule applied; and (b) witness immunity applied to the investigations and reports which were completed when they were only “possible” witnesses in possible future litigation. Instead, he ought to have found that witness immunity would only apply if the “dominant purpose” of the investigations and reports was to prepare evidence for litigation. [106] As noted earlier, I agree with the appellants that the judge erred by finding that their only relationship with the respondents arose from the insurance action. I also agree with them that he cast the net of immunity much too widely. For the reasons which I will develop, my view is that witness immunity does not apply to everyone who investigates anything which may result in litigation in which they may be possible witnesses. As one judge succinctly put it, an expert’s professional negligence is not immunized whenever the expert later relies on it in court: Carnahan v. Coates (1990), 71 D.L.R. (4th) 464 (B.C.S.C.) 477. The immunity applies only to those statements made for the purpose of preparing evidence for legal proceedings and to those statements which must be protected in order to make the immunity for testimony effective. Applying the correct principles as I understand them, the immunity claims by Savage and Matheson should be upheld, but the immunity claims of Wilson, Levesque and Marsh should be rejected. [107] To develop my reasons for these conclusions, I will set out the immunity claims advanced by each respondent, the settled principles of law in relation to witness immunity and their rationale. I will then describe the analysis required for a claim of witness immunity and apply it to the immunity claims advanced by the respondents. 2. The Immunity Claimed by Each Respondent: [108] I turn first to the immunity claimed by the respondents who actually became witnesses at the trial of the insurance action. Wilson (of ICPB), Matheson (of Matheson Engineering) and Levesque (of MDS) say that the immunity extends to statements of a witness made in court and in the preparation of the evidence which is to be so given. They assert that the immunity applies to them as witnesses from the moment that they began their work because, from the outset, they were possible witnesses in possible future litigation. Wilson claims that his investigation, preparation of reports and testimony at trial were linked and interwoven so that it is not possible to distinguish between his role as a witness and his role as an investigator. All these respondents submit that immunity applies because their only relationship with the appellants derived from their role as potential expert witnesses. [109] The claim of immunity on behalf of Savage is based more explicitly on the fact that his initial report, as well as a supplementary report prepared much later, were placed in evidence. It is submitted that, as a result, the initial report prepared in September of 1992, as well as his trial testimony in 1995, are immune from suit. Savage also relies on the principle that the immunity applies because he had no independent relationship with the appellants apart from the fire investigation and as well on the proposition that it is irrelevant that his initial report was prepared before the Elliotts started their action on the policy. [110] Thus, the immunity claimed by the respondents who testified at the insurance action trial may be expressed as follows: The immunity applies to a witness whose only relationship with the appellants derives from the litigation and extends to all reports and investigations undertaken while the investigator is a possible witness in possible future litigation. [111] I turn next to the claim of immunity advanced by Marsh. No one in the Marsh organization was called to testify at the trial of the insurance action. Marsh submits that the immunity applies to all investigators, whether called as witnesses or not, who are retained to investigate a matter which may become the subject of litigation. 3. Basic Principles of Witness Immunity: [112] The witness immunity rule is part of a larger immunity which applies to participants in judicial or quasi-judicial proceedings. As Raymond E. Brown puts it in 2 The Law of Defamation in Canada, looseleaf updated to 2004 Rel. 4, (Toronto: Carswell, 1999) at ¶ 12.4(1) and 12.4(4)(a), “An absolute privilege or immunity attaches to those communications which take place during, incidental to, and in the processing and furtherance of, judicial or quasi-judicial proceedings. No action for libel or slander will lie for words spoken or written during the ordinary course of those proceedings. ... The protection of this privilege extends to all the participants in the judicial or quasi-judicial proceeding including the judge, jury witnesses, parties and their counsel.” See also Watson v. M’Ewan, [1905] A.C. 480 (H.L. (Sc.)); Halls v. Mitchell, [1928] S.C.R. 125. [113] While the immunity is most familiar in defamation cases, it also bars other causes of action. Halsbury says that: “A witness is protected from civil proceedings in respect of the evidence which he gives in judicial proceedings, and in respect of things said or done in the course of preparing evidence for such proceedings. The protection is against actions of any sort, and is not limited to actions for libel and slander.”: Halsbury’s Laws of England, 4th ed., vol. 17 (London: Butterworths, 1976) at ¶ 261 (emphasis added). Similarly, in Martini v. Wrathall, (1999), 180 N.S.R. (2d) 38 (C.A.) (at ¶ 6) and Horn Abbot Ltd. v. Reeves (2000), 189 D.L.R. (4th) 644 (¶ 15), this Court approved the proposition that witness immunity applies to all tort actions. (See also Smith v. Canadian Pacific Railway (2003), 181 Man. R. (2d) 150 (Q.B. Master). It is not disputed in this case that the immunity applies to negligence actions. [114] The immunity applies to words said or conduct performed on a protected occasion, the protected occasion being a judicial or a quasi-judicial proceeding: Brown at ¶ 12.4(2). Thus, publishing defamatory words is not actionable if done in the course of judicial or quasi-judicial proceedings. It is critical to understand that it is not the nature of the conduct or the words which is the focus of the immunity, but the occasion on which the words are said or the conduct is performed. Saying exactly the same words will be either actionable or not depending on the occasion on which they are said. This is true whether the immunity is advanced as a defence in a defamation case or in other types of actions. The immunity applies only to a protected occasion. [115] The core of witness immunity is well established by authority. A witness has immunity in respect of what he or she says and does in court or in testimony before a quasi-judicial proceeding: Darker v. Chief Constable of West Midlands Police, [2001] 1 A.C. 435 (H.L.(E.)) at 463. The immunity also extends to a statement made by a witness if the statement is as to the nature of the evidence the witness can give and it is made to a professional person preparing the evidence to be presented in court: Watson v. M’Ewan; Halls v. Mitchell. 4. Rationale of the immunity: [116] The absolute immunity of witnesses exists because it is necessary to protect the proper functioning of the administration of justice. “Strict necessity” said Duff, J. in Halls, “ is the basis of the privilege”: at 145. Or, as The Lord Chancellor said in Watson v. M’Ewan, taking proceedings in court and compelling witnesses to attend are necessary for the administration of justice and as a result witnesses must be protected from actions in respect of their evidence: at 487. Thus, the core of the immunity protects giving testimony in court – of things said and done on that ‘occasion’. Extensions beyond that ‘occasion’ are made when necessary in order to make the immunity for testimony effective. [117] The immunity, of course, does not exist to protect wrongdoers, but it will sometimes do so. For the immunity to be effective, witnesses must be protected from all law suits, not only unmeritorious ones. This protection of witnesses from the risk of suit is seen as more important than righting a wrong in a particular case. Moreover, to achieve its objective, the immunity must be clear: People must know in advance whether they are protected or not: Darker v. Chief Constable of West Midlands Police and Brown, ¶ 12.4(4)(e). The immunity, therefore, is a blunt instrument, barring all claims in the interests of the broader administration of justice. [118] The absolute immunity of witnesses, therefore, negates the usual rules of civil liability. But such a sweeping exemption from liability is only justified when demonstrably necessary to achieve important objectives: Brown, page 9 -5. As has been said, “... the general rule is that the extension of absolute privilege is ‘viewed with the most jealous suspicion, and resisted, unless its necessity is demonstrated’”: Taylor v. Director of the Serious Fraud Office, [1999] 2 A.C. 177 (H.L., Eng.) at 214 per Lord Hoffman quoting with approval the High Court of Australia in Mann v. O’Neill (1997), 71 A.L.J.R. 903 at 907. Thus, the test for the extension beyond the well-settled core of the immunity is a strict one: Necessity must be shown. Consistent with the usual rules about the burden of proof, the burden of proving a defence based on immunity is on the defendant: Brown at section 12.4(7). [119] Two main policy considerations support the necessity of witness immunity. First, it is critical that witnesses be willing to tell the whole truth as they see it, free of concern about consequences to themselves. The need for both candour and cooperation means that witnesses should be protected from civil liability and the risk of vexatious litigation in relation to their testimony. The rigours of cross-examination and the risk of prosecution for perjury are seen as sufficient checks on the untruthful witness. Second, the immunity protects the substance of the evidence from collateral attack in other proceedings. As Lord Wilberforce put it in Roy v. Prior, [1971] A.C. 470 (H.L. (Eng.)) at 480, the immunity exists “... to avoid a multiplicity of actions in which the value or truth of [the witness’] evidence would be tried over again.” See also Carnahan v. Coates, supra at 475 - 476. 5. Analysing a claim for witness immunity: [120] To determine whether a defendant is entitled to witness immunity, four questions must be answered: 1. What is the conduct that forms the basis of the appellants’ cause of action – that is, what is the “gist and essence” of the appellants’ claim? 2. What is the scope of the claimed immunity and does the “gist and essence” of the appellants’ claim fall within it? 3. Is the scope of the claimed immunity settled by authority? 4. If not, does the claimed immunity meet the test of necessity? [121] I will address the immunity claimed in this case and the arguments raised on appeal within the framework of these four questions. 6. What conduct forms the basis of the appellants’ action? [122] In order for the immunity to apply, the action must be based on conduct carried out on a protected occasion. Therefore, one must characterize the claim to decide whether it is based on, or otherwise sufficiently connected to, some protected occasion. The characterization of the claim for this purpose is not concerned so much with the legal category to which the claim is assigned as with the conduct giving rise to it. One must determine the claim’s “gist and essence” or what has been referred to in another context as its “essential character”: Roy v. Prior, supra; Weber v. Ontario Hydro, [1995] 2 S.C.R. 929. [123] Roy v. Prior is a good example of how this is done. In that case, the defendant solicitor, while acting for an accused in a criminal case, had issued a witness summons to the plaintiff. The plaintiff did not appear, and after the defendant had stated on oath that the plaintiff was evading service, he was arrested on a warrant issued by the presiding judge. The plaintiff sued for malicious arrest. The House of Lords held that the suit was not barred by witness immunity because it was not brought on or in respect of the giving of false evidence. Noting that the court must look to the “gist and essence of the claim” in order to determine whether it was one “in respect of” the conduct protected by immunity, the House determined that the defendant was not being sued on or in respect of the evidence he gave in court, but for his instigation of the plaintiff’s arrest maliciously and without reasonable cause. It was held that the action should not be defeated simply because one step in the course of the abuse of process involved the giving of evidence: at 479-80 (A.C.). [124] Determining the ‘gist and essence’ of a claim is more complicated in a case like this one involving alleged professional negligence by expert witnesses. Such witnesses almost invariably have two roles - advising their client and giving evidence in court. The core principles of witness immunity apply if the gist and essence of the claim is with respect to the evidence in court. For example, in Carnahan v. Coates, an adverse party sued an expert who had been retained to prepare a report for, and to testify in, ongoing child access proceedings. Huddart, J. (as she then was) carefully assessed the plaintiff’s claim and concluded that, however it was expressed, it was in essence a claim based on negligent testimony: at 478. Similarly, in Martini v. Wrathall, the action was based on an allegedly false affidavit and, therefore, barred by immunity. [125] What is the “gist and essence” of the appellants’ claims in this case? To answer this question, I have reviewed the appellants’ statement of claim, their Answers to Demand for Particulars in response to demands from Savage, MDS and Matheson and the agreed statement of facts. As the nature of the claims set out in the statement of claim was expanded by the additional particulars which the appellants delivered, I have assessed the claim as pleaded in the statement of claim and then as expanded in the particulars. [126] I would summarize the “gist and essence” of the appellants’ claim as follows: From the statement of claim: 1. Negligent reporting by MDS, Matheson and Savage up to October 21, 1992, that is prior to Royal’s rejection of the proof of loss; 2. Negligent investigation and reporting by ICPB and Marsh up to just before the Elliotts started their insurance action (in the case of Marsh) or shortly before the trial on the insurance action began (in the case of ICPB). From the particulars: 3. Negligent failure to retract or qualify opinions by MDS, Matheson and Savage once presented with the appellants’ expert evidence. [127] The analysis leading to this conclusion is set out in Appendix I of my reasons. 7. Does the gist and essence of the appellants’ claim fall within the scope of the immunity claimed by the respondents? [128] The immunity claimed is for statements made by investigators during the investigation of something that might possibly result in litigation and in which the investigators are possible witnesses. The question now is whether the gist and essence of the appellants’ claim falls within it. [129] The authorities have described in various ways the required degree of connection between the cause of action and the conduct for which immunity is claimed: that the action be “in respect of” or “derived from” a protected activity; that such conduct “constitutes” the cause of action; that the cause of action be “based upon” or “based solely upon” the protected conduct or that such conduct be an essential element of it: Marrinan v. Vibart, [1963] 1 Q.B. 528 (C.A.); Horn Abbot v. Reeves, supra; Samuel Manu-Tech Inc. v. Redipac Recycling Corp. (1999), 124 O.A.C. 125; O.J. No. 3242 (Q.L.)(C.A.); Surzur Overseas Ltd. v. Koros, [1999] 2 Lloyd’s L.R. 611 (C.A.) and Allen v. Morrison, [2004] O.J. No. 5222 (Q.L.)(Sup. Ct.) . [130] In the present case, the gist and essence of the appellants’ claims is the respondents’ investigation and reporting at various times. The broad immunity asserted by the respondents would apply to all those activities. I conclude, therefore, that conduct which is the “gist and essence” of the appellants’ claims falls within the scope of the immunity which the respondents assert. 8. Is the scope of the claimed immunity established by authority? [131] The respondents have two central contentions. The first, figuring most prominently in the submissions on behalf of MDS and Matheson, is that witness immunity applies in this case because the respondents’ only relationship with the appellants derived from the insurance action. The second is that witness immunity applies to all investigations and reports if, at the time, the investigator is a possible witness in possible future proceedings. In my respectful view, neither of these contentions is clearly established by authority. I will address each in turn. (i) The relationship between the experts and the appellants: [132] It is submitted that the key to the witness immunity in this case is the relationship between the experts and the appellants. It is critical, goes the argument, that the experts’ relationship with the appellants arose solely in the context of the appellants’ insurance action and that there was no other relationship between them. The chambers judge apparently accepted this proposition, citing a passage from Carnahan v. Coates and noting that “[t]he [respondents’] only relationship with the [appellants] derived from the first action and the witness immunity rule applies.” (Reasons ¶ 12 and 15) This proposition was also relied on in other cases cited by ICPB, in particular, Kravit v. Dilli (1998), 56 B.C.L.R. (3d) 150; B.C.J. No. 1479 (Q.L.)(S.C.), N.M.-A. et al. v. I.A.S.M.-A. et al. (1992), 93 D.L.R. (4th) 659 (B.C.C.A.) and Smith v. Kneier (2001). 288 A.R. 144; A.J. No. 458 (Q.L.)(Q.B.). [133] I have already stated my view that the facts in the record do not reasonably support the conclusion that the only relationship between the parties arose out of the insurance litigation. In the case of the insurance investigators, that relationship is rooted in the appellants’ contractual relationship with Royal and long pre-dated any litigation. In the case of Mr. Savage, the relationship derives from his statutory duties and long pre-dates litigation. [134] There is a deeper problem with the chambers judge’s reliance on this proposition. It is based, with respect, on a misreading of the Carnahan case. The judge took that case to stand for the proposition that the scope of the immunity in relation to out-of-court activities turned on the fact that the relationship between the parties derived from the litigation. In my view, the case does not stand for that. Some comments of Huddart, J. in Carnahan have been taken out of context and have led in this and other cases to confusion between the witness immunity analysis and the question of whether an expert witness has a duty of care to either the party retaining the expert or to an adverse party. In my view, to say that the appellants’ and the respondents’ relationship in this case derives solely from the appellants’ insurance action is not only wrong factually but not legally determinative of the scope of witness immunity. I will explain. [135] In Carnahan v. Coates, supra, a mother had applied to vary a child access order to delete a weekend each month with the father, her ex-husband. After she had instituted the application, she contacted Coates, a psychologist, to ask him to assess the children with regard to the father’s access. Coates agreed and formed the opinion that the children did not want the weekend visits. He said so both in a written report that was entered as an exhibit in the proceedings and during his oral testimony in those proceedings. The court deleted the weekend access. Subsequent investigation by another psychologist concluded that there had been no objective basis for taking this step and that the children’s attitudes were simply a reflection of their mother’s manipulation. Coates was found guilty of unethical conduct because he had failed to act with due care. The plaintiff father then sued Coates for negligence. [136] Coates sought summary judgment on the basis (among other grounds) that he was protected by witness immunity. Huddart, J. held that his engagement was solely for the purpose of giving testimony in court, in writing a report filed for use in Court, or both. She concluded, therefore, that the plaintiff’s cause of action, however expressed, was for negligent testimony and nothing else: at 478 D.L.R. The application of the witness immunity rule was clear because the suit was based solely on the defendant’s giving evidence in court. [137] The plaintiff father submitted that the immunity should not apply to experts because the policy justifications for immunity do not apply to professional witnesses. Huddart, J. rejected this submission. She considered the policy justifications for the immunity and concluded that they supported the claim for immunity in the case before her. The need for witnesses to testify fully and freely was implicated, particularly in relation to suits against experts by opposite parties. But she found most persuasive the other main justification for the immunity, the protection of the court process and, in particular, avoidance of relitigation of the original action in the guise of an action against the witness. [138] Huddart, J. referred extensively to Bruce v. Byrne-Stevens and Associates Engineers, Ltd. , 776 P. 2d 666 (1989), a case in which a similar argument had been made in the context of a suit by parties against their own expert. Significantly, there was no dispute among the judges in that case that the immunity would apply to a suit brought by a victim of negligence against an expert hired by the court (and by implication, an opposite party). The issue that divided the court in Bruce was whether the immunity should apply to a suit by a party against his or her own expert. [139] Of course, Carnahan involved an action against an adverse party’s expert. Huddart, J. was careful to point out this distinction between Carnahan and Bruce. She also was careful to note that the Bruce court was unanimous in the view that witness immunity applies to a suit against an opposite party’s expert. This led to her statement, cited by the chambers judge here, that “... the protection of the integrity of the judicial process requires at least that an expert witness be immune from suit by any person with whom his only relationship derives from the judicial proceeding. For the purpose of this application I need go no further”: at 478. This sentence was not intended to define the scope of witness immunity: Coates’ activity on which the suit was based fell within the core of the immunity. The comment simply made it clear that the holding was limited to suits by adverse parties. In effectively limiting her holding to experts being sued by an opposite party, she simply left open the point decided in Bruce as to whether the immunity should extend to suits against experts by the party retaining them. [140] Carnahan, therefore, supports the view that witness immunity applies to negligence actions against an expert witness by an opposite party. The appellants in this case do not argue otherwise and it is not necessary to confront the issue raised in Bruce as to whether the immunity also applies to negligence actions by a party against its own expert. However, Carnahan was a case in which the plaintiff’s action related to giving evidence in court and thus fell squarely within the scope of the immunity. It is not a helpful authority in relation to the scope of the immunity for out-of-court activities done otherwise than for the purpose of preparing evidence for court. [141] As mentioned, Carnahan has been relied on in some other cases for the proposition that the scope of the immunity somehow depends on the relationship between the parties: N.M.-A. et al. v. I.A.S.M.-A et al., supra; Read v. Munt (2004, 71 O.R. (3d) 22( Sup. Ct.); Smith v. Kneier, supra; Boychyn v. Abbey, [2001] O.J. No. 4503 (Q.L.)(Sup. Ct.); Johnson v. State Farm Mutual Automobile Insurance Co., supra; Lowe v. Guarantee Co. of North America, supra and Kravit v. Dilli, supra. In my respectful view, this is a misreading of Carnahan. [142] For these reasons, I do not accept that the fact that the parties’ relationship derives only from the proceedings is “the key” to whether witness immunity applies to out-of-court communications by investigators. Like Carnahan, no issue arises in this case about whether witness immunity applies to claims against a party’s own expert witness. (ii) Witness immunity and investigators: [143] The respondents’ second central contention is that witness immunity applies to all investigators retained to investigate a matter which may give rise to litigation in which they are possible witnesses. In support, they rely on a number of authorities, mostly concerned with police and child welfare or custody investigations. In my view, none of these cases supports the broad claim for witness immunity advanced by the respondents and accepted by the chambers judge. [144] I will turn first to the cases which address the well-settled principles of witness immunity in relation to out-of-court statements by witnesses and potential witnesses. I will then discuss a line of more recent English authority on which the respondents rely and finally address the Canadian authorities which the respondents say support them. In my view, the authorities show that witness immunity applies to out-of-court statements made for the purpose of gathering evidence for litigation which was contemplated at that time whether the person ultimately is called as a witness or not. The authorities do not, however, support the immunity claimed by the respondents for all investigators investigating anything that may possibly give rise to litigation in which they may be possible witnesses. I conclude that the authorities do not support an immunity in relation to the appellants’ claims as pleaded in the statement of claim, but do protect against their claims as expanded in the particulars. (a) Out-of-court statements by witnesses: [145] The core of witness immunity, like that of the broader immunity for judicial and quasi-judicial proceedings, is conduct in court or during a formal step in the litigation. However, the law has long recognized that it is necessary to protect some out-of-court statements of witnesses or intended witnesses in order to make the protection for witnesses effective. It has been held that the immunity applies to proofs of evidence given to a party’s lawyer or other professionals engaged in preparing evidence for trial, to written reports required by statutes or rules of court as a condition for giving oral testimony and to other reports authored for the purposes of pending or contemplated litigation. [146] The leading case is Watson v. M’Ewan. The issue was whether the previous communication by the witness to the opposite party and his legal advisors of statements ultimately given in evidence at trial fell within the scope of the immunity. In holding that they did, the Earl of Halsbury, L.C. said at pp. 487 (A.C.): ... the privilege which surrounds the evidence actually given in a Court of justice necessarily involves the same privilege in the case of making a statement to a solicitor and other persons who are engaged in the conduct of proceedings in Courts of justice when what is intended to be stated in a Court of justice is narrated to them ... If it were otherwise, ...the object for which the privilege exists is gone. ... It is very obvious that the public policy which renders the protection of witnesses necessary for the administration of justice must as a necessary consequence involve that which is a step towards and is part of the administration of justice — namely, the preliminary examination of witnesses to find out what they can prove. ... (emphasis added) [147] In Watson, the out-of-court statements were made for the purpose of communicating to the lawyers and the party what the witness’ evidence would be in relation to a proceeding actually in progress. Thus, the link between the out-of-court statement and the testimony was clear and direct as was the link between the purpose for which the out-of-court statement was made and a pending judicial proceeding. [148] Watson was applied by the Supreme Court of Canada in Halls v. Mitchell. The appellant, a discharged soldier, had applied for workers’ compensation benefits. The benefits were denied on the basis of medical evidence provided by Dr. Mitchell both in writing and in oral testimony. The appellant sued Mitchell for defamation with respect to four publications: one to the Workers Compensation Board in writing and three others to medical practitioners in writing or orally. Only the communications to two other medical practitioners, one of whom was associated with the military and one of whom was an eye specialist who examined the appellant after his initial examination by Dr. Mitchell were in issue on appeal to the Supreme Court of Canada. Among other things, Mitchell submitted that his communication to the eye specialist, Dr. McCallum, was protected by witness immunity. In rejecting this contention, Duff, J., writing for the majority, said at 145-146: It was rather suggested that the letter to Dr. McCallum should be protected as within the principle of Watson v. McEwan [[1905] A.C. 480.]. The basis of the judgment in Watson v. McEwan [[1905] A.C. 480.] is that statements made by a witness as such, in court, are absolutely privileged, and that this privilege would become illusory, were it not applicable for the protection of a statement by an intending witness, as to the nature of the evidence the witness can give, made to professional persons preparing the evidence to be presented in court. As the protection by privilege of the testimony of witnesses in court is regarded by the law as essential to the administration of justice, and as the extension of that protection to such preliminary statements is regarded as essential to the effectiveness of the substantive privilege, such preliminary statements are held to fall within the rule; but, as Lord Halsbury points out, this strict necessity is the basis of the privilege. In Watson v. McEwan [[1905] A.C. 480.] there was no question, as Lord Halsbury observes, of communications to persons other than those engaged professionally in preparing the evidence to be presented in court, and obviously the principle does not extend to such collateral statements. It protects the respondent, whatever his motives may have been, in respect of statements made before the Workmen's Compensation Board and in respect of statements made to the Claims Agent, voluntary though they were, as to the evidence which he was prepared to give; but this privilege has no relation to the statements made to Dr. Campbell, to Dr. Hewitt, or to Dr. McCallum. [149] Thus, in Halls, the privilege was limited to the witness communicating the evidence he was prepared to give to a claims agent. It was held – indeed thought to be “obvious” – that immunity did not extend to persons other than those engaged professionally in preparing the evidence to be presented in court. Thus, the statements made to the eye specialist from whom Mitchell had requested an opinion were not protected. It would appear that while the out-of-court statement to the eye specialist related to the same subject-matter as the evidence given to the Workers Compensation Board, there was little link between the request to the eye specialist and the witness’ role as a witness before the Board. (See also R.G. v. Christison (1996), 150 Sask. R. 1 (Q.B.)) [150] Two decisions of the Ontario Court of Appeal illustrate the application of these principles. [151] In Foran v. Richman (1976), 10 O.R. (2d) 634 (C.A.), the court held that the immunity applied to statements in letters written by the witness to the plaintiff’s lawyer. These letters had been requested by the lawyer as part of preparation for the trial of the plaintiff’s personal injury action in order to determine what the witness’s evidence would be, and the witness had testified at the trial. [152] In Fabian v. Marguilies (1985), 53 O.R. (2d) 380 (C.A.), the appellant had been the plaintiff in a personal injury action. A psychiatrist had conducted a medical examination of him on behalf of the defendant, had prepared a medical legal report for the defendant’s solicitors and had testified at the trial. After the personal injury action was dismissed, the appellant sued the psychiatrist for defamation. The question arose as to whether the psychiatrist’s medical legal report as well as his testimony were immune from liability. [153] Labrosse, J. held that they were and his decision was upheld in a brief judgment by the Court of Appeal. There was no serious issue concerning immunity in relation to the testimony in court. With respect to the report, Labrosse, J. held that the medical legal report was required under the relevant rules of evidence before the doctor could testify. It was, therefore, essential to extend the immunity to it in order to give effect to the immunity of witnesses: 10 ... In order to be permitted to testify at the trial, the doctor had to prepare a report to be used in accordance with the provisions of s. 52 of the Evidence Act, R.S.O. 1980, c. 145. In my opinion, the examination and the ensuing report were prepared in the ordinary course of judicial proceedings. ... The absolute privilege extending to the oral evidence of the doctor would be rendered illusory if he could be sued for the same statements made in a report he is statutorily required to prepare in order to give his oral testimony. [154] The court in Fabian was persuaded of three things: first, that the preparation of the written report was an essential aspect of giving the oral evidence -- the filing of a written report was a statutory pre-condition to the testimony; second, that the extension of the immunity to out-of-court statements was essential to prevent the immunity for in-court statements becoming illusory; and third, that the statements both in and out-of-court were “... the same statements ...”. (b) Out-of-court statements by persons who do not become witnesses: [155] In Watson, Halls and Fabian, the immunity was applied to a witness’s out-of-court statements. The application of the immunity, however, does not depend entirely on whether the person actually testifies. The immunity also extends to statements made by a potential witness to persons engaged professionally in assembling the evidence for pending or contemplated proceedings. [156] An example is Web Offset Publications Limited v. Vickery (1998), 40 O.R. (3d) 526 (Gen. Div.), aff’d 43 O.R. (3d) 802 (C.A.), application for leave to appeal dismissed [1999] S.C.C.A. No 460. The defendant, Mr. Vickery, was a new York lawyer. He had authored a complaint against the plaintiffs under the Racketeer Influenced Corrupt Organizations Act (“RICO”) in New York and had delivered a copy of the complaint and repeated the words in it to lawyers in Ontario. He was sued by the plaintiffs for, among other things, defamation. He successfully applied to strike the claim on the basis of absolute immunity and that decision was upheld on appeal. The judge reasoned that Vickery had provided the RICO complaint to the Ontario lawyers as evidence and that he was a potential witness in the Ontario litigation, which appears to have been under way at the time. In reaching his conclusion, the judge approved the following statement from Gatley on Libel and Slander, 9th ed. (London: Sweet & Maxwell, 1998) at 290: Extent of Privilege. The privilege which protects a witness from an action for defamation in respect of his evidence in a judicial proceeding applies not only to evidence given viva voce, but also to statements contained in an affidavit, or in a document handed in by a witness at the close of his examination. The privilege also extends to statements made by a witness to a solicitor taking his proof, or when interviewing him with the object of possibly calling him at the trial, to documents produced by the witness at such interview which relate to the matter in issue . . . [157] The Court of Appeal dismissed the appeal from the judge’s decision. The Court relied on the same section of Gatley and held that the information supplied was absolutely privileged: At the time it was provided, the Ontario lawyers were seeking information from a potential witness for use in pending litigation. [158] A similar result obtained in Hamouth v. Stuart Video Technologies Inc., [2005] B.C.J. No. 639 (Q.L.)(C.A.). The defendant law firm was representing Smart Video in an investigation by the United States Securities and Exchange Commission (“CES”). The law firm wrote twice to Smart’s brokers to inform them of the inquiry and to advise them of what the firm considered to be the plaintiff’s questionable trading practices. The plaintiff’s defamation action was struck on appeal on the basis that the communication to the brokers was absolutely privileged. The letters, reasoned the Court, had been written by lawyers acting “ ... in the course of their duties to their client in the course of a quasi-judicial proceeding” : ¶ 39. [159] Many of the authorities concerning the immunity in relation to judicial proceedings were recently reviewed by Cullity, J. in Moseley-Williams v. Hansler Industries Ltd., [2004] O.J. No 5253 (Q.L.)(Sup.Ct.), appeal quashed [2005] O.J. No. 997 (Q.L.). Moseley-Williams had been dismissed from his employment with Hansler and had been hired by Ri-Go Lift Truck Ltd. On instructions from Hansler, its solicitor wrote to Ri-Go alleging that Moseley-Williams (among others) had engaged in improper sales activities including breaching a non-solicitation agreement with respect to Hansler’s protected customers. Moseley-Williams sued Hansler for libel. Hansler’s summary judgment application based on absolute immunity was dismissed. [160] Cullity, J. found that at the time the letter was written, no proceedings had been commenced, the lawyer had not been instructed to commence proceedings and there was no evidence that any proceedings had ever been commenced. He found that the lawyer’s letter “ was written in the course, and for the purpose, of asserting Hansler’s rights, but not for the purpose of, or preparatory to the commencement of proceedings by Hansler”. Therefore, the absolute privilege did not apply: at ¶ 60. [161] In cases of non-witnesses, the immunity will generally apply to statements made for the purpose of providing evidence in contemplated proceedings. So, for example, the immunity extends to out-of-court statements of persons conducting medical, psychiatric or psychological assessments ordered by the court or otherwise required for pending or contemplated proceedings. Even though the person may not testify, the whole purpose of the assessment in these cases is the preparation of evidence for a proceeding and it is necessary to protect it as well as the testimony given in court. Such witnesses cannot perform their role as witnesses without undertaking the assessment and so it is necessary, in order to protect them in their role as witnesses, to protect the steps essential to permit them to give their evidence: see, for example, Read v. Munt, supra ; Lowe v. Guarantee Co. of North America, supra and; Varghese v. Landau, [2004] O.T.C. 97 (Sup.Ct.). (c) The English Authorities: [162] I turn next to an important line of English authorities concerned with witness immunity. The respondents rely on the English trial level decision in Evans v. London Hospital Medical College, [1981] 1 All E.R. 715 (Q.B.). While Evans is a helpful and important authority, it does not support the existence of the immunity claimed in this case. Moreover, there is no authority binding on this Court which adopts Evans and, as we shall see, the apparent breadth of the holding in Evans has been circumscribed by more recent decisions of the House of Lords and the English Court of Appeal. [163] In Evans, a pathologist and two toxicologists (the investigators) carried out a post-mortem on the plaintiff’s infant son. They prepared a report for the police and the Director of Public Prosecutions. Shortly afterwards, they made statements under ss. 2 and 9 of the Criminal Justice Act 1967. These were formal witness statements which, on certain conditions, were admissible in court. Allegedly on the basis of the post mortem results, Ms. Evans was charged with murder. At her trial, the Crown offered no evidence and she was acquitted. She then sued the investigators for negligence and malicious prosecution. The allegations of negligence, as in the present case, related both to the investigators’ initial conclusions and to their failure to retract or modify their findings once confronted with contrary expert opinion. The defendants applied to strike out the claim primarily on the basis that it was barred by witness immunity. The issue was how far the absolute immunity extends and, in particular, whether it covers the acts or omissions of a witness or potential witness while they are collecting or considering material with a view to its possible use in criminal proceedings. [164] The judge, Drake, J. held that it did. He rejected any distinction between the preparation of evidence (in the sense of preparing statements for use in court) and the report prepared by the defendants for use by the Director of Public Prosecutions. The question, therefore, became one of fact: Did the negligent acts or omissions arise during the course of preparation of evidence? Drake, J. held that they did. He characterized the case as relating to allegedly negligent collection and preparation of evidence which was to be taken down as a proof of evidence later (at 719). He held -- and his statement of the test has been influential -- that witness immunity applies to statements before proceedings have been commenced “... provided that the statement is made for the purpose of a possible action or prosecution and at a time when a possible action or prosecution is being considered.” (at 721). He added that “[t]he protection exists only where the statement or conduct is such that it can fairly be said to be part of the process of investigating a crime or a possible crime with a view to a prosecution or possible prosecution in respect of the matter being investigated” (at 721). The test set out in Evans, therefore, has two key elements. The purpose of the statement must be in relation to preparing evidence and the statement must be made at a time when a proceeding is being considered. [165] There are clear differences between Evans and the present case. First, Evans was concerned with the investigation by public officials of a suspected crime. The present case is concerned (apart from the respondent Savage) with actions taken on behalf of an insurer in the course of adjusting a loss claimed to fall within the coverage of an insurance policy. In the case of the respondent Savage, the case relates to a discretionary investigation of a fire for a variety of statutory purposes. Second, in Evans, the investigators prepared formal written witness statements which could be admissible in criminal court. From the beginning, they were preparing evidence for use in court. Significantly, the judge in Evans found that the alleged negligence arose during the course of the preparation of evidence for court: at 720. There is, of course, no similar finding in the present case, nor could there be. So far as the agreed facts disclose, no formal witness statements were taken from anyone and litigation was viewed only as a possibility. Third, in Evans, the investigations were carried out for the purpose of a prosecution or possible prosecution. In the present case, all that the agreed facts stipulate is that the investigations were undertaken to determine the cause and origin of the fire in the context of the insurer’s attempt to adjust the loss. Fourth, in Evans, at the time of the investigation, a prosecution was actually being considered. In the present case, there is no evidence that at the time of the investigations, any proceedings were actually being considered or that the investigations were undertaken with a view to, or for the purpose of, gathering evidence for any such proceedings. The most that may be said is that the possibility of litigation was inherent in the situation. [166] Evans involved a public law duty to investigate crime. This aspect of the case was emphasized by the House of Lords in X(Minors) v. Bedfordshire County Council. In that case, one of the issues was the liability in negligence of a psychiatrist who, at the request of a local authority, had examined a child to determine whether the child had been sexually abused. The psychiatrist concluded that the child had been sexually abused and, the same day, court proceedings were launched leading to a wardship application. The psychiatrist prepared a formal written report which was filed in court and the substance of the psychiatrist’s conclusions had been informally placed before the courts early in the proceedings. [167] The House of Lords held that witness immunity applied to the psychiatrist. At the time of the interview requested by the local authority, the psychiatrist “... must have known that, if such abuse were discovered, proceedings by the local authority for the protection of the child would ensue and that her findings would be the evidence on which those proceedings would be based ...” per Lord Browne-Wilkinson at 755. Thus, the investigation undertaken by the psychiatrist had “.. such an immediate link with possible proceedings in pursuance of a statutory duty ...” that it could not be made the basis of subsequent claims: at 755. [168] The holding in X(Minors) was limited to the particular context of a local authority, acting under a statutory duty, investigating whether or not there was evidence on which to bring proceedings for the protection of a child from abuse: 755. Evans was referred to as a case involving the investigation of crime on behalf of the Director of Public Prosecutions. The House of Lords did not opine on whether the same principle could extend to ordinary civil proceedings. The case, therefore, introduces the idea that the principle in Evans is limited to investigations in relation to contemplated criminal prosecutions or, as in X (Minors), to investigations by local authorities pursuant to a statutory duty in relation to suspected child abuse. [169] Evans was also considered in detail by the House of Lords in Taylor v. Director of the Serious Fraud Office, supra. In that case, Mr. Taylor, a solicitor, sued the Director of Serious Fraud Office and others for publishing allegedly defamatory statements to the Attorney General for the Isle of Man and to officials of the Law Society. One question was whether witness immunity extends to statements made to or by investigators for the purposes of a criminal investigation. Lord Hoffman (with whom Lord Goff, Lord Hope and Lord Hutton agreed on this point) upheld (albeit in obiter) the finding of immunity in relation to all of the statements. [170] Lord Hoffman found the test of necessity was met: ... I find it impossible to identify any rational principle which would confine the immunity for out-of-court statements to persons who are subsequently called as witnesses. ... It would be an incoherent rule which gave a potential witness immunity in respect of the statements which he made to an investigator but offered no similar immunity to the investigator if he passed that information to a colleague engaged in the investigation or put it to another potential witness. ... I therefore agree with the test proposed by Drake J. in Evans v. London Hospital Medical College (University of London) [1981] 1 W.L.R. 184, 192: ". . . the protection exists only where the statement or conduct is such that it can fairly be said to be part of the process of investigating a crime or a possible crime with a view to a prosecution or a possible prosecution in respect of the matter being investigated." [171] There are two particularly significant aspects of the decision in Taylor. First, the case concerned a criminal investigation and the rationale of the decision depends on the particular context of the investigation of crime or suspected crime. There was no discussion of whether the same test would apply outside the context of the investigation of crime or child abuse. Second, the House approved the formulation of the test for the application of the immunity set out by Drake, J. in Evans. That test, as noted, limits the immunity to situations in which “... the statement or conduct [of witnesses or potential witnesses] is such that it can fairly be said to be part of the process of investigating a crime or a possible crime with a view to a prosecution or a possible prosecution in respect of the matter being investigated.” In other words, as found in Evans, the statements must be made for the purpose of gathering evidence for a proceeding actually being contemplated at the time. [172] Evans was also considered and, in my view, its scope clarified, by the House of Lords in Darker v. Chief Constable of The West Midlands. The question was whether the immunity extended to police officers who allegedly fabricated evidence, conspired to have charges laid that they knew to be false, knowingly instructed undercover officers to breach police instructions of operation and so forth. The claim was framed in conspiracy to injure and misfeasance in a public office. The officers had not been witnesses: The claims were based on allegations about things they did while engaged in the investigation of crime and preparation of the case for trial. The House of Lords held that the officers were not protected by witness immunity. Five speeches were delivered and Evans was considered in detail. [173] For Lord Hope, witness immunity extends to “... claims made against witnesses for things said or done by them in the ordinary course of such proceedings on the ground of negligence.”(at 446). However, he drew a distinction between claims founded on the giving of evidence and claims founded on some other act. For him, there is a difference between participating in the judicial process as a witness and acting as an investigator. As he put it (at p. 448), “The purpose of the immunity is to protect witnesses against claims made against them for something said or done in the course of giving or preparing to give evidence. It is not to be used to shield the police from action for things done while they are acting as law enforcers or investigators.” (Emphasis added) [174] Lord Hope, while accepting the result in Evans, did not agree with the test laid down there for determining the extent of the immunity. He noted that it is important not to confuse the functions of a witness with those of an investigator or the immunity rule with whether or not a duty of care exists. He concluded at pp. 449 - 450 (A.C.): ... The purpose of the immunity rule is to protect the witness in respect of statements made or things done when giving or preparing to give evidence. The acts of the witness in collecting material on which he may later be called to give evidence are not protected by the immunity. The immunity extends only to the content of the evidence which the witness gives or is preparing to give based on that material. I think that Sir Richard Scott V-C described the position correctly when he said in Bennett v. Comr of Police of the Metropolis (1997) 10 Admin LR 245, 252D-E that the immunity extends to statements made or agreed to be made out-of-court “if these were clearly and directly made in relation to the proceedings in court, for example, witnesses’ proofs of evidence . . .” (Emphasis added) [175] Lord MacKay thought that the immunity should not apply to all of the investigatory and preparatory process. While he did not expressly doubt the test set out in Evans, he emphasized that the immunity relates only to statements made for the purpose of court proceedings: at p. 452. [176] Lord Hutton was of the view that both Evans and Taylor were concerned with liability for statements made, not for actions taken and that the essence of the immunity is concerned with the giving of evidence. He said at p. 469: The underlying rationale of the immunity given to a witness is to ensure that persons who may be witnesses in other cases in the future will not be deterred from giving evidence by fear of being sued for what they say in court. This immunity has been extended, as I have described, to proofs of evidence and to prevent witnesses being sued for conspiracy to give false evidence. But the immunity in essence relates to the giving of evidence. There is, in my opinion, a distinction in principle between what a witness says in court (or what in a proof of evidence a prospective witness states he will say in court) and the fabrication of evidence, such as the forging of a suspect’s signature to a confession or a police officer planting a brick or drugs on a suspect. ... [177] Lord Hutton emphasized that the statements at issue in Evans were made for the purpose of a report that would constitute evidence for a possible prosecution which was actually being considered at the time: On the facts of that case I consider that the decision of Drake J that the defendants were entitled to absolute immunity was correct. Although the plaintiff alleged that it was done negligently, the organs were removed from the body and examined for the genuine purpose of making a report which would constitute a statement of evidence for a possible prosecution and therefore, in my opinion, came within the ambit of the immunity. (at 471 - 2) (Emphasis added) [178] Lord Cooke said that Evans was concerned with the negligent preparation of evidence. He adopted a functional approach to the application of witness immunity. This involves characterizing the activity of the defendant as either part of the investigative phase, which is not subject to immunity, or the judicial phase, which is. Referring to the investigative phase, he said: “Conduct which is primarily and naturally to be seen as belonging to the investigatory function, even though it may have some ultimate link with the giving of evidence, should not be within the general protection”: at 454. [179] Lord Clyde thought that it is not enough that there be an investigation: the investigation must be with a view to an action or a prosecution which is already under consideration: at 459. He adopted a distinction between matters of advocacy, which fall within the immunity, and matters of detection, which do not. The immunity does not attach to things said or done which would not form part of the evidence to be given in the judicial process. This is because the reason for extending the immunity to things said or done out-of-court is to prevent circumvention of the immunity by means of a collateral attack on the witness’s evidence. For him, the critical factor is whether or not the material in question was or was not provided with a view to court proceedings: at 458. Once proceedings are underway and even after a witness has testified, it does not follow that everything said or done in the investigation or preparation for a judicial process is covered by the immunity. Particular cases must be considered in relation to the underlying reasons for the immunity: at 460. [180] Darker is important for the present case for two reasons. First, it emphasizes the importance of the purpose for which out-of-court statements were made. In order for witness immunity to apply to an out-of-court statement, it must be one made for the purpose of preparing evidence for a proceeding. The various speeches express this in different ways: Lord Hope was concerned with whether the statement was made “in relation to the proceedings in court”; Lord MacKay and Lord Hutton insisted that the purpose be the preparation of evidence for court; Lord Cooke referred to those statements that belong to the “judicial” as opposed to the “investigative” stage of the proceedings and Lord Clyde spoke in terms of the statement being made “with a view to” judicial proceedings. Second, Darker draws a distinction between investigation on one hand and preparation of evidence on the other. Whether expressed as it was by Lord Clyde as a distinction between matters of advocacy and matters of detection or, as it was by Lord Cooke, as a distinction between the investigative as opposed to the judicial phase or, as it was by Lord Hope, as a distinction between the role of a witness and the role of an investigator, the fundamental point is the same. Lord MacKay summed it up this way: “[t]he essential character of the immunity ... limits the application of the immunity to conduct which can be called in question only by a founding on a statement in court or a statement which is part of the preparation of evidence for court proceedings”: at 452. [181] I note once again that in the present case, the agreed facts do not stipulate, nor did the chambers judge draw the inference, that the respondents conducted their investigations for the purpose of collecting evidence for a proceeding or that any such proceeding was actually being considered at that time. [182] The importance of the purpose of the investigation and the role of the investigator were also emphasized by the English Court of Appeal in Stanton v. Callaghan, [1998] E.W.J. No. 823 (Q.L.)(C.A.) The plaintiffs sued an engineering firm which they had retained to report to them on how to repair subsidence damage which they had suffered. On the basis of the report, they advanced a claim against their insurers which eventually became the subject of litigation. The principal of the engineering firm was to be an expert witness on behalf of the plaintiffs and before trial, acting on the instructions of the plaintiffs’ solicitors, he met with the insurer’s expert with a view to agreeing on as much as possible and making a list of disputed areas. They agreed on a solution that was considerably less expensive than the one he had reported as necessary to the plaintiffs and, as a result, the plaintiffs felt they had little choice but to settle their action against the insurer on that basis. They then sued the firm and its principal alleging, among other things, that the firm had been negligent in agreeing to an inappropriate solution or, if the solution was correct, had breached a duty to the plaintiffs by failing to advise the plaintiffs of it earlier and by agreeing with the defendant’s expert without notice to, or the consent of, the plaintiffs. At issue was whether there was a distinction between alleged negligence of an expert acting as an expert witness and alleged negligence on the part of the expert acting as an adviser to the party instructing him. The Court of Appeal unanimously allowed the appeal and dismissed the action, holding that witness immunity barred the claim. [183] Chadwick, L.J. carefully examined the bases of the plaintiffs’ allegations and concluded that the plaintiff’s primary claim was this: The method of remedying the subsidence to which the expert had agreed at the meeting was wrong and had been arrived at as a result of negligence. For Chadwick, L.J., the key to the question of whether the immunity applied was the connection between the cause of action asserted and the work of the expert as a witness as distinguished from his work as an adviser: “The immunity would only extend to what could fairly be said to be preliminary to his giving evidence in court judged perhaps by the principal purpose for which the work was done. So the production or approval of a report for the purposes of disclosure to the other side would be immune but work done for the principal purpose of advising the client would not”: at ¶ 66 (Q.L.). He held that the expert’s report, which was the basis of the action, had been made after and as a result of a meeting between the experts on each side of an action that was about to go to trial. Accordingly, witness immunity applied. [184] Otton, L.J. held that witness immunity protects work “principally and proximately” leading to the evidence to be given in court. Nourse, L.J. found that the principal, if not the sole, purpose of the report giving rise to the action was to permit the expert to give evidence at trial. He questioned whether a “substantial purpose” test might be preferable to a “principal purpose” test. Thus, two of the three judges adopted a “principal” or “substantial” purpose test for the application of the immunity. This approach does not seem to me to be different in principle from the “dominant purpose” test which is advocated by the appellants in the present case. But however one describes the required purpose, both Darker and Stanton show that the purpose of the out-of-court statement or conduct is an important aspect of the witness immunity analysis. [185] Stanton was considered by the subsequent House of Lords decision in Hall v. Simons, [2002] 1 A.C. 615 (H.L.(Eng.)). Lord Hoffman ( Lord Browne-Wilkinson and Lord Millett concurring) said that in Stanton, “[t]he alleged cause of action was a statement of the evidence which the witness proposed to give to the court”: at 698. Thus the immunity applied because the cause of action derived entirely from the protected activity. In other words, the essence of the plaintiffs’ claim in Stanton was the conduct of their expert as a witness in a judicial proceeding. [186] Does this line of English authority support the respondents’ claim for witness immunity? In my view, it does not. Drake, J’s formulation of the test for witness immunity in Evans has been approved by the House of Lords in Taylor. However, its application outside of the investigation of crime or of suspected child abuse by those under a duty to do so has been left open. And, in my view, the apparent breadth of the test set out in Evans has been circumscribed by the House of Lords in Darker. Darker emphasized that there is a distinction between the activities of a witness and of an investigator and that only things done for the purpose of preparing evidence for actual or contemplated proceedings fall within the ambit of the immunity. Remember that in Evans, unlike the present case, the judge found that the allegedly negligent acts arose during the preparation of evidence. Moreover, in the present case, there was no public law or statutory duty to investigate as in Evans, (X)Minors, Taylor and Darker. The agreed facts establish nothing more than that, at the time of their investigation, litigation was possible and, if it occurred, the individual respondents were possible witnesses. The facts do not permit a finding that the respondents’ investigations were for the purpose of preparing evidence for litigation that was actually contemplated at that time. (d) The Canadian authorities: [187] I now turn to the Canadian authorities on which the respondents place particular reliance. [188] N.M.-A., supra raised a witness immunity issue similar to that in X(Minors). In N.M.-A., a psychologist was retained by the Superintendent of Family and Child Services to assess whether a child had been sexually abused by a father. The psychologist knew that court action was being considered by the Superintendent. She sent her report to legal counsel for the Superintendent along with a note stating that she knew she would have to testify. The same day that she delivered her report, the Superintendent apprehended the child. The report was filed in court and the psychologist testified. The father sued the psychologist. The issue was whether the action was barred by the witness immunity rule. The Court of Appeal held that it was, accepting the proposition that witness immunity applies to reports made during the preparation of evidence for litigation contemplated at the time. It noted that the witness had been retained for the purpose of preparing evidence for contemplated litigation by a party with a statutory duty to investigate suspected child abuse. The witness had a statutory duty to report the abuse if she had reasonable grounds to believe it had occurred. The witness delivered the report to legal counsel and testified in relation to that report. In short, it was clear that she functioned throughout as a likely witness in proceedings actually being contemplated at the time she carried out her investigation and she actually became a witness in those proceedings. [189] There is no parallel between N.M.-A. and the present case. Here, while proceedings were known to be possible, there is no evidence they were actually contemplated. The facts do not establish that the investigators thought they were preparing evidence for proceedings and their reports were delivered to the insurer, not to counsel. [190] Kansa General International Insurance Co. v. Morden & Helwig Ltd. (2001), 57 O.R. (3d) 58 (Sup. Ct.) provides no support for the sweeping immunity claimed by the respondents. In Kansa, an insurer sued its own adjuster for breach of contract and negligence. The claim related directly to evidence which the adjuster gave in court and statements he had made in preparation for his testimony. Kansa, therefore, has no factual similarity to the present case. Sutherland, J. found that witness immunity clearly covered the evidence given at the personal injury trial. The question was how far back in time the immunity extended. Relying on Horn Abbot v. Reeves, he concluded that the immunity extended to statements made by a prospective witness to solicitors or counsel in preparation for the witness’ testimony: at ¶ 58. Counsel knew that the adjuster would be a witness if the matter went to trial. Accordingly, the immunity extended to the adjuster’s responses to counsel defending the personal injury action during preparation for discovery and trial. In summary, the immunity upheld in Kansa clearly fell within the scope of the rule as set out over 100 years ago in Watson v. M’Ewan. The case provides no support at all for the broad immunity claimed in this case. [191] The respondents also rely on Smith v. Kneier, but it does not support their position. It was a case in which psychologists were preparing a court ordered assessment for use as evidence in court. Nothing of the sort occurred with respect to the respondents in the present case. We were also referred to Boychyn v. Abbey. Like Kneier, Boychyn concerned an action against a psychologist who carried out a court ordered assessment in the context of ongoing custody and access proceedings. Relying on Smith v. Kneier, the court held that witness immunity extends to court-ordered assessments prepared for legal proceedings, even though the assessments are never, in fact, used in legal proceedings. This is a markedly different situation than we face in the present case. [192] Nor do I think that Kravit v. Dilli supports the respondents. There, the Kravits had been involved in a series of proceedings in criminal court arising out of disputes between them and Mrs. Kravit’s family. Crown counsel retained Dr. Dilli to prepare a report to assist Crown counsel in assessing the risk posed by the Kravits to the physical safety of those working in the courts and which would serve as opinion evidence to be used by the Crown in criminal proceedings. The Kravits sued the psychiatrist, but the action was dismissed on the basis of witness immunity. Baker, J. held that the report was prepared to provide evidence for contemplated proceedings and therefore it fell “squarely within” the scope of witness immunity: at ¶ 31. In Kravit, as in Evans, both a public law duty and the specific purpose of preparing evidence for pending or contemplated proceedings were present. Neither is present in this case. [193] There are a number of other Canadian cases dealing with witness immunity in the context of investigations and I will review them briefly. [194] In Ayangma v. NAV Canada (2001), 203 D.L.R. (4th) 717; P.E.I.J. No. 5 (Q.L.) (S.C.A.D.), the question was whether a statement made to an investigator with the Canadian Human Rights Commission could found an action in defamation. The conclusion was that the statement was absolutely privileged. The Court found that witness immunity applies to witnesses or potential witnesses who make statements to an investigator gathering evidence in relation to a complaint made to the Commission and with whom the witness or potential witness is obligated by statute to cooperate. Citing both Taylor v. Director of the Serious Fraud Office and Web Offset v. Vickery, the Court reasoned that the statements would be absolutely privileged if given in evidence at a hearing before the Human Rights Tribunal. Therefore, it would defeat the purposes of both the immunity and the Canadian Human Rights Act, R.S.C. 1985, c. H-6 not to extend the immunity to “... witnesses or potential witnesses in the course of the investigation of a human rights complaint prior to the adjudication of that complaint by the Tribunal”: at ¶ 53. [195] Starkman v. Canada (Attorney General), [2000] O.J. No. 3764 (Q.L.)(Sup. Ct.) shows that the purpose for which an investigation is conducted is essentially a matter of fact. In that case, the Minister of National Revenue (“MNR”) retained the defendant appraisers to assist in his investigation of the plaintiff. Charges were laid against the plaintiff, but subsequently stayed. The plaintiff sued the appraisers for fraud, deceit, fraudulent misrepresentation, negligence, misrepresentation, conspiracy to injure, conspiracy to commit the tort of malicious prosecution and injurious falsehood. The appraisers moved to strike the claim on the basis that they were immune from suit since their report and evidence were prepared and produced in the context of a criminal investigation. For the purpose of the motion, it was assumed that the appraisers had been retained at the point which MNR was simply investigating the possibility of laying charges and that, some 17 months after the report had been completed, it had been relied on to obtain a search warrant and lay charges. Mesbur, J. concluded that it was not “plain and obvious” that the claim was barred by witness immunity. She distinguished the case from Evans on the basis that there were no proceedings contemplated at the time of the investigation and that in Ontario, it is not settled law that the immunity extends to statements made to investigators in preparation of a potential prosecution: at para. 7. [196] Whether witness immunity applies to complaints to the police has been more controversial. [197] Kazas v. Peterson, [1992] O.J. No. 1666 (Q.L.) (Gen. Div.), was an action for defamation based on the defendant’s statements to the police that the plaintiff had assaulted and threatened her. Adams, J. found that absolute immunity applied to both testimony during proceedings and to complaints or communication made as a necessary incident of the commencement of such proceedings. However, in my respectful view, subsequent case law suggests that this formulation is unduly broad. [198] In Rajkhowa v. Watson (1998), 167 N.S.R. (2d) 108 (S.C.), the plaintiff, who was a psychiatrist, sued Maritime Medical Care Inc. and its medical consultant, Dr. Watson, for various things including defamation. In the context of administering MSI for the Province, Dr. Watson met with the plaintiff as part of an audit of his billing practices. Dr. Watson provided information to the RCMP which investigated, but no charges were laid. Hood, J. found that Dr. Watson’s communication to the police was not absolutely privileged. She reviewed many authorities, including Lincoln v. Daniels, [1961] 3 All E.R. 740 (C.A.), Sussman v. Eales (1985), 1 C.P.C. (2d), 14 (Ont. H.C.), aff’d 25 C.P.C.(2d) 7 (Ont. C.A.) and Kazas v. Peterson, and held that for the privilege to apply, three conditions must be met: there must be a direct connection between the making of the statement and the proceedings; the statement must be necessary in order for the witness to give evidence; and, it must be in the public interest that the information be given: at ¶ 39. She found that the first condition was not met. At the time Dr. Watson provided information to the police, there was no proceeding. As she put it at ¶ 46, “[a] judicial or quasi-judicial proceeding does not commence each time the police commence an investigation.” [199] In Teskey v. Toronto Transit Commission, [2003] O.J. No. 5314 (Q.L.) (Sup. Ct.), the TTC hired a firm of private investigators to investigate the conduct of some employees, including Teskey. As a result of the investigation, Teskey was fired for allegedly having sold drugs to one of the investigators. That information was provided to the police and Teskey was charged with a drug offence. He sued for defamation. The defendants sought to dismiss the action on the basis that the statements made by the investigators during the investigation were absolutely privileged. [200] Wilson, J. dismissed the motion. After considering Taylor and Darker, she held that while absolute privilege applies to the preparation of witnesses prior to a trial and to their participation during the trial, it does not apply to statements and documents prepared before an investigation has commenced: at ¶ 75 - 77. She cited with approval the statement of Lord Clyde in Darker that the immunity applies where the statement is made “for the purpose of a possible action or prosecution and at a time when a possible action or prosecution is being considered.” She adopted a distinction between “matters of advocacy” and “matters of detection”, with the former attracting absolute immunity and the latter not: at ¶ 60. Matters of advocacy are those things which are “intimately associated” with the judicial phase of the criminal process: at ¶ 60. (See also Hanisch v. Canada (2003), 16 B.C.L.R. (4th) 310; B.C.J. No. 1518 (Q.L.)(S.C.), rev’d in part (2004), 35 B.C.L.R. (4th) 33; B.C.J. 2159 (Q.L.)(C.A.)) [201] The respondents say that the jurisprudence is clear that witness immunity attaches to all stages of the litigation, including possible future litigation. They cite, among other cases, Watson, Martini , Fabian ,Carnahan, N.M.-A. and Evans. In my view, for reasons already given, none of those cases supports the existence of such a broad immunity. The respondents refer as well to Crossan v. Mortgage and Appraisals Ltd (1998), 164 Nfld. & P.E.I.R. 319; N.J. No. 150 (Q.L.) (S.C.T.D.). However, that case concerned a negligence action based on a court ordered report for use in court proceedings and thus does not support the immunity claimed. (e) Conclusion: [202] I conclude, therefore, that in relation to the claims as pleaded in the statement of claim, the immunity asserted by the respondents is not established by authority. Their claim for immunity, therefore, must be assessed against the test of necessity. [203] However, I reach a different conclusion about the appellants’ claims as set out in the particulars in relation to MDS, Matheson and Savage. Those claims arise out of those respondents’ alleged persistence in asserting their opinions and failure to retract or properly qualify them once confronted with contrary expert opinion. It is a reasonable inference that these allegations relate to the stage at which the appellants provided their expert opinions to the insurer. It is also a reasonable inference that once expert reports are exchanged, the authors are operating in the role of witnesses for the purpose of preparing evidence for litigation. The allegations advanced by the appellants in the particulars, therefore, in my view, relate to allegedly negligent acts performed by witnesses for the purpose of the pending litigation. They, therefore, fall within the scope of the immunity as established by strong authority. [204] It follows that the claims as asserted in the appellants’ Answer to Demand for Particulars by Matheson (at paragraphs 1(i) and (j), 3 (b) and (c), and 5(i) and (j)) all fall within the scope of witness immunity as established by strong authority. The claims advanced by the appellants in their Answer to Demand for Particulars by MDS (in paragraphs 1(k), (l) and 3(b) (c) and 5(k) and (l)) similarly fall within the scope of witness immunity as do the claims set out in the appellants’ Answer to Demand for Particulars by Savage: ¶ 1(l) and (m) and 3(l) and (m). Those claims, therefore, were properly dismissed by the chambers judge on the basis of witness immunity. 9. Does the immunity claimed by the respondents meet the test of necessity? [205] How does one assess whether it is necessary to extend witness immunity to cover a particular situation? An important discussion of this question, albeit in the context of the broader immunity for judicial and quasi-judicial proceedings, is found in the English Court of Appeal decision in Lincoln v. Daniels. Devlin, L.J., in an often cited passage, stated that the absolute privilege which covers proceedings in or before a court of justice may be divided into three categories: the first extends to everything that is said in the course of proceedings by judges, parties, counsel and witnesses and includes the content of documents in evidence. The second covers everything that is done from the inception of the proceedings onwards and extends to all pleadings and other documents brought into existence for the purpose of the proceedings starting with the writ or other document which institutes the proceedings. The third category, the one most relevant for our purposes, consists of situations in which it is necessary in practical terms to extend the immunity “... in order to protect those who are to participate in the proceedings from a flank attack ...”: at 753. He said at 751 - 2: I turn now to the third category. It is obvious that unless there were a category of this sort the absolute privilege granted for matters said and done coram judice might be rendered illusory. This is the consideration that animates the reasoning of the EARL OF HALSBURY, L.C., in Watson v. M’Ewan, ... ... It is not at all easy to determine the scope and extent of the principle in Watson v. M’Ewan. I have come to the conclusion that the privilege that covers proceedings in a court of justice ought not to be extended to matters outside those proceedings except where it is strictly necessary to do so in order to protect those who are to participate in the proceedings from a flank attack. It is true that it is not absolutely necessary for a witness to give a proof, but it is practically necessary for him to do so, as it is practically necessary for a litigant to engage a solicitor. The sense of LORD HALSBURY’s speech is that the extension of the privilege to proofs and precognition is practically necessary for the administration of justice; without it, in his view, no witness could be called. (emphasis added) [206] Lincoln v. Daniels does not provide a bright line test for the point at which the absolute privilege attaches. But it does show that the critical consideration is how closely connected the communication is to a judicial or quasi-judicial proceeding and that this must be assessed in light of the underlying purposes of the immunity. Devlin, L.J.’s reference to the necessity of protecting participants in the judicial proceeding from a “flank attack” refers to both of the immunity’s underlying purposes. Witnesses must be protected from flank attack to encourage candour and cooperation in connection with their evidence and as well to avoid a multiplicity of litigation in which the value of their evidence is tried over again. (See also, Larche v. Middleton (1989), 69 O.R. (2d) 400 (H.C.) and Gursikh Sabha Canada v. Jauhal (2002), 186 O.A.C. 362; O.J. No. 2005 (Q.L.), aff’g [2001] O.J. No. 4243 (Q.L.) (Sup. Ct.). [207] As noted earlier, witness immunity may apply to a person who does not testify. But whether the person has testified or not influences the necessity analysis. The underlying rationales of the immunity apply differently in the two situations. [208] Where a person testifies, the necessity analysis must take account of both the underlying rationales for witness immunity. That is, the necessity analysis in such cases must be concerned both with candour and cooperation of potential witnesses and with possible relitigation of the value or truth of the evidence which the witness has given. [209] When a person does not testify, only the underlying rationale related to candour and cooperation of potential witnesses is implicated. The focus of the necessity analysis, therefore, is on the closeness of the connection between the out-of-court activity and the person’s role as a witness. Is it so close that the preservation of candour and cooperation of potential witnesses requires that the out-of-court occasion have the same protection as testimony? The focus of witness immunity is not the content of the statement, but the occasion on which it is made. The protected occasion is testifying in court. The more closely connected the out-of-court statement or conduct is to that “occasion”, the stronger the case of necessity. [210] How does one assess the closeness of the connection between the out-of-court “occasion” and giving testimony in court? As we have seen, the cases have focussed mainly on the purpose of the conduct for which immunity is claimed. Several factors are relevant to assessing that purpose. If the out-of-court statement is required by law as a precondition to testimony, the case for immunity is compelling. Exactly the same concerns about candour and collateral attack apply to such statements as to in-court testimony: see e.g., Fabian and Carnahan. The same may be said about assessments ordered by the court for use in court proceedings, the preparation of testimony by professional persons engaged in litigation and inquiries by professional persons working on behalf of a party to litigation seeking information from a potential witness in pending litigation. Statements made by or to a person discharging a statutory or other public duty to gather evidence for pending or contemplated proceedings will generally be protected: see (X)Minors; Taylor; Ayangamn, and Evans. The need for candour is clear in those settings and the public duty strengthens the case for protection, particularly where there is a duty to cooperate with the investigation. [211] It is clear, however, that immunity does not extend to all steps in the investigation of a matter that may result in litigation. There is a distinction between the functions of investigators and witnesses even though the same person may, in the end, be both. To decide whether the immunity applies, the case law has looked to whether the action is based on the “judicial phase” of the proceedings, whether it relates to “matters of advocacy” or whether the “principal purpose” was to prepare evidence for court. These are primarily questions of fact and the burden is on the parties claiming the immunity to prove facts which bring them within it. [212] How does this analysis apply to Marsh given that no one from the firm testified? Does the record show that it is necessary to protect all of its out-of-court investigations and reports in order to make the immunity of witnesses effective? [213] The record does not permit the inference that all of Marsh’s activities were for the purpose of preparing evidence for trial. All we know is that Marsh was engaged to investigate and adjust the loss and that its involvement ended shortly before the Elliotts’ commenced their action on the policy. While it may be that, at some point, Marsh’s investigations and reporting were directed to gathering evidence for proceedings that were actually being considered, the facts in the record do not permit that inference. The burden of proving that the immunity applies was on Marsh and, in my view, the agreed facts fall far short of doing so. [214] Respectfully, I conclude that the chambers judge erred in dismissing the action against Marsh on the basis of witness immunity. Of course, my conclusion is based solely on the facts and the pleadings in the record and does not foreclose a different conclusion being reached on the basis of a more complete factual record. [215] In my view, the same may be said in relation to ICPB and Wilson and MDS and Levesque even though Wilson and Levesque testified. The timing of their involvement and the facts that Wilson and Levesque filed reports in court and testified strongly suggest that, at some point, particularly after the Elliotts started their action in March of 1993, ICPB’s and MDS’s investigations and reporting became focussed on preparing evidence for the pending litigation. However, they claim immunity for all investigations and reporting. There is no indication of whether any reports initially prepared during the investigation were ultimately placed in evidence. As noted earlier, there is no evidence that the purpose of their investigations or reports to Marsh was to gather evidence for litigation which was actually under consideration at the time. [216] Counsel for ICPB submits that where an investigator such as Mr. Wilson becomes a witness, his roles as an investigator and as a witness become so inextricably combined that it is impossible to distinguish them. This would be a good argument if there were any facts in the record to support it. But there are not. The onus was on ICPB to show that the appellants’ claim was barred by witness immunity. The appellants’ claim, as pleaded, relates to investigations and reports between the date of the fire and the date Royal pleaded arson in its defence. There is no indication that the purpose of ICPB’s involvement at any point was to gather evidence for a proceeding actually being considered. There is no evidence that any report it prepared for Marsh or Royal was ever placed in evidence or for that matter that Mr. Wilson’s testimony was based on his investigations prior to litigation being commenced. There is no basis in the record to conclude that the appellants’ claims amount to a collateral attack on Mr. Wilson’s trial evidence. There is no way of assessing, on this record, the relationship, if any, between that evidence and his reports and investigations up to April of 1993. In short, ICPB did not discharge its onus of showing that it was necessary to extend immunity in the sweeping fashion it claimed. In my view, the same must be said in relation to MDS and Lesvesque. Respectfully, I conclude that the chambers judge erred in holding otherwise. [217] I reach a different conclusion in relation to Matheson and Savage. The record indicates that in each case, the initial reports prepared by these respondents in September were placed in evidence at trial and relied on by the witnesses during their oral testimony. The filing of such reports is necessary under our Rules in order for the expert to testify. To my way of thinking, these facts affect the necessity analysis in one significant respect. It provides a factual basis for the submission that it is necessary to preclude the action against these individuals for their out-of-court investigation and reporting because those activities, as events unfolded, became so intermingled with their roles as witnesses that it is impossible to untangle them. That being so, the second rationale for witness immunity is strongly implicated. In effect, the appellants’ negligence action would require reassessment of the worth of the trial testimony of these witnesses in the insurance action because, as events unfolded, it is not now possible to distinguish between a claim for negligent investigation and a claim for negligent testimony. To avoid that relitigation is one of the reasons that the protection is afforded to out-of-court statements and conduct by witnesses. [218] I would conclude that Matheson Engineering (and Mr. Matheson) and Savage (and the Province) established the necessity of extending witness immunity to their activities which form the basis of the appellants’ negligence claims against them. The chambers judge was correct to dismiss the action as against them on the basis of the defence of witness immunity. 10. Conclusions concerning witness immunity: [219] The chambers judge, in my respectful view, erred in dismissing the claims against Marsh, ICPB and Wilson and MDS and Levesque on the basis of witness immunity, but did not err in doing so in relation to Matheson Engineering, Mr. Matheson, the Province and Mr. Savage. Of course, these conclusions are based on the record before the court and do not bind another judge in the event that this issue eventually is heard on the basis of a different and more complete factual record. V. DISPOSITION: [220] I would dismiss the appeal. Given the novelty and complexity of the issues and the fact that the appellants, while unsuccessful overall, succeeded in demonstrating significant errors on the part of the chambers judge, I would do so without costs. Cromwell, J.A. Concurred in: Freeman, J.A. Bateman, J.A. Appendix I (1) As set out in the statement of claim, the action concerns the individual respondent’s negligence in investigating for and reporting to the insurer between the date of the fire (August 30, 1992) and the date that Royal pleaded arson in its defence (April 13, 1993). The appellants allege that in carrying out their investigations and making their reports, which they knew Royal would rely on in responding to the appellants’ claim for indemnity on the policy, they breached a duty of care owed to the plaintiffs. The agreed facts stipulate that Royal relied on these reports in claiming that the fire was deliberately set in its defence filed on April 13, 1993. No other date is specified for Royal’s claim that the fire was deliberately set. (While the agreed statement of facts refers to Royal’s rejection of the appellants’ proof of loss on December 21, 1992, the stated reason for that rejection was that the claim was excessive, not that the fire was deliberately set.) (2) In order to put these allegations in the context of the appellants’ insurance action, it will be helpful to summarize the relevant chronology. What follows is based on the agreed facts, the pleadings and the reasons of Boudreau, J. in the insurance action which are referred to in the agreed facts. August 30, 1992 fire Savage begins investigation August 31, 1992 Marsh begins investigation September 1, 1992 ICPB begins investigation September 3, 1992 Savage 1st report (becomes evidence) September 18, 1992 Matheson 1st report (becomes evidence) October 21, 1992 MDS/Levesque report Elliotts’ proof of loss filed December 21, 1992 Royal rejects proof of loss via letter from Marsh February 18, 1993 end of Marsh investigation March 1, 1993 Elliotts commence action April 13, 1993 Royal files defence pleading arson July 27, 1994 ICPB investigation ends September, 1994 further Matheson report; further Savage report (both (date not specified) become evidence) September 6, 1994 trial of insurance action begins (3) No dates are given for the reports of ICPB or Marsh. The facts stipulate that Marsh’s investigation was ongoing between August 31, 1992 and February 18, 1993 and that ICPB was investigating between September 1, 1992 and July 27, 1994. (4) From the chronology, it is clear that the initial investigation by Savage and Matheson was completed on September 3 and 18 respectively. This was some two months before Royal rejected the Elliotts’ proof of loss, over 4 months before the Elliotts started their action and over 5 months before Royal formally alleged arson in its defence. While Savage and Matheson prepared supplementary reports in September of 1994, these are not part of the appellants’ allegations in the statement of claim which, as noted, are confined to the period from August 30, 1992 to April 13, 1993. It follows, that the allegations of negligence against Matheson and Savage, as pleaded in the statement of claim and understood in light of the agreed facts, are confined to their investigations and reports between August 30, and September 18, 1992. (5) We know that MDS reported to Marsh on October 21, 1992. We also know from the reasons of Boudreau, J. in the insurance action that MDS prepared a number of reports which were in evidence before him. But we do not know the dates on which they were prepared. There is no agreed fact that MDS prepared any report other than the one of October 21, 1992 which was relied on by Royal in pleading arson. I will take it, therefore, that the allegation against MDS as pleaded in the statement of claim is confined to the report of October 21, 1992 and the investigations leading up to it. (6) There is no information concerning the dates of reporting by Marsh and ICPB. But the agreed facts state that Marsh’s investigation was ongoing up to two weeks before the Elliotts commenced their insurance action and that ICPB’s investigation and reporting were ongoing until about a month before the trial of the insurance action started. (7) I conclude, therefore, that as pleaded in the statement of claim and understood in light of the agreed facts, the essence of the appellants’ claim against MDS, Matheson and Savage is negligent investigation and reporting up to the date of their initial reports of September 3, 1992 (in the case of Savage), September 18, 1992 (in the case of Matheson) and October 21, 1992 (in the case of MDS) all of which were prior to Royal’s rejection of the Elliotts’ proof of loss in December of 1992. In the case of Marsh, the allegedly negligent reporting continues to shortly before the insurance action was commenced and, in the case of ICPB, to the eve of trial. (8) This understanding of the claim has to be modified in light of the particulars which the appellants provided in relation to their claims as against MDS, Savage and Matheson. (This analysis is based on the Answer to Demand for Particulars delivered by the appellants in response to demands by MDS, Savage and Matheson. Although the particulars relating to Matheson refer to a report by Mr. Matheson dated September 18, 1998, I assume that this is a typographical error as the agreed facts and the reasons of Boudreau, J, in the insurance action refer to a Matheson report dated September 18, 1992.) The statement of claim focusses on negligent investigation and reporting, but the particulars allege negligence which is more directly linked to the role of the respondents, Savage, Matheson and MDS, as expert witnesses in the insurance action. Specifically, the particulars set out allegations of negligence by failing to retract or properly qualify their allegedly improper conclusions and persisting in justifying them vigorously after having been presented with reasonable contradictory expert opinion. It is a reasonable inference that these additional allegations are concerned with the stage of proceedings during which the views of the appellants’ experts were provided to the insurer and presumably to the respondents for their review. At that point, it is only reasonable to infer that the respondents were acting as opposing expert witnesses.