# Lesson 10: Why People Use Substances and Why Risks Differ

*Companion-podcast transcript • Sarah & Kiffer*  
*Office Hours episode to listen to after working through the lesson*

---

**Sarah:** Welcome back to Office Hours. I'm Sarah.

**Kiffer:** And I'm Kiffer. We'll be discussing drug toxicity deaths, residential schools and racism in health care, and support lines are listed on the lesson page. Lesson ten is about why people use substances, and if you ask a room of adults why teenagers try cannabis, a lot of them will say peer pressure.

**Sarah:** In the 2023 BC Adolescent Health Survey, among students aged twelve to eighteen who had used cannabis and no other substance, three percent said they used it the last time because they felt pressured or wanted to fit in.

**Kiffer:** And thirty percent said they used it because their friends were doing it. So friends come up often, and pressure, as a reason students picked for themselves, hardly comes up at all. Both numbers come from the same students. Whether they'd call it pressure when it comes from their friends is another question.

**Sarah:** Question one. What can the reasons people give actually tell us about why they use substances?

**Kiffer:** Quite a lot, as long as we're clear about what kind of evidence they are. A survey reason is a person's own account of one occasion, chosen from a list someone else wrote.

**Sarah:** And in the BC survey, students could mark every reason that applied, and they were asked about the most recent time they used. So the answers describe last times, and they're limited to the options on the list. That list doesn't line up with Cooper's four motives, either. The most common reason, wanting to experiment, at fifty-three percent, doesn't sit comfortably in any of them.

**Kiffer:** I think that's a real gap in the model. Curiosity is a reason to try something, and Cooper's model was built around the motives behind drinking. A model of repeated use and a model of first use may need different categories.

**Sarah:** Back to the pressure question, though. If teenagers don't label it pressure, the three percent could be an undercount.

**Kiffer:** It could. The survey can't separate joining friends to share a good time, which is a social motive, from joining them to avoid being left out, which is conformity. What it can tell you is that young people rarely describe their own use as pressured, and that's worth knowing before anyone writes a prevention message built around peer pressure.

**Sarah:** Here's where I think people go wrong with these numbers. Among youth who used cannabis daily, about half said stress was a reason the last time they used a substance. It's easy to turn that into the claim that most young people who use because of stress end up using daily.

**Kiffer:** Let's test that with hypothetical numbers. Picture one thousand young people who have used cannabis. One hundred of them use daily, and half of those, fifty people, give stress as a reason. Among the other nine hundred, one in five gives stress as a reason. Here's the question for everyone listening. Of all the young people who give stress as a reason, what share use daily? Pause the audio if you'd like more time. Take a few seconds.

*(Pause)*

**Sarah:** One in five of nine hundred is one hundred eighty. Add the fifty who use daily, and two hundred thirty young people give stress as a reason. Fifty out of two hundred thirty is about twenty-two percent. So in this hypothetical, more than three quarters of the young people who use because of stress don't use daily.

**Kiffer:** Right. "Half of the young people who use daily give stress as a reason" and "most young people who give stress as a reason use daily" are two different statements, because they divide by different groups. The first divides by the one hundred who use daily. The second divides by the two hundred thirty who give stress as a reason.

**Sarah:** But stress still goes with daily use. Among the seven hundred seventy who didn't mention stress, only fifty use daily, which is about six and a half percent.

**Kiffer:** So both things are true at once. Giving stress as a reason is associated with daily use, about twenty-two percent compared with six and a half, and most young people who use for stress still don't use daily. That fits the link between coping motives and later problems. It's also a reminder that a survey taken at one point in time can't tell you whether the stress or the daily use came first.

**Sarah:** Question two. The BC survey found that seventeen percent of youth who felt their parents or guardians knew what they did in their free time most or all of the time had tried cannabis, compared with forty percent of youth whose parents rarely or never knew. Does that mean we should tell parents to keep closer track?

**Kiffer:** That's a big gap, and it's tempting to read it that way. The measure, though, is what young people think their parents know, and that knowledge can come from several places. So here's a question. Give one explanation for that gap in which nothing about parents' supervision changes what young people do. Take a few seconds.

*(Pause)*

**Sarah:** The direction could run backwards. A young person who has started using cannabis might tell their parents less about their evenings. Then the parents know less because of the use.

**Kiffer:** Researchers have looked at where parents' knowledge comes from. In a 2000 study of about twelve hundred fourteen-year-olds in Sweden, Kerr and Stattin found that what parents knew was linked to better adjustment, and that young people's own willingness to tell their parents things explained more of that link than parents' efforts to track and watch them.

**Sarah:** So the protective part might be the relationship that makes a teenager want to talk, more than the tracking.

**Kiffer:** That's how they read it. Their study looked at adjustment in general, in Sweden, so I'd use it as a reason for caution in reading the BC finding. And there's a third explanation. A household under strain, with a parent working evening shifts, say, might have less parental knowledge and other risks as well, such as less time together and more stress.

**Sarah:** Which also shows a factor crossing levels. What parents know sits at the relationship level, and evening shifts come from the labour market.

**Kiffer:** Right, a factor can be measured at one level and produced at another. For prevention, I think that turns the question toward what helps young people and parents keep talking, and HSCI 402 takes up which prevention programs have been tested.

**Sarah:** Question three. Why are the harms of substance use spread so much more unequally than the use itself? The alcohol harm paradox is the example I keep coming back to. People with low socioeconomic status experience more alcohol-attributable harm than people with high socioeconomic status, from similar or lower amounts of drinking.

**Kiffer:** And Probst and colleagues' 2020 review found some evidence, though scarce, that the same drinking carries more risk for people with fewer resources. So part of the answer is that the same use carries different risk depending on the conditions around it: housing, money, who is nearby and what help can be reached.

**Sarah:** Then there's the Downtown Eastside. The BC Coroners Service's death review panel reported in 2023 that the death rate there was ten times the provincial average, yet about eighty-five percent of deaths happened outside the neighbourhood. I want to see how both of those can be true.

**Kiffer:** Let's build a hypothetical province of one million people. Ten thousand of them, one percent, live in one neighbourhood. Outside it, the death rate is thirty per one hundred thousand people a year. Inside, it's ten times that, three hundred per one hundred thousand. What share of the province's deaths happen in the neighbourhood? Take a few seconds.

*(Pause)*

**Sarah:** Outside, thirty per hundred thousand across nine hundred ninety thousand people is two hundred ninety-seven deaths. Inside, three hundred per hundred thousand across ten thousand people is thirty deaths. That's three hundred twenty-seven in all, and thirty out of three hundred twenty-seven is about nine percent.

**Kiffer:** So the rate is ten times higher, and more than ninety percent of deaths happen elsewhere. A high rate in a small population adds up to a small share of the total. Now the harder part. Suppose you could bring the neighbourhood's rate all the way down to thirty, or lower the rate everywhere else by ten percent, from thirty to twenty-seven. Which change prevents more deaths?

**Sarah:** That feels obvious. The neighbourhood's rate falls by two hundred seventy per hundred thousand, and the rest of the province falls by three. Two hundred seventy is ninety times three, so fixing the neighbourhood wins easily.

**Kiffer:** Those are drops in rates, though. How many people does each of those drops apply to?

**Sarah:** Oh, right. A drop of two hundred seventy per hundred thousand among ten thousand people is twenty-seven deaths. A drop of three per hundred thousand among nine hundred ninety thousand people is 29.7, so about thirty. The small change spread across the whole province prevents slightly more deaths.

**Kiffer:** That's a version of what the epidemiologist Geoffrey Rose called the prevention paradox. A measure that brings large benefits to a population can offer little to each person who takes part, and many people at small risk can give rise to more cases than a few people at high risk. Here, each person outside the neighbourhood has their risk cut by ten percent, and each resident of the neighbourhood by ninety.

**Sarah:** So which would you choose?

**Kiffer:** I'd read the arithmetic as an argument for doing both. A response that reaches only the neighbourhood misses nine deaths in ten, and a response spread thinly across the province leaves the people at highest risk still facing a rate about ten times everyone else's. The death review panel's real figures have the same shape, which is why the Downtown Eastside needs a response of its own and the rest of the province needs one too.

**Sarah:** Question four. According to the First Nations Health Authority, First Nations people in British Columbia died from toxic drug poisoning at 5.4 times the rate of other BC residents in 2025. Why is it wrong to describe First Nations people as a high-risk group?

**Kiffer:** Because a rate ratio describes how deaths are distributed, and it says nothing about why. Calling a people a risk group puts the cause in who they are. The causes are conditions created by colonial policies and racism: residential schools, which more than one hundred fifty thousand First Nations, Inuit and Métis children attended, the removal of children from their families, and discrimination in health care, along with poverty and the toxic supply. The 5.4 shows how large the gap is, and those conditions explain why it exists. HSCI 214 develops colonialism, intergenerational trauma and adverse childhood experiences in depth.

**Sarah:** And the removal of children has continued into the present. In the 2023 BC survey, nineteen percent of First Nations youth had experience of government care or alternatives to care, compared with four percent of other youth. Racism in care has been measured directly, too. In Plain Sight, the independent review led by Mary Ellen Turpel-Lafond in 2020, found that twenty-six percent of Indigenous respondents to its survey said they were always treated as if they were drunk, or asked about substance use, when they sought health care, compared with five percent of non-Indigenous respondents.

**Kiffer:** And submissions to the review described needs going unassessed or untreated because symptoms were put down to drinking.

**Sarah:** So a person who expects that kind of treatment may put off care or avoid it altogether.

**Kiffer:** That's one pathway from racism to harm, and HSCI 214 takes up stigma theory in depth. Keep one limit of the FNHA figures in mind too. They include only Status First Nations people and their status-eligible descendants, so they leave out First Nations people without status and other Indigenous peoples in the province.

**Sarah:** There's one more figure from the report I want to read carefully. First Nations females died at 9.1 times the rate of other female residents, and First Nations males at 4.5 times the rate of other male residents.

**Kiffer:** Here's a question for everyone. Does the 9.1 mean that First Nations women died at a higher rate than First Nations men? Take a few seconds.

*(Pause)*

**Sarah:** My first instinct says yes. 9.1 is about twice 4.5, so women's rate would be about twice men's. But the report also gives the counts, and they don't fit. One hundred eighty-six First Nations men died in 2025, and one hundred three women.

**Kiffer:** Right. If men and women make up roughly equal parts of the population, those counts mean First Nations men died at a higher rate than First Nations women. The two ratios have different comparison groups. The 9.1 compares First Nations women with other women, and the 4.5 compares First Nations men with other men, so dividing one by the other doesn't give women's rate compared with men's.

**Sarah:** Then why is the female ratio so much bigger?

**Kiffer:** Look at who died among other residents. The report says 20.4 percent of other BC residents who died in 2025 were female, compared with 35.6 percent of First Nations people who died. Can you turn those into a comparison?

**Sarah:** Among First Nations people who died, it's one hundred three women to one hundred eighty-six men, so about fifty-five women for every hundred men. Among other residents, it's 20.4 to 79.6, about twenty-six women for every hundred men. Fifty-five divided by twenty-six is about 2.1. And 9.1 divided by 4.5 is about two, so the two numbers agree, once I allow for rounding and for small differences in how the two populations are made up.

**Kiffer:** That's it. The female ratio is about twice the male ratio because the balance of deaths between women and men differs. Among other BC residents, deaths are concentrated far more heavily among men, so the comparison rate for women is low, and a low comparison rate makes a ratio large.

**Sarah:** So the arithmetic tells us how the 9.1 is built. Why First Nations women's death rate is so much higher than other women's is a separate question.

**Kiffer:** Right. The reasons lie in the conditions we started with, and HSCI 402 develops overdose mortality and who dies in depth. The report also shows change. First Nations deaths fell by a third, from four hundred thirty-three in 2024 to two hundred eighty-nine in 2025, and the ratio fell from 6.5 to 5.4. So the gap is large, and it moved within a single year.

**Sarah:** Okay. Now the part where we disagree. Bruce Alexander's dislocation theory.

**Kiffer:** For me, it ties a lot of this together. Alexander argued that people need psychosocial integration, a balance between belonging and independence, and that its sustained absence, which he called dislocation, is painful and linked to addiction. Residential schools, the removal of children, the loss of housing and the isolation that criminalization produces all break ties between people, families, cultures and land. One theory connects them.

**Sarah:** And that's my worry. A theory that connects everything is hard to test, because almost any finding fits it. If connected youth use less cannabis, that fits dislocation. The same finding fits social learning, or coping motives, or the plain possibility that young people who are already doing well find it easier to stay close to adults. And the Rat Park experiments that first prompted the theory have a contested replication record, as Lesson nine showed.

**Kiffer:** Those are fair points, and I'd still defend the theory on two grounds. First, it points to causes the individual theories leave out. Cooper's motives and social learning describe a person and their friends. They don't ask why generations of children were taken from their families. Second, it makes predictions you can check. Events that cut people off, like eviction or release from custody, should be followed by more harm, and things that restore connection should be followed by less.

**Sarah:** But look at those predictions. Deaths after release from prison are explained well by lost tolerance, which needs no theory of dislocation. Using alone is deadly because no one is there to respond to an overdose. And the strengths that First Nations youth described, culture, language and community, were linked to pride and hope, measured at one point in time. I'd want a study where connection changes and harm is measured afterwards.

**Kiffer:** I agree that tolerance explains why the first weeks after release are so dangerous. Where I'd push back is on the idea that a mechanism rules out a cause. Lost tolerance explains the timing of a death. Dislocation asks why so many people were in custody, or using alone, in the first place. Those are answers to different questions, and I think a structural theory is the right kind of answer to the second one. I'll grant that it's stronger as a source of hypotheses than as a tested explanation.

**Sarah:** That's the part I can accept. As a guide to where to look, it's useful. As an explanation, I'd want it to name what it predicts and what would count as evidence that it's wrong.

**Kiffer:** Then here's a rule we can share. When someone explains substance use with a broad idea like dislocation, ask them to name the specific pathway, the evidence for that pathway, and the finding that would show they were wrong.

**Sarah:** Agreed. I'm more skeptical than Kiffer about the theory itself, but that test applies to every account in this lesson, including the ones we like.

**Kiffer:** Let's pull it together with three things to take away.

**Sarah:** First, a reason in a survey is one person's account of one occasion. Before you turn a percentage around, check which group it divides by. Saying that half of the young people who use daily give stress as a reason is a different claim from saying that most young people who give stress as a reason use daily.

**Kiffer:** Second, a rate describes risk per person, and a count describes how many people are affected. A neighbourhood can have a rate ten times higher and still hold a small share of deaths, and a small change spread across many people can prevent as many deaths as a large change among a few.

**Sarah:** And third, every rate ratio has its own comparison group, and a low comparison rate makes a ratio large. The causes of a gap lie in the conditions people live under, which for First Nations people in British Columbia include colonial policies and racism.

**Kiffer:** If you'd like more practice, rework the hypothetical province from this episode with the neighbourhood at two percent of the population and a death rate twenty times the rest, and work out both the neighbourhood's share of deaths and which change prevents more of them.

**Sarah:** Next time, it's Lesson eleven, Drug Control and the War on Drugs, where the international conventions, policing, incarceration and the iron law of prohibition come in.

**Kiffer:** Take care, everyone.

**Sarah:** See you in Lesson eleven.
