Ending Long-Term Homelessness

written by naehblog
November 2, 2012

Alliance and CEO Nan Roman delivered the following remarks at the Common Ground Alliance/Council of Homeless Persons Pre-conference Convening on in Melbourne, Australia on Wednesday, September 5, 2012. (For our American readers, in Australia permanent supportive housing is referred to as "common ground housing.")

Ending Long Term Homelessness:  Taking Permanent Supportive Housing to Scale

Thank you to the Common Ground Alliance and Felicity Reynolds and the Council of Homeless Persons and Jenny Smith for holding this forum and for inviting me to join you today. When I first came to Australia in 2007, permanent supportive housing and Common Ground housing were quite a new idea, at least by those names. Certainly there were people like Bryan Lippman at Wintringham here in Melbourne who were already doing it – combining housing and services. But it wasn’t the norm; it was not a well-established intervention. Now, as heard this morning, there are quite a few models. That is good, as a variety of models are needed.

But the next step, as the title of this session indicates, is going to scale – moving from the occasional permanent supportive housing program to an adequate supply of permanent supportive housing to meet the need and end homelessness among people with chronic disabilities.

We have not gone to scale on permanent supportive housing in the U.S., but we have certainly ramped up. There are currently nearly 200,000 units of permanent supportive housing in operation. So we have some experiences to share – some good and some bad – that may be relevant here in Australia. Let me start with a few definitions and a little history of permanent supportive housing and why it became such a big thing in the U.S.

In the U.S. you know that we define homeless as people living on the streets, etc. or in programs – not people doubled up or in poor accommodation. When homelessness first emerged in the 1980s, for a long time we thought that everyone who was homeless was a single adult male with a mental health or substance abuse disorder. This seemed an overwhelming problem with no clear solution.

Eventually we discovered, through data and research, that actually most people – 80 percent – who became homeless were homeless briefly and only once. There was a minority (the remaining 20 percent) who had the more serious problems of the stereotype – mental health and substance abuse – and who stayed homeless for long periods of time. This latter became known as the chronically homeless population. When Dr. Dennis Culhane did this typology in the mid-1990s, the size of the chronically homeless population was estimated to be about 250,000. So this new research changed the dynamic and the problem seemed more manageable.

In the earlier part of the twentieth century in the U.S., there was an abundance of what we call single room occupancy housing or hotels (SROs): over 2 million units by the 1960s. These units – though not the optimum as they were very small with shared facilities – nevertheless housed a tremendous number of very poor, often disabled single men, allowing them to pay by the night or week and providing a roof over their heads. Much of this SRO housing was lost between the 1960s and the 1980s, and it’s almost all gone today; which some think can, in and of itself, explain much of the growth of homelessness among single men. But by the 1980s some urban pioneers in New York, Oregon, and elsewhere began to rehabilitate these decrepit SROs into much more decent housing and enrich them with services for the often disabled tenants living in the neighborhoods where they remained. This was the precursor to permanent supportive housing.

Fairly quickly a few national leaders, like the Alliance, the fledgling Corporation for Supportive Housing, and a high ranking housing official in the Bush Republican Administration, saw the potential in this. Clearly it made more sense to provide these vulnerable people with housing than with shelter. A long story follows, but the bottom line is that over time, research showed that this PSH ended homelessness for even the most disabled people.

Thus, by the mid-1990s we had a problem of solvable size AND we had a solution.

I am going to get to the scaling up in a minute, but first I want to reflect just a moment on the different models of permanent supportive housing we have. We heard here today about a number of really great models, and we have many models in the States.

There is the conventional single site, dedicated model in which all the units in a building are devoted to permanent supportive housing. Such buildings tend to have modest studio apartments, often with a kitchenette and/or bathroom (although sometimes in the older buildings the facilities are shared). Some have a fair amount of common area, but others – especially new construction  have very little because of cost. Sometimes fairly intensive services are on-site, but other times the services are really just case management, which links tenants to other services in the community.

There is what you call the Common Ground model in which some of the units are for permanent supportive housing and others are market rate or affordable for other members of the community. Again, the units tend to be small and simple, and the services vary.

Increasingly there is a scattered site model in which people rent units in apartment buildings and the services are mobile or provided off-site. There has been discussion today of the Pathways model. This employs the Assertive Community Treatment or ACT team method of providing services. There are other models that primarily use the intensive case management design to link consumers to other service providers, while ensuring that they get the services they need to reside in the community.

These models are all largely for people with serious mental illness, substance abuse disorders, physical disabilities, age-related issues, or a mixture of these. The story of permanent supportive housing is hugely positive and it is solving vexing problems and generating enormous bipartisan political will nationally and in virtually every city in America. This is a huge success story that is influencing how other social problems are handled.

How did we go from a program model with a scattered set of practitioners to nearly 200,000 units that have significantly reduced chronic homelessness and are targeting the hardest to serve people in literally every community in the country? I think that the recipe for scaling up, for us, is not that complicated, and it is one that we have replicated for other homelessness problems.

The first step was being able to size the problem, and having a problem that is of solvable scale. We had been speaking in the U.S. about a homeless population of millions, which felt too big to solve. But research revealed a problem of a solvable scale – 250,000 people who were chronically homeless, very ill, and clearly should not be on the street or in shelter.

The second step was having a solution to the problem.  A fair amount of research had been done on permanent supportive housing, and much more has been done now. It is clear that it works. Generally speaking a minimum of 80 percent of people with disabilities can be successfully housed in permanent supportive housing, and frequently that percentage is much higher.

The third part of the equation was having a solution that is affordable. Cost studies had shown that it costs the public sector at least as much to leave someone with serious mental illness on the street as it costs to house them with services. Subsequent cost studies on people with HIV/AIDS and chronic street inebriates have found significant cost savings from housing high cost people. By shifting money that was being spent on not solving the problem to solving the problem, we could get there. Of course, that is more easily said than done as a practical matter.

These three things together  a problem of solvable scale, for which we had a solution that was effective and cost effective  provided us with a framework for solving the problem. If there are 200,000 chronically homeless people whose homelessness can be ended with permanent supportive housing that costs less than what we do now, by providing 20,000 more units per year for the next ten years, we can solve the problem in ten years. So this was less about program models and more about the overall framework of solving the problem.

This framework allowed policy makers and practitioners to see a way forward. By going to scale, they could solve a problem, not just endlessly address it. And they could do it cost-effectively.

I want to point out that these three factors all involve something very important, and that is data and research. Without data and research: we would not have known the number of chronically homeless people; we would not have been able to verify the effectiveness of the solution; and we would not have been able to make the cost argument.

One other essential ingredient in getting this going was advocacy. I can promise you that this research, once it emerged, did not just automatically cause policymakers to say, “Let’s create 200,000 units of permanent supportive housing and end chronic homelessness!” It was a few key national organizations with strong credentials and good capacity that moved the agenda forward through advocacy.

Once the scaled problem and a cost-effective solution were in place, what were the steps in scaling this up? There was no single thing that happened, but a series of things.

On the political will side, there were commitments made by the Administration and by Congress to end chronic homelessness. In both cases, these commitments were expressed in budget terms, so there were some dollars behind it.  That was the beginning.  Over time, at the federal level:

  • Congress targeted 30 percent of the nation’s largest homeless assistance grant program to permanent supportive housing. This essentially required that every community in the nation begin to provide permanent supportive housing, and created a steady stream of funding over time.
  • The Department of Veterans Affairs (VA) started funding thousands of units of permanent supportive housing for homeless military veterans; a large group in the U.S. The federal housing department provides rent vouchers for scattered site housing and VA provides the services.
  • The goal setting continued, with the current Administration vowing to end chronic and veteran homelessness by 2015.

In this way, federal leadership and a pipeline of resources were established.

States and localities also began to take up the challenge:

  • Some State housing finance agencies – public entities that fund affordable housing – began to require that a percentage of multi-family housing units they financed be permanent supportive housing, or to target a percentage of funds to permanent supportive housing each year.
  • New York State and New York City created the New York/New York agreements to provide thousands of units of permanent supportive housing.
  • Plans to end homelessness, which exist in nearly every major city in the nation, almost all contain commitments to end chronic homelessness.

And the private sector has gotten on board:

  • There is a consortium of foundations dedicated to getting their peers to fund permanent supportive housing.
  • The new Social Impact Bond movement in the U.S. is focused on ending chronic homelessness through permanent supportive housing because of the cost savings.

We also had to increase our capacity to provide permanent supportive housing. Many homeless organizations were initially disinclined to engage in the strategy: they had no will or expertise to develop housing or provide sophisticated services. Some felt that their mission lay in helping people who were homeless – not in ending people’s homelessness. But the federal leadership, expressed through money, pushed the development of will and capacity among nonprofit providers. National intermediary NGOs emerged to help with financing, technical assistance, and training. Eventually the capacity developed.

All of these elements are geared to increasing the scale of the intervention. Certainly individual organizations still have to have the mission and the gumption to do the deals and put up the housing, and in the U.S. as here that is a gargantuan effort. But increasingly the systems are being created that will provide and sustain enough of this housing to meet all the needs.

All of this is not without challenges, some of which I have identified. You might ask, for example, if we had 200,000 chronically homeless people and we created 200,000 units of permanent supportive housing, shouldn’t the problem be solved? And yet there are 107,000 chronically homeless people remaining. Why?

The main challenge has been that the permanent supportive housing units do not all go to chronically homeless people. Some of this is policy, in that the agencies do not require that every unit go to chronically homeless people. Some of it is the providers, who may not take the high-need tenants, either because they just don’t feel they have the service supports for the most needy people; because they really don’t want to take them; or because performance benchmarks disincentivize that.

Some of these targeting issues are being handled by changes in policy, and also by things such as the vulnerability index, and the 100,000 Homes Campaign, which I know have been here in Australia. These campaigns – sort of like the Ten Year Plan campaign of early 2000s  get people geared up to identify and house the hardest to serve.

There are other issues as well. There are mismatches between where the units are and where the people who need them are. Funds are given out by formula and go to every community. But the majority of the chronically homeless population is in the big cities like New York and Los Angeles. So we don’t have the units where we need them.

And a significant lesson we’re learning is the importance of different models. Not every one of us wants to live in the same type of housing with the same type of people. I have noticed that where there is a proliferation of one type – say single site – there is more failure because if that model doesn’t work for someone, there is no option. Let a thousand flowers bloom and don’t get too caught up in the purity of any model. Different models means there is more choice for consumers and if you want to end homelessness, you will need someplace that works for every single person.

At the end of the day, we have not ended chronic homelessness in the U.S. nor solved all of the problems related to providing permanent supportive housing. But we have cut chronic homelessness significantly. And some communities have essentially ended it.

Since we started a little earlier than you, there might be some things to learn from us. I would summarize them as follows.

  • Data is critical to building the case and then measuring progress and adjusting as you go along.
  • Having a framework for how you’re going to end the problem works.
  • Setting numerical goals is important.
  • You need strong, talented advocacy.
  • You need technical assistance and capacity building.
  • You need sustained funding.
  • Targeting is critical – otherwise you won’t reduce your numbers.
  • Use a lot of models, but stay focused on the population you want to address.
  • Try to get the money moving in this direction – if every project is a one-off it will never scale up.  You need a pipeline.

I appreciate and admire the terrific work that is going on here in Australia. I commend the efforts of the Common Ground Alliance to bring you all together. And I very much I look forward to the discussion about how you could scale up in Australia.  Thank you.