# Lesson 12: Disability, Diversity, and Integrating the Foundations

*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. This episode goes with Lesson twelve, Disability, Diversity, and Integrating the Foundations, which is the last lesson of the course.

**Sarah:** It's meant for after you've worked through the lesson. We'll add some perspective on how these ideas get used in practice, some critique of the frameworks and the numbers, and we'll work through the questions students tend to find thorny with this material. We'll also do some extra worked examples, including a few that are harder than the ones on the lesson page.

**Kiffer:** There are a few places in the episode where we'll ask you to work something out before we give the answer. When that happens, you'll hear a few seconds of quiet. Pause the audio if you'd like more time, and then come back to hear how we worked through it.

**Sarah:** We have four questions today. First, does the social model of disability mean that impairment doesn't matter? Second, why do careful surveys disagree so much about how many people have a disability? Third, when disabled people have worse health, how much of that gap comes from the impairment itself? And fourth, when a disability rights law is passed, how would we know whether it worked? Along the way we'll come back to the three habits of thought that the course has tried to build.

**Kiffer:** Question one. Does the social model mean impairment doesn't matter? Sarah, I'd like to start with a made-up example and the ICF, because the framework makes the question concrete. Imagine a hypothetical man named Marcus, who is forty-one and worked as a warehouse supervisor. After a COVID infection he developed long COVID. He tires after small amounts of effort, and he has trouble holding information in mind.

**Sarah:** So let's sort his situation into the ICF components. His fatigue and memory problems sit under body functions. His trouble lifting boxes or concentrating through a long task sits under activities. And the fact that he's stopped working and dropped out of his recreational hockey league sits under participation.

**Kiffer:** Good. Now one more detail. Marcus asked his employer for shorter shifts with rest breaks, and the employer refused. Where does that refusal go in the ICF, and which model of disability would point to it as the main problem? Take a few seconds.

*(Pause)*

**Sarah:** The refusal is an environmental factor. The ICF counts the attitudes of other people and the policies of workplaces and services as part of the environment. And it's the social model that would point to it, because the barrier sits in the workplace, outside Marcus's body.

**Kiffer:** That's right. A medical-model response would focus on his fatigue and memory, through rehabilitation and treatment where it helps. A social-model response would focus on the refused shifts. The ICF lets us see both on one page. There's also a fifth component, personal factors, such as his age and the way he tends to push through tiredness. The ICF names personal factors as part of the picture, but it doesn't give them classification codes, because they vary so much across societies and cultures.

**Sarah:** So back to the question. I'll take the side of the strong social model, because I think it deserves a fair hearing. Michael Oliver set out the social model in a 1983 book, building on British disabled activists from the mid-1970s. Its power was political. It said that the problem to fix is the staircase, the hiring practice, the inaccessible clinic. If you start conceding that impairment is part of the problem, I worry it slides back into fixing the person.

**Kiffer:** I'd take Tom Shakespeare's side here. In 2001, he and Nicholas Watson asked whether the social model had become an outdated ideology. Their argument is that impairment and barriers both cause disability. For Marcus, a perfect employer would help a great deal, and he would still be exhausted after climbing stairs. Leaving that out makes it harder to argue for research and care on long COVID itself.

**Sarah:** But I'd point out that Oliver himself wrote that he never claimed the social model was an all-encompassing framework. He described it as a way of putting one simple idea to work in everyday practice. If you treat it as a practical tool, the criticism loses some of its force.

**Kiffer:** That's a fair point, and it narrows our disagreement. I think the strong version works well as a political tool and less well as a full description of why people are disabled. In my view, conditions with pain or fatigue that doesn't go away, like long COVID or chronic pain, are where the full description matters most.

**Sarah:** So here's a rule we can share. When you describe a disability problem, name the impairment effects and the barrier effects separately, and ask the people living with the condition which barriers matter most to them.

**Kiffer:** Agreed. That rule also follows the principle of nothing about us without us, which disability rights activists around the world took up in the 1990s.

**Sarah:** Question two. Why do careful surveys disagree so much about how many people have a disability? The lesson gives about fifteen percent globally and about twenty-two percent for Canada in 2017. Those can't both describe the same thing.

**Kiffer:** They describe different operational definitions, which is the first habit of the course in action. Let's build it with made-up numbers. Imagine a hypothetical survey of one thousand adults that uses the Washington Group short set. That's six questions, about difficulty with seeing, hearing, walking or climbing steps, remembering or concentrating, self-care, and communicating. Each question has four answers: no difficulty, some difficulty, a lot of difficulty, or cannot do it at all.

**Sarah:** And the Washington Group's recommended cut-off counts someone as having a disability if they answer a lot of difficulty, or cannot do it at all, on at least one of the six questions.

**Kiffer:** Right. Here are the made-up counts of people at that level in each domain. Walking, fifty-two. Seeing, twenty-four. Hearing, twenty-eight. Remembering, thirty-three. Self-care, fifteen. Communicating, ten. What share of the adults in this survey have a disability by the recommended cut-off? Take a few seconds.

*(Pause)*

**Sarah:** Let me try it. I'll add the six domains. Fifty-two plus twenty-four is seventy-six, plus twenty-eight is a hundred and four, plus thirty-three is a hundred and thirty-seven, plus fifteen is a hundred and fifty-two, plus ten is a hundred and sixty-two. So a hundred and sixty-two out of one thousand, which is 16.2 percent.

**Kiffer:** Before we accept that, here's a sanity check. Suppose one person has a lot of difficulty walking and also a lot of difficulty with self-care. How many times did your sum count that person?

**Sarah:** Twice. Oh, I added up difficulties when the question asked about people. Anyone with difficulty in two or three domains was counted two or three times, so a hundred and sixty-two can only be an upper limit.

**Kiffer:** Exactly. And the lower limit is fifty-two, because the walking group alone contains fifty-two different people. To get the real answer, you need to count each person once.

**Sarah:** So suppose the made-up survey shows that a hundred and ten people had a lot of difficulty in at least one domain. That's a hundred and ten out of one thousand, or eleven percent. And it sits between the limits, which is reassuring.

**Kiffer:** Now watch what the cut-off does. Suppose that in the same made-up survey, three hundred and forty people reported at least some difficulty in one or more domains. With that looser cut-off, the prevalence is thirty-four percent. Same people, same questions, and the estimate triples.

**Sarah:** And the Canadian Survey on Disability uses a different approach again. Its screening questions ask about difficulty and also about how often a condition limits daily activities, and they cover pain, learning, mental health and developmental disabilities, which the six short-set questions don't ask about directly.

**Kiffer:** Which helps explain why the Canadian survey finds a higher figure. In 2017 it found that twenty-two percent of Canadians aged fifteen and older had a disability, about 6.2 million people. In 2022 the figure was twenty-seven percent, or 8.0 million people.

**Sarah:** And the fifteen percent figure from the World Report on Disability, where did that come from?

**Kiffer:** It came mainly from two other sources, the World Health Survey and the Global Burden of Disease study, which gave adult estimates of about sixteen and nineteen percent. The report also explained that it was higher than the older estimate of about ten percent from the 1970s, because populations had aged, chronic disease had become more common, and measurement had improved.

**Sarah:** So the jump from twenty-two to twenty-seven percent in Canada is worth thinking about too. Statistics Canada linked part of it to growth in the senior population and part of it to more mental health-related disability, especially among youth and working-age adults.

**Kiffer:** That's a good example of reading a trend carefully. The first question is whether the definition changed. The second is whether the population changed. Only after that do you ask whether the underlying health of people changed.

**Sarah:** Question three. Disabled people have higher rates of nearly every preventable chronic condition and more unmet health care needs. When students see that gap, a common first thought is that it simply reflects the impairment. How much of it does?

**Kiffer:** Some of it does. Someone with a spinal cord injury has a higher risk of pressure injuries and urinary infections because of the injury itself. But a large part of the gap is in outcomes the impairment shouldn't affect much, like cancer screening, dental care and blood pressure control. That's where barriers, poverty and ableism come in.

**Sarah:** Let me set a short exercise, using a made-up example. A hypothetical community clinic finds that its patients who use wheelchairs are much less likely to be up to date on cervical cancer screening. Name one possible explanation at each of the three levels of ableism from the lesson: structural, institutional and interpersonal. Take a few seconds.

*(Pause)*

**Kiffer:** Here's one set of answers. Structural: people who use wheelchairs in this community are more likely to live on low incomes and to depend on accessible transit that's unreliable, so getting to appointments is harder. Institutional: the clinic has no height-adjustable exam table and books every appointment for the same short slot. Interpersonal: a clinician assumes the patient isn't sexually active and doesn't offer the test.

**Sarah:** And each of those has a different fix. The structural one needs transit and income policy. The institutional one needs equipment and booking rules. The interpersonal one needs training. None of them involves changing anything about the patient's body.

**Kiffer:** Diagnostic overshadowing is a closely related idea. The term comes from research by Steven Reiss and colleagues in 1982, who studied how a diagnosis of intellectual disability led clinicians to overlook signs of mental illness. Today it's used more broadly, for any new symptom that gets put down to an existing disability without being investigated.

**Sarah:** The lesson mentions Brian Sinclair as a Canadian example. He was a forty-five-year-old Indigenous man and a double amputee who used a wheelchair. In September 2008 he died in the emergency department of the Health Sciences Centre in Winnipeg after waiting thirty-four hours, from a bladder infection that could have been treated. The inquest report in 2014 found his death was preventable.

**Kiffer:** And it's worth being careful about what the case shows. Reports said that staff assumed he was intoxicated, or that he was homeless and sheltering from the cold. Those assumptions draw on stereotypes about Indigenous people and about disability together. The case is a reminder that these forms of bias combine, which is why it's hard to assign the harm to any single one.

**Sarah:** Question four. When a disability rights law is passed, how would we know whether it worked? The history in the lesson keeps showing a gap between the law and what happens next.

**Kiffer:** It does. Section 504 of the Rehabilitation Act passed in 1973, and the regulations sat unsigned for about four years until activists occupied the federal building in San Francisco in April 1977. In Ontario, the Accessibility for Ontarians with Disabilities Act set a goal of full accessibility by 2025. A review by former lieutenant governor David Onley, published in January 2019, found that progress had been painfully slow.

**Sarah:** So we need a way to estimate what a law changed. The lesson mentions difference-in-differences. Let me build the harder example, which uses made-up numbers. A hypothetical province passes an accessibility law that covers employment. In a survey before the law, forty-five percent of disabled adults are employed, compared with seventy-eight percent of non-disabled adults. In a survey a few years after the law, the figures are forty-three percent and seventy-nine percent.

**Kiffer:** So here's the question for everyone. What's the difference-in-differences estimate, and what would it seem to say about the law? Take a few seconds.

*(Pause)*

**Sarah:** For disabled adults, employment went from forty-five to forty-three percent, a change of minus two points. For non-disabled adults, it went from seventy-eight to seventy-nine percent, a change of plus one point. The difference-in-differences is minus two, minus plus one, which is minus three points. So the law seems to have lowered the employment of disabled adults by about three percentage points compared with what we'd otherwise expect.

**Kiffer:** That's the right arithmetic. Now here's the trap. Suppose the before survey identified disability with questions about difficulty in daily activities. And the after survey switched to a single question: does a long-term condition limit the kind or amount of work you can do?

**Sarah:** Then the disabled group in the after survey is partly defined by having trouble working. People whose condition is well accommodated at work might say no and drop out of the group. So the employment rate in that group would fall even if the law did nothing.

**Kiffer:** Exactly. The two groups called disabled aren't the same group. Suppose the after survey also kept the old questions, and with those, disabled employment went from forty-five to forty-seven percent.

**Sarah:** Then the change for disabled adults is plus two, the change for non-disabled adults is still plus one, and the difference-in-differences is plus one point. So the sign flips from minus three to plus one, just from holding the definition constant.

**Kiffer:** And this happened in the real debate over the Americans with Disabilities Act. Two influential early studies, published in 2000 and 2001, found that employment of disabled people fell after the law. In 2003, the economists Douglas Kruse and Lisa Schur showed that the decline appeared with a measure based on reported work disability, and that measures based on functional limitations showed increases in employment for some groups. They concluded that the findings depended heavily on how disability was defined.

**Sarah:** So a difference-in-differences is only as good as two things. The comparison group has to show what would have happened without the law, and the definition of the affected group has to stay the same.

**Kiffer:** And I'd add that the debate over disability and rights in Canada is still moving, and it's worth following. Medical assistance in dying for people whose only condition is a mental illness was delayed by Parliament in 2024 to March 2027. Then in late October 2026, the federal government introduced a bill that would cancel that expansion, citing concerns from medical experts and a parliamentary committee. Disability rights organizations have been among the strongest voices in that debate.

**Sarah:** Which connects to the critique in the lesson. When housing, income and disability supports are inadequate, many disability advocates argue that a request to die can't be treated as a free choice.

**Kiffer:** Let's pull it together. Each of today's questions used one of the three habits the course has tried to build, and the takeaways follow them.

**Sarah:** First, the operational habit. A disability count depends on the questions and the cut-off, so before you compare eleven percent with twenty-seven percent, ask which instrument and which definition produced each number.

**Kiffer:** Second, the historical habit. The social model, the ICF and the disability rights laws each came from a particular moment and a particular fight, and knowing that history helps you see what each one was built to do and what it leaves out.

**Sarah:** And third, the structural habit. When disabled people have worse health, separate the effect of the impairment from the effects of barriers, poverty and ableism at each level, because those are the parts that public health can change.

**Kiffer:** If you'd like more practice, rework the difference-in-differences problem from this episode with your own made-up numbers, and then ask what would happen if the definition of the disabled group changed between the two surveys.

**Sarah:** That brings us to the end of the course. Lesson twelve is the last lesson, so this is the last episode of Office Hours for HSCI 130, and the courses that follow will build on everything we've talked about.

**Kiffer:** Take care, everyone.

**Sarah:** Thanks for spending the course with us, and good luck in the courses ahead.
