We can take a better approach than what was put forth in the failed “Compassion Seattle” City Charter Amendment.

City Charter Amendment 29, posed by “Compassion Seattle” is dead, at least formally. Bruce Harrell seems intent on screaming “long live the Charter Amendment!” by making it into mayoral policy. In fact, his promise to sweep encampments and punish those who won’t go into shelters suggests for some Seattle “centrists,” CA29 was never about housing or compassion at all, but about shoving homeless Seattleites out of sight so we don’t have to be confronted by the ugliness of inequality in our city. 

Still, CA29 was quite popular, a poll suggesting 61% of Seattle residents in favor. But in favor of what, exactly? The text was sprawling and ambiguous, with very little agreement on what the amendment would actually accomplish. For many, the promise of quickly building 2,000 units of housing, plus requirements restricting park removals to situations where real housing is on offer, was attractive. But in the end, the text contained too many trojan horses and the Harrell campaign looked likely to use too many of them as a pretext for sweeping parks.

Harrell wears a blazer and gestures toward tents in the background.
Bruce Harrell speaks near a homeless encampment on Broadview-Thomson school grounds. (Photo courtesy of KING 5)

With CA29 dead, but the mayoral race heating up, it’s worth asking: what would a set of policies that are genuinely compassionate and effective look like going forward? Here are nine steps.

  1. Get real buy-in from homelessness experts. Early on, the campaign backers claimed that many human services organizations, like DESC, Plymouth Housing, Fare Start, the Public Defender Association and Evergreen Treatment services endorsed the measure. This turned out to be false. Instead, many are a part of the Seattle/King County Coalition on Homelessness, which sued to stop CA29 from going on the ballot. Organizations with expertise and people with lived experience need to be at the center of these conversations.
  2. Build affordable housing that is actual housing, fast. CA29 required 2,000 units to get built in a year. This sounded great, but then it included a trojan horse that allowed a bed in a big open room to count as housing. We know congregate shelter is often unhealthier, less hospitable, and more dangerous to our unhoused neighbors than are tents in parks. This is not compassionate; it’s just a cheaper way to empty the parks. This was a deal-killer. These should have been permanent supportive housing units for the chronically homeless.
  3. Build enough housing for all people chronically experiencing homelessness. King County’s point-in-time count suggests at least 3,355 people suffer from chronic homelessness. This is our highest-needs population, and they need housing now.
  4. Expand to housing for the episodically homeless. Many people who are homeless need help for shorter or more medium length periods of time. When you consider the 4,000 people living in shelters, or our other 3,600 rough sleeping neighbors in King County, Seattle needs to shoulder housing for at least 5,000 of them. That is to say nothing of the tens of thousands of units we need to end the housing affordability crisis.
  5. Fund behavioral health support for all at-risk populations. CA29 required “low-barrier, rapid-access, mental health and substance use disorder treatment and services . . . [for the] chronically homeless” with a “behavioral health rapid-response field capacity” that worked with “non-law enforcement crisis response systems,” using “culturally distinct approaches.” This was to be available to everyone living in housing for the otherwise homeless. It should also include the recently homeless or others who are at high risk of losing their homes as well. (We should have fully funded social medicine with complete behavioral support. But I will limit my scope to homelessness for now.) We should also be clear not to insinuate that homelessness is primarily about mental health or substance abuse, a favorite conservative canard. An overwhelming supermajority of unhoused people are mentally fit, just as a lopsided majority of folks with behavioral or substance abuse problems have housing. That being said, this population does suffer from somewhat higher rates of severe behavioral problems and mental illnesses, in part because homelessness exposes them to much more trauma. More at-risk people should be able to receive this support.
  6. Tie eligibility for future contracts to performance. CA29 also mandated that the city publish quarterly information regarding the “effectiveness of strategies and services designed to transition homeless individuals to housing,” and inform the public “which city services, activities, and practices may contribute to people entering or experiencing homelessness.” While some say sunshine is the best disinfectant, I say the city already measures the impact of the providers who serve the homeless and it doesn’t work. Some of the providers that flunk just go straight to the city council and get written directly into the budget instead. Data dashboards don’t necessarily prevent cronyism or incompetence. Tying funding to performance would work much better. Organizations should be given contracts only if they are high-performing. Performance should be measured by the ratio of dollars spent to months housed, with a minimum magnitude for average months housed. This should be measured separately by risk tier, since chronically unsheltered folks are much more expensive to house. Racial groups and other marginalized groups should also be measured separately. Efficiency should be maximized for interventions within the groups, not across the groups, to prevent performance data from driving biased funding away from marginalized citizens.
  7. Reduce the cost and time to build housing for the homeless. CA29 required the City to waive land use codes, regulation and permitting fees, move projects serving the homeless to the front of the permitting line, and refund various city-related costs, fees and sales taxes. This was welcome, but it could have been better. These changes should be permanent, to not only help us get out of this emergency, but prevent future ones, and should include all subsidized housing. We must also waive the design review process for all affordable housing, as well as all parking lot coverage, and rear and side setback requirements. Finally, we need to increase heights for affordable housing to 85 feet in multifamily and mixed-use neighborhoods, and 45 feet in lower density neighborhoods, with increased floor-to-area ratios to match. At minimum, this should apply in all high opportunity and high frequency transit neighborhoods, and to community land trusts in remaining neighborhoods.
  8. Fund it using progressive revenue. CA29 was an unfunded mandate. We cannot house over ten thousand homeless people, nor prevent floods of future Seattleites from losing shelter, without building many thousands of units of housing. Real units are expensive. Simple math and chronic inaction show that nothing in our current budget could come close to the cost of building them. It doesn’t do us any good to pretend. McKinsey, not exactly the most progressive organization on earth, says King County needs to spend between $450 million and $1.1 billion a year in incremental public funding on subsidized housing just to stabilize housing for rent burdened, extremely low income people. 
  9. Legislation on camp removals should offer clear and limited criteria for any kind of removals. Sweeps without housing on offer are cruel and unusual punishment, according to the Ninth Circuit court in Idaho. [The Supreme Court declined to hear the case, allowing the decision to stand.] That holding does, however, allow people to be swept into shelter. But Seattle can do better than the bare minimum required by the conservative Court. Given what we know about beds in congregate shelters, we can comfortably say, they aren’t housing. Shelters offer no privacy, very little safety, and extremely little autonomy. They often require sobriety, which is completely out of step with current housing-first best practice. They may force people to throw away their things, separate from partners or family members, or get rid of their pets. And shelter may expose them to assault or trauma. This is not a compassionate alternative, and it is not housing. Forcing people into these congregate shelters should easily meet our definition of cruel and unusual. We cannot do so and respect the dignity of humans living outside our doors. 

Legislation on removals should offer clear and limited criteria for any kind of removals. CA29 did have some good language requiring individualized interventions that paid attention to culture, family structure, and disability. It also acknowledged the possibility of harm from encampment closures. It should explicitly give the aggrieved simple administrative grounds to challenge the manner in which removals take place, to have harm redressed, and it should put transparency and accountability mechanisms in place.

It should also limit when removals can take place. Real housing, not fake congregate housing, should be on offer. Otherwise, removals should be limited to public health risks, real threats to public safety, or blocked access to major public goods, not counting the land itself that is being occupied.

While CA29 is dead, these policy questions live on. Hopefully our region can come together, this time around something genuinely compassionate and genuinely effective.

Correction: An earlier version of this article incorrectly stated the Supreme Court affirmed the Ninth Court’s Martin v Boise ruling. By declining to hear the case, the Supreme Court let the lower court ruling stand, which is legally distinct from affirming that ruling.

We hope you loved this article. If so, please consider subscribing or donating. The Urbanist is a 501(c)(4) nonprofit that depends on donations from readers like you.

Ron Davis (Guest Contributor)

Ron Davis is an entrepreneur that has spent most of his professional life working to improve the lives of workers and seniors. He's a former member of the Citizen Oversight Panel for Sound Transit and is active in trying to make Seattle a more just, inclusive, clean, walkable, city. He has a JD from Harvard Law School and lives in Northeast Seattle with his wife, a family physician, and their two boys.

Inline Feedbacks
View all comments
A Joy

The minority of King County’s homeless have behavioral health issues. All the data collected backs this up. As long as behavioral health support is artificially tied to homelessness solutions we will be wasting money as opposed to actually solving the problem. Do we need more money spent on behavioral health issues? Absolutely. Does that have anything to do with the average homeless person’s plight? Absolutely not.



Interesting article (although pre-Covid, 2019) about primary causes of homelessness in Seattle. Alcohol or drugs was the second most commonly cited cause, as identified by homeless individuals themselves. Mental health accounted for 8.8%, although I would venture that individuals with severe mental illness would be significantly less likely to fill out a voluntary survey and thus are likely underrepresented.

Ron Swanson

“homelessness is [not] primarily about mental health or substance abuse”

Tell me you don’t go outside without telling me you don’t go outside…

CJ Fenn

I agree. My biggest frustration is when this issue is just framed as an affordability crisis, which is a big part of the problem. This article suggests that substance abuse is not party of the primary problem. I coached my daughters little league team and had to sweep dugouts prior to games to clean up needles. Which I found more often than not. Try to imagine the rage of a father trying to coach kids and there is someone yelling at the girls during practice. Until the progressive side can incorporate in their 9 part plan how they will address the fact that drugs are legitimate part of the problem, they will continue to misunderstand why CA29 got traction.


Homelessness leads to needles in dugouts. What does that say about the solution? Actually very little. If the implication of needles in the dugout were that “we just need to get those people treatment,” then why are most homeless people not experiencing mental illness? (Only 36% of chronically homeless people have a mental and/or substance use disorder).

The fact is you look at what’s changed. Availability of treatment hasn’t changed. Substance use disorders, although more deadly now, haven’t really increased. But rent has doubled in the past 20 years.

A lot of people on one side of this issue can say, it’s because society has become too tolerant. But how are you measuring that? Are you saying police behavior has changed? Because I don’t think any person who becomes homeless is thinking, “well, gee. Cops aren’t as mean anymore. I’m going to move out and live in a tent.” It’s the rent!

The surprise is to think about that fact that most people who have substance use disorders leave their needles and bottles in the comfort of their home. Why is there 2.1 million in the us with opioid use disorder, but only 550,000 homeless? Why 15 million with alcohol use disorder, but only that number homeless? Because they have enough to get by. But when rent increases by half in the past 10 years, doubles in the past 20, or triples in the past 30, now they don’t have enough to get by and you get more people who are on the streets – doing whatever they would use to do in their house or apartment.

So why not figure out how to build both *temporary supportive* and *permanent* housing to address the needles in the dugout?

Edit: or improve temporary supportive housing. Person below has a good point about storage units.

Last edited 2 months ago by Andrew

Drug use is often correlated with homelessness. Sometimes one leads to the other, sometimes the other way around.

I don’t think anyone denies this. This is in fact a big part of Point 5 in the article above. If we have more public funding to help people with issues such as addiction and mental illness, perhaps fewer people struggling with these things will become homeless in the first place.


Your words exemplify the triumph of anecdote over statistics. They are the perfect showcase the silliness of saliency bias. This is a hard, complicated issue – emotion cannot be taken out of it. But let’s try to let reason rule the day as much as we can please.

I go outside every day, take my small kids into Seattle parks most day, including those with many homeless people. But when I think about cause/effect and homeless people, I try to pay attention to actual representative data, not the biased interactions of my daily walk to coffee, run through the woods, or to the playground.

Beside the fact that you probably have no idea what percentage of people are using needles in the camps, the people in parks represent a fairly small fraction of the total number of homeless people. So even if every single person in a park were using a needle every day, your observation would be completely irrelevant in light of the information I presented above.

Ron Swanson

The plural of anecdotes is data. I’ve been observing the situation long enough to get a representative sample of the homeless population.

The majority of homeless individuals killing people’s dogs, throwing bricks off of overpasses, throwing hot coffee at children, molesting female pedestrians, etc., have severe mental health or substance abuse problems.

Denying this fact by claiming they’re a small percentage of the homeless population is simply obfuscation. They’re the majority of the highly visible, highly disruptive campers metastasizing all over the city.

Better housing options for those who are ready to accept rules, jail for those who aren’t. It’s that simple. The polling is showing that’s likely to be the majority position – Harrell is leading for a reason.

Bryan K

The majority of homeless individuals killing people’s dogs, throwing bricks off of overpasses, throwing hot coffee at children, molesting female pedestrians, etc., have severe mental health or substance abuse problems.”

But that’s not actually about homelessness, it’s about people with problems. Nobody wakes up and says “hey my shelter’s made of wood, no need to kill a dog today” versus “oh my shelter’s made of canvas, need to kill a dog.”

Ron Swanson

Yes, that’s addressed by the later part of my comment – massive increase needed for better housing for those whose main problem is “does not have house,” consequences needed for those whose main problem is drugs and mental illness.

More capacity is needed for treatment for both, but the community needs to stop tolerating antisocial behavior from those who won’t accept it.

Bryan K

But we don’t need consequences for “those whose main problem is drugs and mental illness” if they aren’t doing things that you describe – we instead need consequences for “people whose main problem is doing bad things” – like killing dogs or throwing coffee at children – regardless of why they’re doing it (including simply being an otherwise healthy bad person).

A Joy

The plural of anecdote is not data. That is the crux of the issue here. Subjective reports in quantity are never a replacement for objective statistics.


Yes, we’ve got someone who is deeply confused about the nature of data. Many biased interactions do not make a randomized, non selection biased sample.

We have more than ten thousand homeless people in King County, and only 800-1200 tents (in Seattle, where the majority concentrate). Just the visibly homeless make up a small fraction of the homeless population.

And the percentage of folks in tents who are committing the violent kinds of crimes described above that make it into the papers is extremely small. It’s the unusual stuff that makes it into the papers, because that’s what gets clicks. Apparently the power of distortion is strong enough to convince people its statistically representative.


I think there should be least some opportunities to make temporary emergency housing at least a little bit less awful without too much cost. One obvious target that comes to mind is lack of storage. There needs to be both on-site storage lockers for people to keep clothes and stuff, rather than carrying it around all day and off-site storage for bulkier items, such as furniture, which is too big for a shelter, but likely to eventually be useful when they find more permanent housing. Compared to housing humans, storage space for stuff is cheap.

Another thing that should be looked at is the feasibility of allowing residents to get at least some basic degree of privacy. I’m thinking along the lines of foldable partitions between different family units, or even simply allowing people to pitch a tent on the shelter floor. It doesn’t have to be fancy, just something that allows people to have some basic personal space that’s out of sight of others.

There also needs to be some solution for people with pets, at least those that don’t cause problems to others. Forcing people already on the edge mentally to give up their beloved dog or cat is not helping things.


“Sweeps without housing on offer are cruel and unusual punishment, according to the Ninth Circuit, a holding affirmed by the Supreme Court. That holding does, however, allow people to be swept into shelter. But Seattle can do better than the bare minimum required by the conservative Court.”

Wait, what? Are you saying the 9th Circuit is conservative? Or SCOTUS? Or both?


This is poor reportage. The Supreme Court never took the appeal from the Ninth Circuit decision. That decision, Martin v Boise, is the most progressive decision on the issue. It’s also quite acrobatic.


Thanks, that clarification makes sense. I’m a bit surprised a Harvard JD doesn’t understand that declining to take an appeal is not the same as affirming.


This is how our legal system works. The Supreme Court doesn’t make a habit of explicitly affirming lower court decisions they agree with, but their agreement can be inferred from the fact that fewer than four justices wanted to allow an appeal to proceed.


This is not at all how SCT jurisprudence works. It is not a court of error correction, unlike the circuits. You shouldn’t infer anything from a cert denial, except that 5 justices don’t find it weighty enough to hear. (Could you point to the 5 justices who “agreed” with this decision, while keeping a straight face?). You can bet that the ACLU is not pushing this issue in the other circuits, though, because a circuit split might tip the balance for the supremes. And I’m pretty sure the ACLU would prefer this particular Court not weigh in on this issue.


You are right, there is a difference. But at the time I drafted this, the distinction between affirming and letting it stand didn’t strike me as functionally relevant for the the policy article, especially when turns of phrase like “let it stand by declining to take up the case” were the obvious alternative. While such ugly writing is fine when writing to lawyers, I try to avoid it elsewhere! =)

Still, “affirm” is a legal term of art, and so I should have found another word to avoid confusion.

I apologize for the lack of clarity!


No worries – just wonking out. It does mean, though, that you’re urging the city to be more progressive than the 9th, not the SCT.


10 — Change the zoning. The main reason Seattle has a big homelessness problem is because rent has skyrocketed. The main reason that rent has skyrocketed is because jobs have increased, while housing has not. The main reason housing has not is because of our zoning laws. Change the zoning laws and more people will be able to find a place to live.


Yes, the combination of our economic success and our poor supply of housing have made costs run out of control.