What the 2020 census data tells us about housing in Seattle

This is the second in a series of articles diving into the 2020 Census and what it reveals about Seattle. The first article looked at the impact on the upcoming City Council redistricting. This one looks at the housing data and how the numbers have changed over the past ten years. Future articles will look at changes in the racial and ethnic makeup of Seattle’s residents, and how each individual Council district has changed over the past ten years.

Let’s start with some relatively good news about housing: we built a lot over the past ten years. Between 2010 and 2020, Seattle’s population grew by 21.1%, and housing unit growth almost kept pace at 19.4%. Obviously more would be better, but compared to most American cities — and most cities in Washington State — that is a phenomenal achievement, especially when one considers that Seattle is by far the largest city in the state. In Everett, housing actually outpaced population growth; in Vancouver (the suburb of Portland, not the one up north) housing was virtually on pace. Bellevue grew its housing base by 17.1%, but still couldn’t keep up with the frightening pace of population growth the city saw (24.1%). Of the larger cities in the state, the laggard was Spokane, where population grew 9.4% but housing only increased by 3.5%.

Seattle added almost 60,000 new units over the past decade. Over half of those were concentrated in two Council districts: D3 and D7. Put another way: the big housing growth was in the center of the city, and the farther you move north and south, the less housing was built. While D7 added over 19,000 units, D5 in the north end of the city added only 3,300 — a huge disparity, and worthy of pondering the implications given that by design all seven Council districts began the decade with nearly identical sized populations (though admittedly not equal shares of housing units).

Here’s another look at the before-and-after of Seattle’s housing base:

When we compare housing growth to population growth in each of the districts, we can see clearly how differently the past ten years treated various parts of the city.

The number of housing units in District 7 grew by 34.4% between 2010 and 2020, but even that astounding rate couldn’t keep up with the 42.4% increase in population. The slowest housing growth was in D5 at 7.6%, well below the 12.8% population growth.  Surprisingly, in two districts — 2 and 3 — housing growth outpaced population growth; in a future article we’ll discuss the issues around whether that’s a good thing or a bad thing and how it ties in with gentrification and economic and cultural displacement.

In 2020 there were about 22,600 vacant units in the city, a drop from about 25,000 ten years earlier. Vacant units are slightly more evenly spread around the city than total housing units, though 46% of them are still within the combination of Districts 3 and 7.

Vacancy rates dropped across the city over the past ten years, with the exception of District 4 where they remained fairly flat (keep in mind that the census is measuring vacancies in all housing units in Seattle, not just rental units). We also can see that vacancy rates track with overall housing growth: highest in the center of the city, lowest in the north and south ends. This is not terribly surprising, for two reasons: first, the vacancy rates are no doubt related at some level to the relative pricing of housing, which is more expensive in the city center. Second, it takes time for newly-built housing units to fill up — developers suggest up to a year for a large apartment or condo building.

And yet despite all that population and housing unit growth over the past ten years, within each district the number of housing units per person has changed little — a small tick down in D5 and D7, but overall remarkably steady.

There is, however, still a significant variance across districts, with District 7 50% higher than District 2. It’s a bit of an abstract metric, though, so let’s make it real: the average household size, i.e. the number of people in each occupied unit (leaving out the vacant ones). Once again in retrospect probably not too surprising, household size is largest in D2 (home to many communities of color) and D4 (home to UW), and smallest in D3 and D7, the city center full of young urban professionals.


So what does this all mean?  A few observations:

  • The uneven housing growth across the city is, of course, tied to zoning and so-called ‘”single family” residential zones, as well as other issues such as transportation, “urban villages,” and infrastructure. There are many policy implications as we approach another major revision of the city’s Comprehensive Plan. I will leave it to others to opine at length on those topics, as we will all do over the next three years (and beyond).
  • The fact that in Districts 2 and 3 housing growth outpaced population growth — and in District 3 both were above the average rate for the entire city — is attributable to a combination of private and public housing investments that were concentrated in those areas. And yet, as we will see in later installments in this series of articles, the issues around gentrification and economic and cultural displacement are very complicated. The two districts have gone down very different paths for their respective Black communities: D3 saw a major exodus, while D2 maintained approximately the same number (though as a percentage of the total population in the district it shrank a bit). Almost certainly housing prices contributed to that differential outcome. Recently some economists have argued that the best insurance against gentrification and economic eviction is simply to build more housing; however, what we are seeing play out in Districts 2 and 3 suggests that it’s far from that simple.
  • The fact that as a whole Seattle’s housing construction nearly kept pace with the torrid pace of growth is somewhat surprising. It certainly didn’t feel that way over the past ten years, and we didn’t talk about it as if we were keeping up. To be fair, the building boom wasn’t steady and even across all ten years. Still, there’s an opportunity to do a retrospective on the past decade to figure out what we did right (and wrong) so that we can continue the trend — and perhaps even have housing growth outpace the population over the next ten years.

I hope you found this article valuable. If you did, please take a moment to make a contribution to support my ongoing work. Thanks!

7 comments

  1. Sorry to ask what may be an embarrassingly ignorant question, but when there is a gap and the population grew more than the housing what does that mean in real terms? That those people are living in the street? Or five to a room? Or—? Given the intense conversation in the city around housing, homelessness and affordability this mismatch seems like a key piece of information, but I’m not sure what it means in physical terms.

    1. It’s a good question. It doesn’t mean one thing; it was probably a combination of factors. The census counted homeless individuals. Vacancy rates went down across the city, which suggests that some of the population growth was handled by more fully utilizing the existing housing units. The number of persons per occupied unit didn’t change much, which suggests that where population growth outpaced housing growth it largely wasn’t accommodated by cramming more people into existing units.

  2. Thank you for the work you do! I always look forward to your posts.
    In reading this one I get the impression we should be building one unit per person, but my understanding is that most Seattle homes hold 1.9 on average. What is the fix?

    1. I don’t think we should be building one unit per person — that’s too much. But we do need to understand a whole panoply of policy and city planning questions around how the housing base does or does not match the population we have, and will have in the future. What is the distribution of household sizes, and how does that align with the size of housing units that we’re building? Are we building those housing units in the places where people want to live, and in the right sizes, and with the right accompanying transportation, amenities and infrastructure? And understanding that it takes 2-3 years to build a unit of housing (including the permitting/approval process), are we building housing in advance of anticipated population growth, alongside it, or lagging it? Within that, there’s also a question of what a desirable vacancy rate is that allows for enough mobility for people to move to stay close to jobs and family, while also keeping the housing market attractive enough to encourage more housing construction to keep up with future growth; in other words, we want the vacancy rate high enough for people to move when necessary or desired and to keep housing affordable, but not so high as to create a glut of vacant housing and make it financially infeasible to build more housing.

  3. I just read a report from a real estate company that showed there were over 100,000 units in various stages of planning right now in greater Seattle. This included unit now under construction plus those already permitted an awaiting a permit. Will you be writing about that in a future part of this series?

    1. Well not for this series, because that’s outside the scope of census data. And broader than just Seattle.

      Be cautious with permit data, especially those still in the pipeline. A lot of projects get abandoned before they break ground. And over the last couple of years many permit applications have been filed just before new regulations took effect in order to “grandfather” in the permit under the old regulations. Some of those are speculative and won’t ever see the light of day.

Leave a Comment