I have just returned from Canada, after spending a week as a guest of the University of Ottawa. I was contributing to a new initiative exploring innovation in housing and homelessness. Alongside these events, I met with municipal leaders, federal officials, researchers, health practitioners, community organisations and frontline homelessness services. 

One of the things that stayed with me was a short walk through Ottawa’s city centre. Each day, on my way between meetings, I would pass many of the same people on the streets. One woman in particular seemed to occupy roughly the same spot throughout the week. I know nothing about her circumstances and have no idea what support she may or may not have been receiving. Yet after a week spent discussing innovation, prevention, evidence and systems change, I found myself returning to the same question: how do we ensure that all of these conversations ultimately translate into different realities for people like her?

That question followed me throughout my visit. The contexts of the UK and Canada were different, but the underlying questions felt remarkably familiar. Canada is considerably more decentralised than England, with responsibilities shared across municipal, provincial and federal governments. The geography is different. The housing system is different. The indigenous peoples mean that the distribution of risk is different. The scale of its drug crisis has created challenges that are difficult to fully appreciate from afar. 

Yet beneath these differences sat many of the same tensions that those of us working on homelessness encounter elsewhere. How do we respond to visible and immediate pressures while investing in long-term change? How do we create the conditions for prevention? How do we ensure that knowledge translates into action? And how do we build systems that learn?

One of my first conversations was with two senior city leaders. They spoke about Ottawa’s commitment to ending youth homelessness by 2030 and to moving away from large congregate shelters towards smaller supportive housing developments embedded across the city. Doing so requires not only investment but also political leadership: navigating understandable local concerns, aligning action across different levels of government and maintaining focus on longer-term outcomes. The discussion reminded me of conversations I have had with leaders across the UK. Everyone agrees that preventing homelessness is preferable to responding to it. Yet creating the conditions for prevention requires sustained political commitment, coordination across multiple agencies and a willingness to invest in outcomes that may not become visible for years.

A day later, the conversation looked very different but sounded surprisingly similar. Federal officials reflected on the role national government can play in helping connect research, policy and practice. Like many countries, Canada possesses world-class data, universities, researchers and practitioners. The challenge is not a shortage of knowledge. It is how knowledge travels. How promising ideas spread. How evidence informs decisions. And how learning accumulates across a system rather than remaining isolated within individual organisations. 

In between those conversations, I visited Shepherds of Good Hope, one of Ottawa’s largest homelessness and supportive housing organisations. What struck me most was the contrast between the realities of the present and the ambitions for the future. Outside the shelter, people gathered during the middle of the day, some visibly struggling with the effects of addiction and poor health. It was impossible not to reflect on the devastating impact of the toxic drug crisis and the immense challenges facing those working on the frontline. At the same time, inside the organisation, leaders spoke with pride about reducing average lengths of stay from around a month to just 11 days through diversion efforts designed to help people access alternatives more quickly. 

That combination of realism and ambition stayed with me. It would be easy to look at scenes of visible homelessness and conclude that systems are failing. The reality is more complicated. Across the week, I met dedicated people working incredibly hard to improve outcomes. Yet many conversations returned to the same challenge: how do we translate what we already know into consistent action at scale? .

In many ways, this is the central challenge facing homelessness policy today. Over the past decade, homelessness has benefited from a remarkable expansion in evidence and data. The Centre for Homelessness Impact’s Evidence and Gap Maps show that between 2018 and 2024 the number of effectiveness studies identified globally increased from 227 to around 800, while implementation studies grew from 212 to approximately 840. Canada’s contribution has been particularly significant, with effectiveness studies increasing from 34 to 144 and implementation studies from 46 to 164. Compared with even 10 years ago, we know substantially more about the causes of homelessness, the effectiveness of different interventions and the conditions that support successful implementation.

Yet throughout the week I found myself reflecting on a different observation. One participant referenced the frequently cited estimate that it can take around 17 years for research evidence to become embedded in routine practice. Across many fields, there remains a substantial gap between what we know and what we routinely do.

Healthcare has wrestled with this challenge for decades. Some of its most important advances have involved not only discovering effective treatments but developing the methods, institutions and cultures required to spread and sustain improvement. The emergence of quality improvement, implementation science and learning health systems reflected an important insight: knowledge alone does not change outcomes. Systems need ways of absorbing, testing, adapting and applying what they learn.

Increasingly, homelessness appears to be confronting the same challenge. Several of my discussions in Ottawa touched on the Centre for Homelessness Impact’s Test and Learn approach and the advantages of testing promising interventions across multiple areas simultaneously. What interested me, however, was not simply the methodology. It was the ambition behind it. The objective is not merely to generate better evidence. It is to help different communities learn faster than the problems they are trying to solve.

Perhaps nowhere is this more important than in prevention. Compared with 20 years ago, we understand far more about the warning signs that often precede homelessness. We know that interventions delivered earlier are frequently more effective and less costly than those delivered after a crisis has occurred. We know that poverty, housing insecurity, poor mental health, institutional transitions and relationship breakdown frequently create predictable routes into homelessness. Yet despite this knowledge, much of our collective effort remains concentrated downstream.

This is understandable. Crises demand attention. Visible need creates moral urgency. Political leaders, service providers and communities cannot simply ignore people who are desperate for help today. The challenge is that responding to today’s crisis can easily consume the time, money and attention required to prevent tomorrow’s.

What struck me most in Ottawa was not the differences between municipal leaders, federal officials, researchers and practitioners. It was the degree to which they were increasingly arriving at similar conclusions. The Mayor of Ottawa, Mark Sutcliffe’s ambition to end youth homelessness, federal efforts to strengthen connections between evidence and practice, and Shepherds’ determination to reduce reliance on traditional shelter models all pointed in the same direction. Different responsibilities. Different levers. Different time horizons. Yet all were attempting to answer the same question: how do we create systems that become better at preventing homelessness rather than simply managing it?

The past decade has seen extraordinary progress in our understanding of homelessness. The next challenge may be ensuring that our institutions evolve at the same pace as our knowledge. If there was one lesson I took from Ottawa, it is that lasting progress is unlikely to come from choosing between emergency responses and prevention, or between evidence generation and implementation. The communities that make the greatest progress in the years ahead are unlikely to be those that generate the most evidence alone. They are likely to be those that become best at learning from it, acting on it and adapting in response to it.

• Ligia Teixeira is Chief Executive of the Centre for Homelessness Impact

‍