Edit: In R v Ceissman [2010] NSWCCA 50, Latham J sets out the approach taken to coincidence evidence in NSW.

CW v The Queen [2010] VSCA 288 went to the Court of Appeal as an interlocutory appeal from the trial judge’s ruling, under s 295 of the Criminal Procedure Act 2009.

The accused was charged with setting fires in three commercial properties. The fires occurred on the same night, in the same suburb. The prosecution alleged the accused had issues with each of the torched businesses.

Permission was granted by the trial judge to allow the evidence for each count of arson to be used cross-admissibly in deciding the other two. If the jury members were satisfied the accused started one of the fires, they could use that to draw the inference that he’d caused the others. They could (if they wanted to) conclude that it was highly improbable that different people might have lit each one.

This was a coincidence ruling under s 98. These are rare compared with tendency rulings under s 97. Unlike tendency cases which typically rely on a striking similarity or calling card feature, there was no suggestion here that the way each fire was lit made it uniquely identifiable.

The prosecution relied on the circle of persons reasoning from R v Perry (1982) 150 CLR 580. (Perry was the notorious case where the accused’s husband was poisoned by arsenic, and her first husband and brother had also mysteriously died of arsenic poisoning in separate incidents). As in Perry, the prosecution asserted it was innately improbable that three persons known to the accused would fall victim to the same kind of crime in a short space of time, unless the accused had committed all of them.

Maxwell P, Buchanan and Neave JJA [at 19]:

The argument for the applicant was that the present case was distinguishable on the facts, in that the class of business associates was much wider, and more diffuse, than the groups of persons (characterised as ‘small family circles’) under consideration in the poisoning cases. Hence, it was contended, the mere fact that all three victims of these fires were members of that class did not give the evidence any particular probative force. We disagree. Not only was each victim a person with whom the applicant had had business dealings, but he was in a current dispute with each of them. In the circumstances, the judge was entitled to conclude, as would the jury as the tribunal of fact, that (in the language of Perry) it would be contrary to ordinary experience for this series of fires, affecting these particular victims, to have occurred by coincidence.

The Court rejected the need for strikingly similar features [at 21]:

Counsel for the applicant contended, further, that it was necessary for the prosecution to show some ‘striking similarity’ in the circumstances, before it could be concluded that the evidence had significant probative value. He relied for this purpose on statements in the authorities that, where what was in issue was not whether a crime had been committed but the identity of the perpetrator, a stringent requirement of similarity should be applied. With respect, this submission is misconceived. As we have already explained, the basis of the coincidence reasoning in a case such as the present is quite different. It relies on the existence of a relationship which uniquely links the accused person with two or more victims of similar crimes. There is no separate requirement of ‘striking similarity’.

As mentioned earlier, the prosecution did rely on certain similarities in the fire events, as follows:

* each fire was deliberately lit by spreading an accelerant;

* each fire was lit in commercial premises;

* each fire was started at or near the entrance door to the premises;

* all of the fires occurred on the same evening, within a four hour period; and

* all of the fires were in the same suburb.

The argument for the applicant was that the use of an accelerant was ‘the stock-in-trade of the arsonist’ and that, likewise, there was nothing distinctive about fires being lit at the front door of commercial premises. We accept that, by themselves, these features might be insufficient to give the evidence significant probative value. The close proximity in time is of more significance, at least in pointing to the improbability of there having been more than one arsonist active on this particular night. But nothing further need be said on this aspect since, as we have said, what gave the evidence its significant probative value was the link between the accused and each intended victim.

The Court also found that the evidence passed the s 101 test, describing the trial judge’s finding [at 26] as ‘unimpeachable’. The Court did say that applying s 101 could sometimes put the test as high as the no rational explanation test from R v Pfennig (1995) 182 CLR 461, but adopted Spigelman CJ’s reasoning in R v Ellis (2003) 58 NSWLR 700 in finding that the no rational explanation test isn’t generally appropriate under the Evidence Act 2008.