FIghting Bias Against Drug Users in Jury Trials

People who use drugs experience serious stigma in Canadian society. They are blamed for problems they did not create and denied the compassion extended almost automatically to others. When a person who uses drugs is charged with a criminal offence, there is a real risk that bias in the community will result in an unfair trial.

A recent decision from the Alberta Court of Appeal takes that reality seriously. In R v Sullivan, 2026 ABCA 140, the Court unanimously overturned a second degree murder conviction because the accused was not permitted to screen prospective jurors for bias against drug users, and because the jury was never told not to reason from that bias. This is an important and welcome development in the criminal law.

What Happened

John Sullivan admitted killing Camille Houle. His defence was self-defence: he testified that Mr. Houle had tried to inject him with a lethal dose of fentanyl. Mr. Houle had significant levels of methamphetamine in his system when he died. Mr. Sullivan described himself as addicted to fentanyl and methamphetamine. The killing took place in a residence where people were regularly using heroin, methamphetamine, and fentanyl.

Essentially every non-expert witness at the trial was a person who used drugs. So was the accused. So was the deceased.

Before the trial, defence counsel asked to "challenge the jurors for cause" — to question prospective jurors about whether they could be impartial. His concern was precise: jurors might assume that people who use drugs are less believable, less reliable, and more likely to be violent. Given that the case turned on whether Mr. Sullivan or Mr. Houle was the aggressor, and on whether the jury believed Mr. Sullivan's testimony, these assumptions could decide the outcome.

The trial judge said no. Drug use, in his view, was not an "immutable" characteristic like race — it was, he wrote, alterable by conscious choice, and something common sense tells us a person would want to change. In any event, he found the evidence did not establish widespread bias against drug users.

Instead, he promised general anti-bias instructions. Those instructions were given twice, and they named gender, race, ethnicity, sexual orientation, and employment status as sources of bias. They never mentioned drug use. And although the trial judge told counsel he would repeat them in his final charge to the jury, he did not. The jury retired to deliberate having received no warning at all against concluding that Mr. Sullivan was more likely to be a killer because he used drugs.

Mr. Sullivan was convicted. The Court of Appeal set that conviction aside and ordered a new trial.

"Immutability" Was Always the Wrong Question

The trial judge's reasoning rested on the idea that prejudice only counts when it attaches to something a person cannot change. Because a person could, in principle, stop using drugs, bias against them was somehow less deserving of the law's attention.

The Court of Appeal firmly rejected this proposition. A personal characteristic, the Court held, is not excluded from consideration just because it involves an element of personal choice or is not entirely beyond a person's control.

The Court explained why “personal choice” is a red herring at paragraph 24:

A wide range of characteristics can create a risk of prejudice. While an immutable characteristic may give rise to generic prejudice, there is no necessary correlation between the two. Some immutable characteristics – for example, height or eye colour – do not ordinarily create a risk of generic prejudice, while other non-immutable characteristics – for example, membership in a gang or being a sex worker – can give rise to generic prejudice.

In short, “immutability” says nothing about whether prejudice exists. The only question that matters is whether there is a realistic possibility that some jurors will not be able to set their assumptions aside.

There is also something worth naming in the trial judge's phrasing. The suggestion that drug use is simply changeable by choice, and that any sensible person would want to change it, is not a neutral legal observation. It is itself a version of the belief that puts people who use drugs outside the circle of ordinary human sympathy — the idea that they have brought their circumstances on themselves and are therefore fair game for judgment. Whatever one thinks of the underlying moral question, a criminal trial is not the place for a jury to work it out on the back of an accused person's liberty.

Believing Anonymous Strangers: The Evidence Question

Sullivan also offers a practical takeaway for defence counsel thinking about how to show that bias exists in their community. Trial counsel offered three Edmonton news articles — about opposition to a supervised consumption site, about drugs and crime, about a proposed bylaw banning open drug use — and a collection of posts from the Edmonton subreddit. The Reddit posts are ugly. They describe people who use drugs as violent, filthy, and disposable. One commenter suggested they go into the woods and use drugs there, out of sight.

The trial judge gave this evidence “no weight”. The posts were anonymous, unverifiable, and possibly written for provocation. The news articles, he said, merely reported social problems and public reaction — not prejudice.

The Court of Appeal took a dim view of this assessment. The Court accepted that social media posts should be approached with some caution. But they did not need to establish as fact that Edmontonians in general, or even some slice of Edmontonians, held the biased views articulated in the comments or reddit posts. Rather, needed only show a reasonable possibility that prejudicial attitudes exist in the community and could taint a jury. That is a much lower bar.

Prejudicial beliefs carry social consequences, so people express them where they can do so anonymously. The polished, attributable, institutionally credible sources a court would ordinarily prefer are precisely the sources least likely to reveal the bias that actually exists. To demand only respectable evidence of prejudice is to risk that prejudice goes unproven.

Why This Matters

Drug use in this country is criminalized, moralized, and policed. The people most affected experience intersecting and compounding discrimination across a host of other social axes, like race, gender, income, and disability. While a criminal courtroom cannot cure social prejudice, it can prevent that prejudice from resulting in an unfair trial.

A jury trial is supposed to be a proceeding in which the state must prove its case, and in which the accused person is presumed innocent. That presumption is worth nothing if twelve people walk into the box already convinced that people like the accused are liars and thugs. Sullivan recognizes this. It says the courts must take the risk seriously, must let defence counsel probe for it, and must speak to the jury directly about it when it is present.