The City of Seattle released new construction data on January 15th and it shows a remarkable amount of new building. There were 8,311 housing units added in Seattle in 2014. Of those 7,538 were multi-family or mixed-use buildings. This means that over 90% of new housing units are mixed-use or multi-family. Single-family permitting accounted for the third most units, 679 but only about 8% of the total. The data also made it clear that Detached Accessory Dwelling Units and other Accessory Dwelling Units (DADUs and ADUs) are not popular, accounting for about 1% of all new units. This change is remarkable in a city that is 65% single family zoning.

During 2014 the city also permitted 760 demolitions. This means that the total additional housing added in 2014 is really only 7,551 units.


Putting the Numbers in Context

Most importantly though, we need to measure whether or not the city is successfully adding enough new housing. Over the last ten years the city averaged about 4,287 new units constructed per year. Initial planning in 2004 projected a need to accommodate 47,000 new households between 2004 and 2024. If 2010 average household size persists (which is unlikely since renters tend to have smaller household sizes and most new housing is rental housing), this translates to 4,841 new residents per year.

It turns out though that the most recent population growth numbers show the city of Seattle grew by 17,770 people in just one year. It is also obvious that the number of built units will exceed what the city projected. In ten years, the city produced 91% of twenty year projections.

Sustainable Housing Production

Seattle is exceeding expectations for both population growth and housing construction. Unfortunately, the latter isn’t keeping up with the former. The city appears to have a shortfall in additional housing units. To accommodate 17,770 people the city needed to build between 8,626 and 9,710 new housing units but only managed 7,551, missing demand by 10-22%.

The city is also undergoing a comprehensive plan update and will amend the projection for additional construction. Unfortunately, the most recent recommended update only adds 23,000 units to the projection, even though growth nearly doubled original expectations. A conservative estimate* would require 33,133 new units, 44% more than the proposed amendments.  A high estimate** might require as many as 113,385 units.

There are a number of problems with Seattle’s housing market, including low vacancy rates at the bottom of the market, homelessness, displacement, boom and bust market cycles and housing quality. Solving those problems will require various solutions but it will definitely involve building enough housing for new residents. The city already has a shortfall in housing production and will need to produce more units in the future to make up for this shortfall. The city must be more realistic about growth.

*a conservative growth rate of 1% with a conservative household size of 2.06 

**2.8% (the 2013-204 rate) and a smaller household size of 1.83 would require

We hope you loved this article. If so, please consider supporting our work. The Urbanist is a nonprofit that depends on donations from readers like you.


  1. Let’s see, there are about six times as many single family houses being built than ADU or DADU (combined). There are over seven times as many apartment units (including those in multi-use buildings) than houses. This is obviously not a free market, and the ADU/DADU market is being artificially constrained.

    Some of the development in the city is obviously fed by wealthy people. There are a lot of new houses being built, and a fair number being torn down. Those are basically people building bigger houses. Some of the houses are being replaced by apartments as well. It is impossible to tell which is which, but what strikes me as interesting is that almost as many apartment units were demolished as old houses. Since a substantial amount of the demolished houses were replaced with bigger houses, it stands to reason that more apartment units, overall, were demolished to make room for other apartments than houses making room for apartments. This makes no sense, financially. Why would you tear down a small apartment to build a bigger apartment, when you can tear down a house to make room for an apartment? The answer is you can’t, legally, in most of Seattle.

    All of this pushes up the cost of rent. Build in the handful of places where they allow you to build, and costs will be sky high. It doesn’t make sense to tear down a big house, let alone a small apartment and then build a bigger apartment unless rents are really high (and expected to be high). It makes way more sense to just add on a little apartment to a house, or build a little backyard bungalow. But, again, you can’t. The law is so restrictive that only a handful of folks bother. The end result is very high rents.

    The solution isn’t complicated. We can simply liberalize our ADU and DADU laws. The result would be cheap housing (much cheaper than most of what is being built) while preserving the existing single family homes.

    • Sightline has pretty good coverage of this issue in case you haven’t seen it:

      One complication to keep in mind when it comes to liberalizing construction in SF zones is that the property values will almost perfectly mimic the potential earnings that can be achieved from the property. This means that upzones often simply result in higher land values, adding to the cost of development. We hope to have an article before long showing how this works.

      As for DADUs and ADUs, I’m not completely sure if there’s been research on how those affect land values.

      • Thanks — yes I’ve read those articles. I think they are great.

        >> This means that upzones often simply result in higher land values, adding to the cost of development.

        OK, but then the value of existing areas goes down by the same amount. Why wouldn’t it? Put it this way, if you simply zoned all of King County as low rise, then existing low rise zones would go down in value. Of course they would. There is nothing special about them.

        If all of the lots in a newly upzoned area suddenly go up in value to match their potential as a tear down, then it means that our zoning is ridiculously restrictive. It means that the restrictions have artificially constrained the value of property so much that a mansion, or a two story building is worth the same as a parking lot. That can only happen if demand is sky high, and supply is (artificially) held really low. Such is the case in Seattle right now.

        >> As for DADUs and ADUs, I’m not completely sure if there’s been research on how those affect land values.

        I personally don’t care. I would assume that land values would go up a bit. If you are allowed to make more out of your property, then the property might be worth more. But I’m not sure if you would see much of a spike (and again, if you did, you would see other property values go down). I can remodel my kitchen, and get more money from my house, but until I actually do the remodel, I won’t see the value of my property go up. Like a kitchen remodel, if the change was widespread, then the value of the property is based on the overall value of the current property, and less on the potential. There are plenty of tear downs in Seattle, but they tend to happen with smaller places (except in areas where there are great views). In other words, the property is not worth more as a lot than as a house (and lot).

        This is how it should be, and how some of the market actually works. In a lot of areas, only the cheap buildings are replaced. You don’t see two story buildings being replaced by four story ones (it is just isn’t worth the cost of construction). This is how ADU/DADU liberalization would work. In many cases, folks would add an apartment or a bungalow. But in other cases, if they see that rent is low, then it wouldn’t be worth the cost of construction. Of course, another constraint is that a lot of people own their own house, so they would only bother with increasing density on their own property if it is also worth the bother.

        Speaking of which, I actually have neighbors, and can see the city’s restrictive policies in action. The place next to me has two houses (one main one and one tiny one). Both are really nice and would grab a lot more money than the asking price if sold separately. Another neighbor across the street lives in a place that has a basement apartment. He has no problem renting out the unit, and was actually interested in buying the two houses. He said that he wanted it as an investment. I assume that means that he would rent out both places. I don’t think he realizes that he can’t. No one can. You have to live in one, and rent out the other. This is why the houses sit and sit. A rental company would just snatch them up, rent them both, and make good money (as would the other neighbor). But this restriction alone really puts a crimp in building ADUs and DADUs. Given the guys experience with the two houses, I’m not sure that building the house even covers the cost of construction. It is pretty cheap to add a new house on a bare lot (100 grand maybe) but if that only adds a hundred grand to the overall value (and in his case, that is about it by my estimation) then it really doesn’t pay for itself. There are just too few people who want to be a landlord, and live next to the tenant.

        • I agree with almost everything you say. Demand is high, zoning is restrictive and supply is low. But I think you over simplify when you draw a straight line between development costs and rents. Land isn’t a traditional commodity in which any piece of land is interchangeable with another. For this reason, many places could be upzoned and the expected rents would simply raise property values. In other words, development costs would increase and there would be no discernible affect on rents, at least not initially.

          But overall, yes I agree, upzoning is critical and reducing the cost of development is important.

          • Oh, I agree. Land and land use is not very fluid at all. That is why there are plenty of parking lots sitting on very valuable land. The landlords aren’t waiting for the zoning to change (and the land is worth a lot more with a big building on it) they just haven’t got around to selling it.

            I also agree that some land is more valuable than others, which is why some areas that allow big buildings see a lot less development than others. But I stick with my original assessment, nonetheless. If we are talking about an area where property values of all land immediately jump to the cost of a tear down, then property must be extremely valuable. It is probably valuable right next to it as well. The old land suddenly becomes a lot less valuable if the new land is upzoned. That is just supply and demand. But of course, a lot depends on the area that is rezoned. If that area is next to the freeway, or doesn’t have a view, or is otherwise less attractive, it will have less of an effect on the other property.

      • Relaxing the regulations around mother-in-law units and backyard cottages, especially the off-street parking requirement, is not an upzone. It would apply equally to all SF zones and would therefore not affect land values. It would not change the character of the SF neighborhoods.

        • I agree, it wouldn’t change the character of the neighborhoods. In many cases it would help preserve it.

          As to your other point, I probably wasn’t clear. Changing ADU laws would change the value of higher zoned lands, by putting negative pressure on rents. Right now, if I own a house in a low rise or mid rise neighborhood, I’m thinking about tearing it down and putting up an apartment. It will cost a lot to build an apartment, and the value of the house will be thrown away, but rent is really high right now. So before I mow down the old house, I’ll have to double check the rent in Seattle. If ADU laws are changed, and more of them get built, then rent goes down, and suddenly I’m thinking about converting the house to an apartment, not tearing it down. I would get less income overall, but capital costs are lower, and with rent lower, that is the wise thing to do.

          Likewise, the same thing would happen in reverse for single family homes (especially in areas that are popular with renters). For example, a house in Fremont would sell for a lot more if it could be converted more easily to an apartment. In this case, it would probably be more of a neighborhood by neighborhood thing (some neighborhoods are more popular with renters than others).

          Changing the ADU laws is our best bet for both preserving the existing houses and reducing the cost of rent. There are other things that can be done (to reduce rent) but most of them involve construction that tends to be more dramatic and destructive (involve the removal of houses). For example, we could get rid of FAR limits on mid rise (or high rise) buildings, which would enable putting a lot more units on a lot. Likewise we could do away with parking and unit limits, so that the buildings that are built can hold a lot more people. But those other steps would do less to preserve or retain the value of the existing property. They would still do something, because they would put downward pressure on rent, which would in turn lead to fewer lots being developed. In other words, if you allowed one block to house more people, then fewer blocks would be changed.

  2. Great report, Owen. I was surprised as many new SFR units there were until I realize they’re probably just rebuilding on the same lots, which is interesting.

  3. On your chart, it’s not clear if the numbers for demo permits refer to the number of permits (likely) or the number of units demo’d (less likely). Could you verify this? It’s a big difference. Also, nothing in the article refers to the destruction or creation of affordable housing, defined as 80% of AMI or less (Mayor’s definition) or 60% of AMI or less (affordable housing advocates). The market only builds market-rate housing (!). As you point out, it’s 10% short of keeping up, but the Amazon employment bubble will burst. Meanwhile, our biggest public policy problem is displacement–unless you think it’s OK for the people who serve us and clean for us, our seniors, people with disabilities and minorities can’t afford to live here. How about a guest column on policies that offer a more equitable solution?

    • Hi Sarajane,

      The numbers above are for actual construction and demolition, not permitting.

      As for the affect on affordable units, the city doesn’t track this so there is no way to know the rents of demoed units. Our piece on the linkage fee discusses some of the points you make but we would always be happy to get guest contributions.


      • I don’t doubt that the buildings were actually torn down, which requires a permit. Is the chart showing the number of buildings demolished or the number of units? I suspect it’s the number of buildings, because it’s easier to track.

        • It’s supposed to be the number of units. The city has the unit counts for each building so if it knew the building being demolished it wouldn’t be hard to track this.

Comments are closed.