Decision tree or impenetrable thicket?

A difficulty with the balancing exercise described in R v Grant (noted here 18 July 2009) is that there is no indication of when an offence is sufficiently serious to make admission of the improperly obtained evidence a likely outcome.

This difficulty is illustrated in another decision delivered the same day as Grant and applying that case: R v Harrison [2009] SCC 34 (17 July 2009).

The appellant had been driving a hired car which was stopped on a highway by a police officer. The officer did not have sufficient grounds to carry out the search of the car which revealed two cardboard boxes. These contained a total of 35 kg (yes, kilograms) of cocaine.

Again, the Court split. This time the only dissenter was Deschamps J. She would have ruled the evidence admissible.

All judges agreed that the offence was very serious. How could they not? Deschamps J said the offence was at practically the highest point in the spectrum of importance of the public interest in adjudication on the merits (68). McLachlin CJ for the majority said (34) "While the charged offence is serious, this factor must not take on disproportionate significance."

This highlights the underlying forward-looking approach to protecting the repute of the administration of justice, encapsulated in Grant at (84):

" ... The short-term public clamour for a conviction in a particular case must not deafen the s. 24(2) judge to the longer-term repute of the administration of justice. Moreover, while the public has a heightened interest in seeing a determination on the merits where the offence charged is serious, it also has a vital interest in having a justice system that is above reproach, particularly where the penal stakes for the accused are high."

Critics (and I am one) will say this undervalues the public interest in prevention of the harm caused by criminal activity. Does the breach of Charter-protected interests in Harrison really outweigh the harm to society that 35 kg of cocaine could cause?

The majority in Grant described the evaluation of interests as a "decision tree" (Grant, 86). Unfortunately they did not draw the tree for us. Had someone stolen their crayons?

To be fair, I think the appropriate model is not a decision tree but rather a list of quantities of each of the three items. The decision depends on the total of the quantities. Score for the seriousness of the official misconduct, add a score for the impact of that misconduct on the defendant's rights, and then add (or subtract, if it goes against the trend of the others) another score for society's interests in adjudication of the case on the merits. If the end result is a total that supports admissibility, then that is the decision. The problem becomes - or is revealed to be - one of anticipating total scores that will be required for admission of the evidence.

Update: The approach to the model just mentioned, a move away from a decision tree, is consistent with the Court's summary of the way Grant is to be applied in R v Lafrance, 2022 SCC 32 (22 July, 2022) at [90] (majority judgment; the minority did not need to address this point). The first two factors are to be assessed cumulatively, and then against the third.