Lee v Lee
(2019) 266 CLR 129 · [2019] HCA 28 · (2019) 372 ALR 383
Summary updated 4 November 2026
Overview
Lee v Lee; Hsu v RACQ Insurance Limited; Lee v RACQ Insurance Limited [2019] HCA 28 is a decision of the High Court of Australia (Kiefel CJ, Bell, Gageler, Nettle and Edelman JJ), given on 4 September 2019, in three appeals from the Court of Appeal of the Supreme Court of Queensland. The appeals turned on who was driving a Toyota Tarago in a collision in which the appellant in the first appeal, then 17, was rendered an incomplete tetraplegic [14], [16]. The joint reasons of Bell, Gageler, Nettle and Edelman JJ held that the Court of Appeal erred by dismissing the appeals on the footing that the trial judge's decision was neither glaringly improbable nor contrary to compelling inferences, when it was its duty to decide for itself which of two competing hypotheses was more probable [56]. Kiefel CJ agreed with the orders proposed in the joint reasons for the reasons given there, adding observations of her own [1]. The High Court allowed all three appeals, found that the father was the driver, entered judgment for the son against RACQ Insurance Limited for $3,350,000, dismissed the insurer's counterclaim against the son and gave judgment for the mother and the father on it [65]–[66].
Facts
The collision happened at around 1.30 pm on 25 September 2013 [17]. The Toyota, rounding a bend on a relatively narrow, unmarked bitumen road on North Stradbroke Island, met a Nissan Patrol driven by David Hannan and collided with it head-on on the Nissan's side of the road [16]. In the Toyota were the appellant, then 17, his mother (Chao-Ling Hsu, the appellant in the second appeal), his father (Chin-Fu Lee, the appellant in the third appeal) and his two younger brothers [14]. The collision was caused by the negligence of the Toyota's driver, and the sole issue at the trial was the identity of that driver [16].
Mr Hannan carried his dog about 30 metres away and returned to the vehicles, which he estimated took between 30 and 90 seconds, and at that time no one was in the Toyota's driver's seat and the father was standing between the first and second rows of seats [17]. A police officer, Mr Harvey, spoke with the father through a civilian interpreter, believed the father identified himself as the driver, and recorded the father in the driver's seat and the appellant behind the driver [18]. The investigating officer, Senior Constable Pepper, noted apparent bloodstaining on the driver's airbag and that the driver's seatbelt was buckled [19]. DNA testing established that the blood on the airbag came from the appellant [21].
The appellant sued the father, the mother and RACQ Insurance Limited (the RACQ), the compulsory third-party insurer of the Toyota, for damages for negligence, pleading that the father was the driver [15]. The RACQ counterclaimed in deceit against all three for the recovery of payments made on the representation, pleaded as false, that the father was the driver [22]. The RACQ's case was that the blood on the airbag came from direct contact with the appellant's bleeding face and that, immediately after the collision, the father pulled the appellant back from the driver's seat into the seat behind it [22].
The appellant and the mother each gave evidence that the father was driving [23]. On the appellant's case, the father, who was photographed with blood on his hands, likely wiped the appellant's blood from his hands on the deflated airbag while trying to help the trapped mother [23]. The father did not give evidence, but a statement he made to an insurance investigator, in which he said he was driving, was tendered in the RACQ's case [24].
Dr Grigg, a mechanical engineer called by the appellant, gave evidence that an airbag of this type inflates in about 0.04 seconds and deflates about 0.2 seconds after impact, and that the Toyota's front seatbelt pre-tensioners fire when the airbag deploys and produce a powerful force pulling the wearer back against the seat [29]–[30]. The bloodstaining was predominantly on the windscreen side of the airbag (when inflated), and there was none on the central front portion that faced the driver [32]. Dr Robertson, a specialist in forensic medicine and pathology retained by the RACQ, considered the staining consistent with close contact with a person with heavily bleeding facial injuries, and thought the absence of fingerprints or handprints made transfer by a third person less likely [33]. Dr Hallam, a DNA forensic biology consultant, considered it possible that the blood was transferred to the airbag by an intermediary [37].
Procedural history
The trial judge, Boddice J of the Supreme Court of Queensland, formed an adverse view of the credibility of the appellant and the mother, rejected their evidence and found that the appellant was driving [15], [25]. His Honour mistakenly recorded that he had made allowance for the appellant's evidence being given through an interpreter, when the appellant gave evidence in English [25]. The trial judge found, contrary to the case put by each party, that the driver had not been wearing the seatbelt [40]. He reasoned that if the father could get himself out of the driver's seat within 90 seconds, there was no reason why he could not have moved the appellant into the back seat in the same time [39]. He found Dr Robertson's evidence highly persuasive and rejected as implausible the suggestion that the father transferred the blood from his hands [41]. The appellant's claim was dismissed, judgment was given for the RACQ on its counterclaim, and the appellant and his parents were ordered to pay the RACQ $439,840.96, with the parents ordered to pay a further $234,428.41 [25].
The Court of Appeal (Fraser, Philippides and McMurdo JJA), in a leading judgment by McMurdo JA, identified critical errors in the trial judge's findings and concluded that, save for the DNA evidence, it was "much more likely" that the appellant was not the driver [26]. McMurdo JA considered it much more difficult for the father to move the paralysed appellant from the driver's seat into the rear seat than to get himself out of that seat [44]. He found that the driver was likely wearing the seatbelt, and said that if the driver was not wearing it, as the trial judge had found, there was "no real prospect that the driver was the appellant" [45]–[46]. On the DNA evidence, he regarded the RACQ's hypothesis as "quite probable" and said the question for the trial judge was which of the two hypotheses was more probable [49]. He concluded that the trial judge had not misused his advantage in seeing and hearing the appellant and the mother, and that the decision was neither "glaringly improbable" nor "contrary to compelling inferences", and the appeals were dismissed with costs [26]. On 16 November 2018, Bell, Keane and Nettle JJ granted the appellant and his parents special leave to appeal [27].
Issues
The appeals were brought on two grounds [27]. The first was that, on the inferences to be drawn from the DNA evidence, McMurdo JA failed to engage with a critical argument based on unchallenged expert evidence on which the appellant relied [27]. The second challenged the Court of Appeal's restraint in the face of the trial judge's advantage, where the finding that the appellant was the driver was said to be contrary to the compelling inferences from uncontroverted evidence [27].
Held
Bell, Gageler, Nettle and Edelman JJ
Their Honours treated the complaint about the adequacy of the Court of Appeal's reasons as subsumed by the second ground [28]. McMurdo JA referred to Dr Grigg's evidence when reviewing the evidence but did not return to it when deciding between the competing inferences from the DNA evidence [51]. Because success on the first ground depended on success on the second, their Honours said the appeals did not provide the occasion to consider when an appellate court's failure to deal with a particular line of reasoning might be legal error [52].
On the second ground, their Honours said of McMurdo JA's concluding passage: "This passage is not without difficulty." [54] The rejection of the evidence of the appellant and the mother did not establish that the appellant was the driver, and the trial judge's conclusion rested on inferential reasoning from circumstantial evidence, particularly the forensic evidence [54].
Their Honours stated that a court of appeal is bound to conduct a "real review" of the evidence at first instance and of the judge's reasons, citing Fox v Percy (2003) 214 CLR 118 and Robinson Helicopter Co Inc v McDermott (2016) 90 ALJR 679 [55]. Appellate restraint unless findings are glaringly improbable or contrary to compelling inferences applies to findings likely to have been affected by the trial judge's impressions of the credibility and reliability of witnesses seen and heard, including secondary facts based on those impressions combined with other inferences [55]. Beyond that, quoting Warren v Coombes (1979) 142 CLR 531, "in general an appellate court is in as good a position as the trial judge to decide on the proper inference to be drawn from facts which are undisputed or which, having been disputed, are established by the findings of the trial judge" [55]. Here the trial judge's findings of primary fact were not disturbed, but the Court of Appeal found that, in material respects, the inferences he drew from them were wrong [55]. On the Court of Appeal's analysis, the trial judge's seatbelt finding, if correct, meant there was no real prospect that the appellant was the driver, and the RACQ's account of moving the appellant in less than 90 seconds was unlikely [55].
Having rejected the essential planks of the trial judge's reasoning, it was not to the point for the Court of Appeal to ask which hypothesis the trial judge considered more probable, or whether he had been unduly influenced by the DNA evidence [56]. It was an error to dismiss the appeals in a case the Court of Appeal itself called very closely balanced on the footing that the trial judge's decision was neither glaringly improbable nor contrary to compelling inferences [56]. "It was the duty of the Court of Appeal to decide for itself which of the two hypotheses was the more probable." [56] In words the joint reasons quoted from Dearman v Dearman (1908) 7 CLR 549, which itself quoted The Glannibanta (1876) 1 PD 283, the Court of Appeal had to persist in "weighing [the] conflicting evidence and drawing its own inferences and conclusions" [56]. It did not, and the second ground was made good [56]. That restraint meant McMurdo JA did not return to consider whether his tentative conclusion was displaced by the DNA evidence [57]. His agreement with the trial judge that the DNA evidence was persuasive was apt to overlook that the trial judge's acceptance of Dr Robertson's opinion assumed the appellant was unrestrained by the seatbelt, and so to overlook the significance of Dr Grigg's unchallenged evidence [57].
Their Honours held that a new trial was not appropriate, because the evidence of the appellant and the mother was not critical to this largely circumstantial case, and that the High Court could give the judgment that ought to have been given at first instance (citing s 37 of the Judiciary Act 1903 (Cth)) [58]. The RACQ did not challenge the Court of Appeal's conclusion that the driver was wearing the seatbelt [58]. Their Honours rejected the RACQ's argument that the appellant had to establish how his blood came to be on the airbag: the question was whether it was more probable than not that the father was the driver [59]. As McMurdo JA reasoned, the appellant's presence in the rear passenger seat within 90 seconds of the collision was "a circumstance powerfully in favour of the conclusion that he was not the driver" [59]. The likelihood that the driver wore the seatbelt made Dr Robertson's direct-contact opinion distinctly less sustainable in light of Dr Grigg's evidence [59].
The RACQ relied on photographs in Dr Grigg's second report showing an airbag of the same description partly deflated, but their Honours said the evidence was silent on whether the photographs showed the airbag at the moment of deployment or later, and that even if deflation is gradual, within about 0.2 seconds of impact the airbag would have been subsiding towards the steering wheel [60], [62]. Contrary to the Court of Appeal's conclusion, Dr Robertson's opinion that a small movement of the face could explain the staining on the windscreen side of the airbag was not readily reconciled with the finding that the appellant would have been held against the driver's seat [62].
Taking all the circumstances into account, their Honours noted that the father appeared to have been cooperative at the scene and told Mr Harvey, about an hour after the collision, that he was the driver [63]. As the Court of Appeal recognised, the RACQ's hypothesis that the father moved the paralysed appellant to the rear seat in less than 90 seconds was "to a high degree, improbable" [63]. A civilian, Mr Hough, saw the father cradling the appellant and touching his face, with bright red blood on the appellant's face, so a likely explanation for the blood on the father's hands was that it came from the appellant [64]. How the blood came to be largely on the windscreen side of the airbag was unknown, but given the seatbelt finding the likelihood of direct contact with the appellant's face could not be accounted greater than the likelihood of transfer from the father's hands [64]. Their Honours concluded: "McMurdo JA's tentative conclusion that it is much more likely that the father was the driver is correct for the reasons that his Honour gave." [65] That conclusion was not weakened by the blood on the airbag, given Dr Grigg's unchallenged evidence, and the appeals had to be allowed [65].
Kiefel CJ
Kiefel CJ agreed with the orders proposed by the other four judges for the reasons they gave, and added observations on the Court of Appeal's analysis of the evidence [1]. Her Honour said it is useful to bear in mind what fact a party has undertaken to prove, the evidence by which it seeks to prove it and the evidence by which the other party meets that case [3]. The rejection of the evidence of the appellant and the mother did not mean that the appellant could not otherwise prove his case, or that the RACQ had established that he was the driver [4]. Her Honour understood McMurdo JA to have meant that, unless some evidence prevented his preliminary conclusion that the appellant was not the driver, the appellant's case was made out [6]. Given McMurdo JA's finding that it was improbable that the father could have moved the appellant in the time available, it was not obvious that the DNA evidence compelled a conclusion that the appellant was in the driver's seat, and each of its two hypotheses had to be considered in light of the level of certainty attending that finding rather than in isolation from it [9]. Other matters that had to be considered were not taken into account, and, as the joint reasons explain, it was by this means that the Court of Appeal fell into error [10]. The DNA evidence was based on an erroneous assumption, because Dr Robertson assumed the appellant was unrestrained by the seatbelt, while McMurdo JA found the driver was more probably than not wearing it [11]. On Dr Grigg's unchallenged evidence, a restrained driver would have been pulled back into the seat on impact, so that, had the appellant been the driver, the blood could not be explained by direct contact between his face and the airbag [12]. In her Honour's view, the seatbelt finding and Dr Grigg's evidence "rendered the DNA evidence of little, if any, value", the RACQ could not make out its case, and nothing prevented the Court of Appeal acting on its foreshadowed conclusion [13].
Orders. In the first appeal (B61/2018), the parties agreed upon the orders, and the High Court allowed the appeal, ordered the third respondent (the RACQ) to pay the appellant's costs, and set aside the Court of Appeal's order of 1 June 2018 [66]. In its place, the appeal to the Court of Appeal was allowed, the trial judge's orders of 23 March 2017 were set aside, judgment was entered for the appellant against the RACQ for $3,350,000.00 (clear of rehabilitation expenses under the Motor Accident Insurance Act 1994 (Qld)), with the date of judgment to take effect on 23 March 2017 under r 660(3) of the Uniform Civil Procedure Rules 1999 (Qld), and the RACQ's counterclaim against him was dismissed [66]. The RACQ was ordered to pay his costs of the claim on the indemnity basis, and his costs of the counterclaim and of the appeal to the Court of Appeal on the standard basis [66]. In the second and third appeals (B62/2018 and B63/2018), the appeals were allowed with costs, the Court of Appeal's orders were set aside, the trial judge's orders were set aside and judgment was given for the mother and the father respectively on the RACQ's counterclaim, with the RACQ to pay their costs of the counterclaim and of the appeal to the Court of Appeal on the standard basis [66].
Why it matters
The joint reasons, with which Kiefel CJ agreed, explain the limits of appellate restraint on a rehearing: restraint attaches to findings likely affected by the trial judge's impressions of witnesses, and where an appeal court rejects the essential planks of the trial judge's inferential reasoning it must decide for itself which hypothesis is more probable [1], [55]–[56]. The joint reasons left open when an appellate court's failure to deal with a particular line of reasoning amounts to legal error [52].
Cited in 173 later casesRelies on 11 casesSee every one in casewise