Homeless needs assessment paints a detailed picture of the city’s homeless population

Late last summer the Mayor promised that the Human Services Department was working on an assessment of the needs of Seattle’s homeless population. Originally due out in November, it was finally released today.

The report gives the most detailed view to-date of who our homeless neighbors are, how they became homeless, and the issues they are struggling with. Along the way, it debunks several persistent myths about the homeless population and suggests the services that the city could provide that would do the most to lift them out of homelessness.

First, a preface on how the survey was done. It was conducted by Applied Survey Research, a company that has done similar surveys in other cities including San Francisco; ASR also  conducted the “one night count” here in Seattle this year. In addition to their own staff, they used “peer survey workers” who had themselves experienced homelessness, both to increase their ability to find a wide variety of homeless people and to increase the trust among the homeless people being surveyed so that they would be willing to fill out the survey. They varied geographic locations to try to capture data from across the city, and they surveyed people from unsanctioned encampments, sanctioned encampments, various shelters, and other locations. In all, they collected 1,050 survey responses — a huge number for a survey such as this.

They also held a series of focus groups involving a total of 80 homeless people from various sub-communities (youth, families with children, people living in vehicles, etc) to explore in depth some of the issues and trends they were seeing in the survey results.

That said, this is not a statistically rigorous survey: you won’t find a margin of error or a p-value anywhere in it. Mainly that’s because we don’t actually know how large the homeless population is in Seattle; without knowing what portion of the total you sampled, you can’t estimate the error. They also didn’t fully randomize the sampling method (though they did try to mix in a bit of random sampling). Nevertheless, with a sample size this large the quality of the information is high. The 2016 one-night count found almost 3000 homeless people in the city, which is certainly an undercount; but if in November of last year the true number were even as high as 10,000, then a 10% sample would likely produce reliable results. So it’s not a perfect tool, but it’s a very good one. HSD also points out that the survey results match the demographics in two other data sets: the one night count, and the HMIS system that Seattle and King County use to track, assess and serve homeless people. That adds further validity to these survey results.  It’s worth stating this up-front because the results directly contradict several widely-held beliefs about homeless people and will no doubt be challenged by some as “fake news.” This is not fake news; it’s a good piece of work done by a reputable organization that provides useful insights that help us make decisions about how to address homelessness in Seattle.

Let’s dive in to the data, beginning with the myths that the survey debunks.

  1. The majority of Seattle’s homeless population are “locals.”

70% were living in King County at the time they became homeless, with another 11% in Pierce, Snohomish or Thurston County. Only 15% came here from outside of Washington state. It’s a persistent myth in most large American cities that the homeless are coming from elsewhere.

2. Most homeless people didn’t come to Seattle to access homeless services.

It’s a myth (also common in many cities) that they are moving to the city to access its homeless services. The graph below takes a bit of explanation: since many of our homeless here in Seattle are natives, only about 700 out of 1050 respondents answered the question “why did you come to Seattle?” Of those 700, only 15% (about 100 people) said that they came here to access homeless services. That’s about 10% of the total respondents. Also keep in mind when you look at these figures that even though they asked for the primary reason, some people gave more than one reason, so the items don’t add to 100%. Homelessness is complicated. But the point remains: anyone who tells you that there is a mass influx of people to Seattle because of our services doesn’t know what they’re talking about.


3. There is no single dominant cause of homelessness.

When asked what the primary event was that led to homelessness, the people surveyed answered all over the map. Again, homelessness is complicated. For many, it represents a “cascade failure” where more than one thing goes wrong. If you’re living in poverty and have no savings, losing your job — even if you can find another one within a month — is a traumatic event. Similarly, if you’re living in poverty and get divorced, or you’re living paycheck to paycheck and get evicted (for whatever reason), you don’t have the money to pay the move-in fees for a new apartment. This means that the solutions we put in place to help people need to be diverse and customized.  We should also note that “alcohol or drug use” is only 12.9% of the survey responses, about half the rate of “lost job.”


4. The vast majority of homeless people don’t want to live outside.

93% of those surveyed would choose to move into safe, affordable housing if it were offered to them.


5. Alcohol and/or substance abuse among the homeless is high but not even close to universal.

About 45% said that they don’t use drugs or alcohol. That’s probably under-reporting their substance abuse, just as the rest of us also under-report drug and alcohol use, but it’s still nowhere near 100%. When broken-out by type of substance, some of the numbers are surprisingly low, and especially heroin use is much lower than the popular perception:

One more point on these numbers: we saw earlier that only 12.9% said that alcohol or drug use was the cause of their becoming homeless. Even if we allow that there is another group who were substance abusers before they became homeless, even if it wasn’t the primary cause of their homelessness, that’s still far less than the 55% of the homeless population who currently use drugs or alcohol.  The clear take-away: homelessness leads to substance abuse far more than substance abuse leads to homelessness.


6. The homeless aren’t uneducated.

81% of those surveyed had a high school diploma or greater.

That’s good news; it means that if we can help them with their other issues, they are employable.


7.  The homeless have good reasons for not using existing shelters.

Lots of them, in fact.

And this survey doesn’t even mention the other top reason: that the vast majority are only overnight shelters that kick people back out on the street at 6am.


Beyond the myth-busting, there are several other important insights to be gained from the survey. Several  are simple demographic statistics to help us understand the phenomenon better. For instance, people of color (except Asians) and members of the LGBTQ community are over-represented in the homeless community:

So are survivors of domestic violence: about 42% overall, 51% of those under 25, 58% of women, and 63% of transgender persons.

Health issues, and especially a variety of mental health issues, are widespread. Here, as with substance abuse, the percentage of homeless people with mental health issues — particularly depression — is far higher than the rate of people who said that mental health issues caused their homelessness (7.9%). We can draw the same conclusion: homelessness drives mental health issues far more than the reverse.

Also, nearly 50% of respondents reported going without needed dental care.

We can also learn much about how people experience homelessness. For example, 50% had been homeless for a year or more.

30% were “chronically homeless,” meaning that they had been homeless for over a year and also had a disabling condition. That’s double the national average for people experiencing homelessness, though similar to other West Coast cities.

Not surprisingly, there is a higher prevalence of drug use among the unsheltered compared to the sheltered (let’s look at that graph again).

Three more relevant stats:

  • Only 1/3 of those surveyed had been assessed and enrolled into the Coordinated Entry system that King County uses to track and serve homeless people.
  • 41% of respondents said that they had experienced being “swept” from an unsanctioned encampment.
  • 11% of female survey respondents (about 40 women) reported that they were pregnant at the time of the survey.

But here’s the jaw-dropper: 23% of respondents — and 40% of those under the age of 25 — had gone through the foster care system. 4% of respondents with an experience of foster care reported aging out of foster care as the primary cause of their homelessness, and 6% reported living in foster care immediately before they became homeless. According to ASR, one in four former foster youth experience homelessness within four years of exiting the foster care system. Today’s foster care system is extremely efficient at generating homeless people. Focus group participants confirmed that lack of support for youth transitioning out of the foster care system, combined with instability and abuse within the system, are major contributors to the problem.

Those surveyed had many ideas about the things that would help them get out of homelessness. As expected, those related to housing issues rated highly.

The focus groups brought out the notion that the most effective outreach is peer outreach: that the city should use outreach workers who have themselves experienced homelessness.


When we reflect on this wealth of data, we can draw up a list of things the city should prioritize in order to help address the needs of our homeless community. Some are well-known, including:

  • Safe, affordable housing;
  • Jobs (even those requiring a high school diploma);
  • Health services, especially dental and mental health;
  • Domestic violence support programs;
  • Support for LGBTQ homeless, especially LGBTQ youth.

But there are also some new items that point to holes and failures in the current human services system:

  • Reforming the foster care system, and providing support for youth transitioning out of the system;
  • A larger push to enroll homeless people in the Coordinated Entry system;
  • Support for pregnant women experiencing homelessness;
  • Building a peer outreach program.


HSD is far from done with analysis, though. Combined with the forthcoming 2017 One Night Count (aka “Count us In”) data, this needs assessment could provide a first approximation of the actual demand for human services: how many people need mental health services, how big should domestic violence support programs be,  how much affordable housing is needed to bring all the homeless indoors, etc. But the department also needs to pay attention to the other side of the equation: it needs to start generating crisp data on the capacity of the city, county and private human service programs. Until that’s done and the “gap” between need and capacity is understood, we simply don’t know how to invest wisely to bring the right new resources online.

The Needs Assessment Survey is a good, and necessary step. We learned a lot from it. But there is still much work to be done.



  1. Thank you for covering this, and for your humane, compact, and accurate analysis!

  2. I’m with David B. — solid analysis. I especially appreciate the explanation of the survey’s statistical limits.

  3. I agree with DK and DB; good reporting.

    Is the report available on line anywhere? If you linked to it, I missed it.

    1. We were briefed yesterday under an embargo of 6am this morning. The report is supposed to be released some time today, and I’ll post an update with the link when that happens.

  4. Do you still believe that HSD isn’t being held accountable and isn’t trying to gather data to help drive its decisions for how to spend people’s tax dollars better?

    1. Yes, I still believe they aren’t being held accountable. And I didn’t say they weren’t trying to gather data; I said that they were not being particularly competent at gathering and sharing data about what’s happening with the communities of people they are trying to serve, and they are especially not being transparent with data about their own operations. And I stick by all that.

      As I said in the post, this is good stuff, but it’s only one piece. They need to take this data and convert it from percentages into real numbers of people who need specific services. They also need to study and publish information on the capacity of existing providers to serve that many people and identify where gaps exist.

Comments are closed.