Margot Kushel is having the moment she never wanted to have.

Kushel is an internist at the University of California, San Francisco. She started specializing in low-income populations shortly after graduating from the Yale School of Medicine and has spent two decades researching the underlying causes and consequences of homelessness in relative anonymity.

Lately, however, she’s seen her profile rise, as the problem she has spent her career trying to solve has escalated.

Last year, Kushel was named director of the Benioff Homelessness Initiative at UCSF, which was endowed with $30 million from Marc Benioff, the billionaire founder of Salesforce. The initiative is focused on translating proven homeless solutions into widespread adoption and continuing to research what isn’t known.

Here’s the conversation, edited and condensed for length:

Q: Tell me about your career and how you ended up specializing in homelessness.

A: When I started residency, I realized that approximately half of the inpatients we cared for were homeless. We would admit patients to the hospital, give them all this very high-quality, expert medical care, and then, eventually, we would have to discharge patients back to their homelessness, meaning to outside. Patients would ask me to please not discharge them, but eventually we wouldn’t have a choice. Inevitably, a few days later, the patient would be back, often in worse shape than they had been in before. I remember thinking that there had to be a different way and decided to change my career plans.

Q: What was the state of understanding of homelessness when you first got started versus what we know now?

A: When we first started, people believed that to provide housing, people needed to go through steps. First, a shelter. Then, if they “behaved well” (didn’t use drugs, took medicines, etc.), they could get to transitional housing. If they did everything “right,” then they could be offered permanent housing. As a result, only a tiny proportion of people with behavioral disabilities became housed.

Housing First turned that upside down, recognizing that when people were homeless, they couldn’t attend to their mental health or substance use needs (or anything else). This has been enormously successful, housing about 85% of the most complex folks. There is overwhelming and incontrovertible evidence that this works — people are housed successfully, and then the other things follow.

Q: What are some of the myths around homelessness?

A: You hear people saying things like, “You can’t just house people who have addiction problems.” You can, and you must. Another is that homelessness is caused by mental health and substance use problems. We know that most homelessness is driven by economic forces. The vast majority of people who become homeless could be easily housed if there were housing that they could afford on their income. Yes, having mental health and substance use problems are risk factors. But, most people with these disabilities are housed.

Q: What don’t we know about homelessness?

A: There is a lot more work to be done in homelessness prevention. We know that for some people, a small infusion of resources (cash, services) can prevent homelessness. But, for every 100 or so people at high risk, only one will become homeless. So, we need to do a better job of figuring out how we can target efforts.

Most people who become homeless “self-resolve,” meaning they find housing. We don’t know how long that takes and whether we could shorten it substantially by intervening. We know that for most people, long-term subsidies are the answer, but there may be people who need shorter-term help. We are going to try to figure out who needs what, while working to solve the main problem, which is the shortage of extremely low-income housing.

Q: What would it take to end homelessness?

A: We’ve always known that most homelessness is a result, pure and simple, of poverty: the lack of a living wage, the lack of affordable housing and the insidious effect of racism. If we don’t fix the fundamentals, we are just patching a leaking ship. And that is what has happened.

It would take an investment in creating and sustaining extremely low-income housing and efforts to increase the minimum wage and to close the existing housing gap. Right now in California there are 22 units available and affordable for every 100 households with extremely low incomes.

Some people will have disabilities that require assistance, and we need to provide that. The VA has leaned in hard to the Housing First principle and has seen huge reductions. Taking a page from what the VA has done, we can solve chronic homelessness by fully funding permanent supportive housing.

For everyone else, we need to focus on increasing the supply of extremely low-income housing by building very low-income housing, preserving what exists and providing sufficient vouchers (right now, only a quarter of households who qualify get them).

It has been profoundly dispiriting to see, for so many years, the issues of housing affordability/living wage, and homelessness ignored on the federal level. And, it has been hard to see the disconnect in people’s mind between housing and homelessness.

This article originally appeared in The New York Times .