What happened
Ana Louise Hardwick was found dead in her bedroom in the Australian Capital Territory on the morning of 2 October 2002. She had died from neck compression before a fire was lit in the room. There was an abrasion on her nose, a complex abraded injury to the left side of her neck, and small bruises on each wrist. The pathologist could not determine whether the compression had been caused by ligature, rod or manual strangulation. Steven Wayne Hillier, her former domestic partner of 12 years and the father of their two children, was charged with her murder.
The prosecution case was wholly circumstantial. It rested on three main components identified by prosecuting counsel in final address: first, that Mr Hillier had the opportunity—he was alone on the night of Monday 30 September 2002 after arranging for the children to stay at his father's house; secondly, that he had a powerful motive arising from the bitterly contested Family Court proceedings. In June 2002 the Family Court had ordered that the children reside with their mother. An appeal had been filed. Interim shared-care orders were discharged on 20 September 2002, so the children were due to move to their mother. The Crown contended that Mr Hillier decided to “take the law into his own hands”. Evidence was led of a flurry of telephone calls in the week beginning 23 September 2002 to lawyers, psychiatrists and others directed to strengthening the pending appeal, which calls ceased abruptly after the weekend of 28-29 September. Thirdly, DNA evidence: a tape lift from the right-side flap of the collar of the pyjama top Ms Hardwick was wearing when she died produced a mixed profile consistent with her and Mr Hillier. Two prosecution experts assessed the probability that the DNA came from them rather than from an unknown person as very high. A defence expert, Dr McDonald, considered transfer via the children a real possibility but could not exclude Mr Hillier.
Additional evidence concerned chemical injuries to Mr Hillier's fingertips first noticed when he attended to give fingerprints on 1 November 2002. The Crown invited the jury to infer that the injuries were self-inflicted to avoid leaving prints at the scene. The trial judge gave a strong consciousness-of-guilt direction limiting the use of this evidence.
After a 15-day trial before Gray J and a jury in the Supreme Court of the Australian Capital Territory the jury deliberated for only a few hours before returning a verdict of guilty. Mr Hillier was convicted and sentenced.
He appealed to the newly established Court of Appeal of the Supreme Court of the Australian Capital Territory. The notice of appeal ultimately contained six grounds, the first two being that the verdict was unsafe and unsatisfactory and against the evidence and the weight of the evidence. By majority (Higgins CJ and Crispin P, Spender J dissenting) the Court of Appeal allowed the appeal, quashed the conviction and, although no formal order of acquittal was entered, the majority reasons made plain that they considered a verdict of acquittal the only proper outcome.
The Director of Public Prosecutions applied for special leave to appeal to the High Court. The application was referred to the Full Court. On 22 March 2007 the High Court granted special leave, treated the appeal as instituted and heard instanter, allowed the appeal, set aside the orders of the Court of Appeal and remitted the matter to that Court for rehearing before a differently constituted bench. Gleeson CJ agreed with the joint reasons of Gummow, Hayne and Crennan JJ. Callinan J agreed that the Court of Appeal had erred but would himself have ordered a retrial rather than a remittal.
Why the court decided this way
The joint reasons (Gummow, Hayne and Crennan JJ, with whom Gleeson CJ agreed) held that the majority in the Court of Appeal had misapplied the legal principles governing appellate review of a jury verdict said to be unsafe or unsatisfactory. The Court began by carefully tracing the statutory framework. Unlike the States, the Australian Capital Territory has never had a “common form” criminal appeal statute derived from the Criminal Appeal Act 1907 (UK). Appeals from the Supreme Court of the ACT were originally regulated by s 52 of the Seat of Government Supreme Court Act 1933 (Cth) and later by provisions of the Federal Court of Australia Act 1976 (Cth). Part 2A of the Supreme Court Act 1933 (ACT), inserted in 2001, gave the new Court of Appeal powers to set aside a verdict and order a verdict of not guilty to be entered (s 37O(1)(d)) and to order a new trial (s 37O(1)(e)), but said nothing expressly about the principles on which those powers were to be exercised.
The joint reasons held that there was no reason to read Pt 2A as conferring narrower powers than those developed under the common-form statute and applied to Territory appeals in Chamberlain v The Queen [No 2] (1984) 153 CLR 521. The applicable principles were those stated in Davies and Cody v The King (1937) 57 CLR 170 and authoritatively synthesised in M v The Queen (1994) 181 CLR 487. The appellate court must ask whether, on the whole of the evidence, it was open to the jury to be satisfied beyond reasonable doubt of the accused's guilt. Only if the evidence contains discrepancies, inadequacies or lacks probative force in such a way that, even making full allowance for the jury's advantage, there is a significant possibility that an innocent person has been convicted, must the verdict be set aside.
The majority in the Court of Appeal had instead identified five or six items—unidentified DNA on the pyjamas, bruises on the wrists, marks said to be consistent with handcuffs on the bed-head, footprints in soot, unidentified fingerprints and hair—and treated each as raising a reasonable hypothesis consistent with innocence. They concluded there was “a real possibility that another person was responsible”. The High Court held that this reasoning was erroneous because it examined those items in isolation. The joint reasons quoted at length from Chamberlain (No 2) and Plomp v The Queen (1963) 110 CLR 234 to the effect that in a circumstantial case all the circumstances must be weighed together; no single circumstance can be put to one side merely because, viewed alone, it is consistent with innocence.
The Court noted that the prosecution case, taken as a whole, included a strong motive (the imminent loss of custody and the sudden cessation of the intensive preparatory activity for the appeal), opportunity (being alone on the relevant night), the DNA evidence on the collar (which the jury could prefer to the defence expert's innocent-transfer hypothesis), the fingertip injuries, and Mr Hillier's own evidence which the jury was entitled to disbelieve. The Court of Appeal had acknowledged that “aspects of the evidence, particularly that relating to motive, timing and DNA … provided ample grounds for grave suspicion”, yet had allowed discrete items said to point to a third person to override that suspicion. That was not the task required by M.
Gleeson CJ added observations that the point about unfairness in the use of the telephone-call evidence in final address had not been taken as a ground of appeal and, even if it had been, would not necessarily have required the conviction to be quashed. Callinan J, while agreeing that the Court of Appeal's reasoning was flawed, considered that the failure to put the “take the law into his own hands” imputation squarely in cross-examination before pressing it forcefully in address constituted a material irregularity. He would have ordered a retrial so that a jury could assess Mr Hillier's response to that specific suggestion in the context of the whole case. The majority preferred remittal so that the Court of Appeal could itself perform the correct appellate task.
The disposition therefore turned on the need to correct a fundamental error in the application of appellate principle. Because the error required a fresh look at the whole of the evidence, the interests of justice were best served by remittal rather than by the High Court itself entering a verdict or ordering a retrial.
Before and after state of the law
Before Hillier the law on unsafe verdicts in the Australian Capital Territory was unsettled in its statutory expression but was understood to be the same as under the common-form criminal appeal provisions. Chamberlain (No 2) had decided that the Federal Court, when hearing Territory appeals, possessed the power and duty to set aside a verdict if it would be “unsafe or dangerous” to allow it to stand. Conway v The Queen (2002) 209 CLR 203 had dealt with the treatment of misdirections but left the unsafe-verdict ground for later clarification. M v The Queen supplied the authoritative synthesis: the appellate court does not ask whether it itself entertains a doubt; it asks whether the jury's verdict was one that was open to them on the whole of the evidence, making full allowance for their advantage in seeing and hearing the witnesses.
Hillier confirmed that Pt 2A of the Supreme Court Act 1933 (ACT) imports exactly the same principles. The spare statutory language does not produce a narrower jurisdiction. The decision reiterated, in the specific context of a wholly circumstantial case, the propositions from Plomp and Chamberlain that evidence cannot be examined in “hermetically sealed compartments”. After Hillier it is beyond argument that an ACT appellate court must undertake a cumulative assessment and must not treat the existence of items consistent with innocence, viewed in isolation, as sufficient to require an acquittal. The decision also reinforced the High Court's reluctance to grant special leave to the prosecution on an appeal against an acquittal, while making clear that correction of a clear misapplication of M is a sufficient “exceptional circumstance”.
The practical effect has been to align ACT appellate practice more explicitly with that in the States and to emphasise that the “unsafe” ground is not a licence for the appellate court to substitute its own view of the facts.
Key passages with plain-English translation
Paragraph [27] (joint reasons): “It is only where a jury's advantage in seeing and hearing the evidence is capable of resolving a doubt experienced by a court of criminal appeal that the court may conclude that no miscarriage of justice occurred … the ultimate question must always be whether the court thinks that upon the whole of the evidence it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty.”
Plain English: The appeal judges do not simply ask themselves whether they personally have a doubt. They must decide whether the evidence as a whole was capable of persuading a reasonable jury of guilt beyond reasonable doubt. Only if the evidence is so weak that no reasonable jury could convict may the appeal court intervene.
Paragraph [58] (joint reasons, quoting Chamberlain): “In a case depending on circumstantial evidence, the jury should not reject one circumstance because, considered alone, no inference of guilt can be drawn from it. It is well established that the jury must consider ‘the weight which is to be given to the united force of all the circumstances put together’.”
Plain English: You cannot look at one piece of evidence by itself, say “that could be innocent”, and stop. All the pieces must be put together like a cable; the question is whether the whole cable is strong enough to prove guilt and exclude any reasonable innocent explanation.
Paragraph [41] (joint reasons): the powers under Pt 2A “extend to setting aside a conviction ‘whenever it appears unjust or unsafe to allow the verdict to stand because some failure has occurred in observing the conditions which, in the court's view, are essential to a satisfactory trial, or because there is some feature of the case raising a substantial possibility that, either in the conclusion itself, or in the manner in which it has been reached, the jury may have been mistaken or misled’.”
Plain English: The appeal court can quash a conviction either because something went badly wrong at the trial or because, even though the trial was fair, the result is one that no reasonable jury could have reached on the evidence.
Paragraph [105] (Court of Appeal reasons, quoted and criticised): the majority said it was “impossible … to conclude that it was open to the jury to find that the guilt of the [accused] had been proven beyond reasonable doubt” because of “other aspects of the evidence, such as that relating to the unusual features of the injuries she suffered and the apparent use of the handcuffs [which] make it difficult to reconstruct what actually occurred on the night in question and the evidence suggesting that another person may have been present”.
Plain English translation of the High Court's criticism: The appeal judges were saying “we cannot work out exactly what happened, so the jury could not have been sure”. That is the wrong question. The jury did not have to reconstruct every detail; they only had to decide whether, on all the evidence, the only reasonable conclusion was that Mr Hillier was the killer.
What fact patterns trigger this precedent
Hillier is triggered whenever an intermediate appellate court in a Territory or, by analogy, a State is asked to set aside a conviction on the unsafe-verdict ground in a case that depends on circumstantial evidence. The precedent applies with particular force where the appellate court has:
- identified discrete items (unidentified DNA, marks on the body, footprints, unidentified fingerprints or hairs) each capable of an innocent explanation;
- treated those items as creating a “real possibility” of a third-party perpetrator without weighing them against the cumulative strength of motive, opportunity, forensic links, post-offence conduct and the accused's own evidence; or
- expressed its conclusion in terms of difficulty in reconstructing events or the existence of alternative hypotheses without asking whether a reasonable jury could exclude those hypotheses.
The case is also relevant whenever the prosecution relies on a combination of motive (especially custody or family-law disputes), sudden cessation of activity consistent with a change of plan, and low-copy DNA evidence. It further applies to any appellate reasoning that treats expert disagreement on the source of DNA as automatically creating reasonable doubt without reference to the jury's entitlement to prefer one expert over another having seen them give evidence.
Because the decision rests on the M test, it governs any appeal in which the court is asked to decide whether “upon the whole of the evidence it was open to the jury to be satisfied beyond reasonable doubt”.
How later courts have treated it
The joint reasons carefully positioned Hillier as an application, not an alteration, of M, Chamberlain (No 2) and Plomp. Subsequent appellate courts have therefore treated it as confirmatory of the obligation to undertake a cumulative assessment. The decision has been cited for the proposition that an appellate court must not examine circumstantial evidence “piecemeal” or in “hermetically sealed compartments”. It has reinforced that the question is always whether it was open to the jury to convict, not whether the appellate court would have convicted.
The emphasis on the statutory powers under Pt 2A has been accepted as settling the content of appellate review in the ACT. Courts have continued to grant prosecution appeals only in exceptional cases, but have readily done so where an intermediate court has demonstrably failed to apply the “whole of the evidence” test. The discussion of the telephone-call evidence and the Browne v Dunn issue has been treated as a reminder that unfairness in the conduct of the trial must be raised at first instance if it is to found appellate intervention. Overall, Hillier has been followed as restating orthodoxy rather than breaking new ground, but its clear condemnation of isolated analysis of exculpatory items has tightened the discipline required of appellate courts in circumstantial cases.
Still-open questions
The judgment leaves open precisely how an appellate court is to weigh expert disagreement on the statistical significance of a mixed DNA profile when that evidence forms only one strand in a circumstantial cable. The joint reasons noted the three experts' differing opinions but did not prescribe the weight to be given to the possibility of secondary transfer via the children. A future court will have to decide whether a jury may rationally prefer the prosecution experts' “very high probability” assessment even in the face of a credible innocent-transfer hypothesis.
The precise boundary between permissible appellate evaluation of “grave suspicion” that falls short of proof and impermissible substitution of the court's own view remains a matter of degree. The Court of Appeal on remittal will have to perform the M assessment without the High Court having itself weighed every item. That task may raise questions about the extent to which an appellate court may take into account the jury's implicit rejection of the accused's evidence.
Callinan J's separate reasons leave open whether a material irregularity in the use of circumstantial evidence in final address, without cross-examination, can of itself justify quashing a conviction even when the evidence taken as a whole would otherwise support it. The majority did not decide that point. Future cases will have to determine whether such an irregularity requires a new trial or can be cured by a strong jury direction.
Finally, the decision does not resolve how the “real possibility” language used by the Court of Appeal should be reconciled with the criminal standard. The High Court made clear that a mere possibility consistent with innocence is not enough to require acquittal if the combined evidence excludes every reasonable hypothesis. The exact formulation that trial judges and appellate courts should use when directing or reasoning about “reasonable” as opposed to “mere” possibility remains a live practical question.