For Sale By Owner Sign
Very Honest For Sale By Owner Sign by Casey Serin on Flickr

In many cities where supply is constrained by various factors – geography, NIMBYism, etc.  – the price of housing tends to be expensive.  One way that cities have tried to solve for this is by increasing the potential supply of housing through zoning changes.  Naturally, this has been met with some resistance in the neighborhoods slated for upzones, as has been documented on this blog many times over.  Alternative approaches to increasing affordability, like rent control, often have unintended consequences.

What if we took a different approach to housing affordability? Let’s consider the median multiple: a metric of housing affordability that compares the median house to the median income in a given area. For example, a metro area with a median income of $50k/year and a median house price of $150k/year would have a multiple of 3, which is the historical average in the US (Seattle hit 5.2 in 2008, making it one of the least affordable cities in the country).

What if a municipality were required to issue as many housing and zoning permits as needed until the median multiple hit some number, let’s say 4.0?  Why 4.0? Well, 3.0 would be nice, but since the income data trails the housing price data somewhat, the municipality would risk overshooting the target and crashing the housing market.  The precise number isn’t important.  What’s important is that there would be a target to aim for, similar to the way the Federal Reserve buys and sells short term interest rates to hit an inflation target.

How would that look in practice? King County median income was $67,806 in 2009. Multiply that by 4 and you get $271,224.  King County’s median housing price, meanwhile, was $382,160 in September of 2009.  So there would be quite a ways to go to hit the target.  Build, baby, build!

What’s nice about this approach is that it doesn’t presuppose any particular development pattern.  The city could choose to build up OR out.  Now, I’m pretty sure that between the Growth Management Act and the geography of the region, building up would be the path of least resistance. In reality, the city still would run into resistance and likely have trouble meeting such an aggressive target.  But the momentum would change. Instead of asking “should we build?” the question becomes, “where should we build?” That’s a question the urbanists have a better answer for.

I’ll caveat this by saying that “affordability” is very complicated and one statistic or ratio will never capture all the nuances.  Transportation costs and interest rates, for example, are ignored by the median multiple.  But I’m coming around to the idea that crude metrics, even if they don’t meet all our wonkish ideals, have a certain political appeal.

48 Replies to “Rethinking Housing Affordability”

  1. The catch is that municipalities would *also* have to have enough restrictions on expanding into actual greenspace (agricultural land, parks, etc.) — otherwise they’d just sprawl out more single-family houses, and you’d end up with “affordable” housing which wasn’t affordable once you added in transportation-to-work costs.

    1. Dunno whether the GMA would actually do the trick for that.

      No, probably no problem in King County (what agricultural land is left in King County?) but it might have unintended consequences if applied further out….

      1. Actually, there are CONSIDERABLE amounts of land in East King. Look at the 203 and 202 corridors, all the land around Issaquah and N. Bend, and then even S. King. Lots of land surrounding the agriculture which *could* be developed, but shouldn’t.

      2. One Big parcel going on the market on Eastside that seems to have been under the radar:

        Bellevue-based Mutual Materials manufactured bricks on the 53-acre site, in Newcastle just off Coal Creek Parkway, for more than 50 years. The plant closed last year, a victim of “the length and depth of the recession,” said Joe Bowen, the 112-year-old company’s president.

        The property is zoned mixed-use — residential, retail, office. About 30 acres is developable, said Newcastle City Manager Rob Wyman. The rest is streams, wetlands, steep slopes and old coal-mining areas.

  2. I love this. I have the same concern as Nathanael about sprawl, but done well this won’t be an issue. Our development will be as good as our civic leadership. If they really ask the question “where should we build?”, the right answer will be in clusters with good transit.

  3. The City of Seattle has historically tied public infrastructure investments to willingness of neighborhoods to accept more housing, including more “affordable” housing (which could mean unaffordable, but limited to those with low income, but that’s another scam for another post). The City is ahead of the curve on answering the question of where to build.

    A big part of the disconnect is that only *city* infrastructure is then tied to willingness to upzone. Rainier Valley, Beacon Hill, and Pioneer Square got their regional transportation infrastructure investment without the neighborhoods buying in to upzoning. Same with the U District and Northgate, but at least there is modest upzoning in Roosevelt.

    The U District has been fighting increasing building heights to stop “encroachment” by the University of Washington. But if UW can’t build up in the U District, where can they build up? And if UW can’t build up, where will all the 18-year-olds go off to to get an affordable 4-year degree? Yep, our NIMBYism is building up Wazzou. Tragic.

    1. The U-District is an especially silly place to have restrictive zoning.
      1) It has already got a number of high-rises
      2) It has pretty good transportation infrastructure
      3) It has a ton of retail, and is already walkable
      4) Has some of the lowest rates of car ownership, and some of the highest rates of transit use.

      1. Agreed. This is a critical issue for Seattle’s future. U-District ought to be 300/240 zoning at least. Step down towards Roosevelt.

    2. When you say “The U District has been fighting increased building heights” – who exactly do you mean? I live in the U District, and would LOVE to see higher buildings (and all those hideous parking lots on Brooklyn turned into apartments). Do I have a city council member who is trying to limit growth? If so I’d like to know who it is so I can write some emails…

  4. We should now know that increasing the number of housing units doesn’t reduce housing costs. First, you’d need a very large supply of vacant units for rents to be affected–somewhere between 15 and 19%. That’s a lot of unused space. Second, we’ve already seen that (a) new construction can be geared to the second-home market, particularly in cities with an existing shortage and (b) the “boomtown effect” can actually increase the cost of housing, even as the supply increases. If you want affordable housing, you have to specifically develop affordable housing.

    1. These comments make zero sense. What is any of this based on? That 15-19% number sure sounds precise – I’d love to hear a source for that. Basic economics tells us tha any change in supply affects prices.

      At least your “boomtown effect” is an understandable mistake. You’re only looking a prices, not affordability. The only reason such an effect would happen is from increased jobs and prosperity, which increases average salaries.

    2. Three things:

      a) There are two sorts of “second homes” that I can think of. One is you are a wealthy foreigner looking to move money offshore, so you buy a house in a stable place. Think Vancouver or Melbourne high-rise condos. The second is that home prices vs rents are pretty good, so you want to buy a house to rent it out. If this is happening, then rents should actually fall, because more rental units will come online. So this is a boon for affordability. I am sure I missed others.

      b) “Boom town” effects are usually caused by other factors. Fracking becomes a big deal in North Dakota or something and lots of people move there for jobs, but housing supply lags. In this case, you build more housing to solve the problem.

      c) The best way to create very-low income housing may be to build it directly. That’s different from housing that’s affordable to the average person.

  5. How would that look in practice? King County median income was $67,806 in 2009. Multiply that by 4 and you get $271,224. King County’s median housing price, meanwhile, was $382,160 in September of 2009.

    Brilliant. Seattle’s finally starting to dig out from an economic collapse caused by a building bubble and you want to somehow create public policy that will finance overbuilding at today’s prices with the goal of deflating that wealth by more than what lead to the great recession. Loan baby loan!

    1. Was the bubble problem in Seattle that we over built or that prices went up to high? Or both?

      Over building is bad at the time, but in the long run you end up with extra houses which isn’t the worst thing in the world. It’s not like a massive boom in cupcake businesses or something. I’m not saying either is good, but one is certainly better than the other.

    2. “Overbuilding” by what metric? Did Frank propose public “financing” for this? I think you’re misunderstanding. I believe he’s suggesting reducing our limitations on construction enshrined in our building and land use codes.

      1. Frank failed to propose any financing. If he’d thought it through that far he never would have publish such a silly piece. You could eliminate all zoning and building permits and it still costs money to build houses, duh. Who’s going to invest when the stated policy of government is to drive down the value of what you just built? If you hadn’t noticed, housing starts have been at a virtual standstill because of steadily declining real estate values. They’re starting to pick up now because prices are rising. STB’s really jumped the shark with this post.

      2. Bernie – The counterpoint to your argument (and Mark Johnson’s below) is the Texas real estate market. There are no restrictions on sprawl, so builders can easily build as much as they want. The elastic supply kept prices from rising too much during the boom, and falling too much during the bust. Builders and buyers knew these homes wouldn’t appreciate much (because another builder would always come along if demand increased), but builders built because they could profit and these prices, and buyers bought because they needed a place to live.

        Frank is proposing a similar elastic housing market in Puget Sound, but with unlimited “building up” capacity where Texas used unlimited “building out” capacity.

        I do agree with you that the government will never have an official policy of reducing housing prices from $382,000 to $271,000. Existing homeowners/voters will prevent significant implementation. The most likely scenario is for affordable housing policies to hold prices steady even during boom periods or wage increases.

      3. Texas has a lot to do with a strong economy than zoning: Fuel and food drive leaders.

        Just four states — oil-producing strongholds North Dakota, Louisiana, Alaska and Texas — have returned to peak employment, riding a global energy boom. Other energy and agricultural states in the nation’s breadbasket, including Oklahoma, Nebraska and South Dakota, are within a percentage point of their peak. All these states lost relatively few or no jobs in the downturn and have benefited from a worldwide surge in food and energy prices.

        You’re right about the sprawl. Houston, a city of 2 million has a density of only 3,600/sq-mi; the same as Issaquah. And the median household income is only $42k, less than Everett or Tacoma, so that puts a damper on speculative development. Not a big profit margin when the median home price is only $123,800.

  6. @ Matt the Engineer- I agree that the wage level does need to be part of the equation, and the original post says it would be. Encouraging a more democratic distribution of business profits to all the workers who help to produce them would also take us closer to meeting this particular affordability goal.

    The bigger problem: I am not sure who you would issue building permits to if the market was soft to enough to have declining prices. The issues influcencing housing prices are cost of land, cost of materials, cost of labor, cost of credit, and cost of operation. Zoning affects cost of land, but it is difficult to control- massive upzoning in an area that has no demand has little effect, and upzoning where demand is high can push land prices higher. Building code probably affects costs of materials, but I don’t have any building safety issues that I want to see eliminated to bring housing costs down. Tax structure has some effect on cost of operation- there is already a little help in that arena for multifamily housing.

    It’s a complex problem. Good to keep trying new things, and this is a simple metric we should probably work on but we should not give up on other metrics as well.

  7. Great idea with one large exception being that all units are not equal, particularly in terms of construction cost. stacking units vertically is more expensive per SF than ground-related housing. If there is not a willingness to accept smaller units (at a higher price/sf, holding unit price point constant for this metric) than there could still be a major affordability gap despite ample zoning. This is the case in Seattle, where there is a 50 year+ supply of stacked flats but very little new land for new ground related townhouses or new single-family homes.

    1. The construction cost can be higher per square foot for taller buildings, but the overall cost can still be much cheaper when land is expensive. Take my house, a small single-family home in north Seattle, as an example. The house itself is appraised at roughly $150k, while the land it sits on is appraised at closer to $200k.

      Suppose there was an empty lot of identical size next door to my house. You could buy the land and put an identical house on it for about $350k total. Suppose that the zoning allowed you to put three units on that land instead of just one. Suppose also that the difficulty in building vertically means that you need to spend $200k per unit to get the same amount of space as the $150k single-family home.

      In this case, you’ve still spent the same $200k on the land, and $200k each to construct three apartments/townhomes/whatever. That’s a total of $800k, or only $266k per unit when the land is included. That’s a significantly lower price per unit than the $350k for a single-family home, and you can house three times as many people this way.

      However, as you state, there are few vacant lots in Seattle. If you want to build new townhomes, you’re probably going to have to knock down an existing single-family home. That means the cost of acquiring the land in the first place will include the fair-market value of the existing home. In our three-unit example, add $150-200k to the overall cost because existing house has intrinsic value and also costs money to tear down.

      Now the cost per unit has gone up closer to $333k. That’s going to be a hard sell if you can get a single-family home on a larger lot in the same neighborhood for $350k. If the zoning instead allows for 4-6 units on that same land, maybe then the math will work out. It seem that for a redevelopment to be really enticing, either the existing building can’t be worth much at all, or the new zoning has to allow so much more density that the cost of the existing building becomes mostly irrelevant.

      1. A couple of problems with your numbers. The big one is that if the land were rezoned to allow apartments it would then be worth a lot more than the current $200k. Second is the appraised value of a structure reflects it’s depreciated value. Replacement cost for new construction is many times the appraised value of an older home.

      2. “if the land were rezoned to allow apartments it would then be worth a lot more than the current $200k.” How is that a problem? High land values make this argument stronger – the higher the ratio of land values to construction cost, the more it makes sense to build up.

        “Replacement cost for new construction is many times the appraised value of an older home.” Sure, but the math still applies. If this didn’t work out in real life, you would never see houses torn down to build townhouses.

      3. Nothing “wrong” with the value of the land going up. The “wrong” is not working it into the numbers cited. Yes, when you get a rezone for higher density, be it town homes or midrise a developer has to build more densely for the project to pencil out. It also virtually assures that the cost per dwelling will be more than any single family zoning it displaced. That always happens when you tear down an old structure to build new even if the zoning doesn’t change.

      4. “It also virtually assures that the cost per dwelling will be more than any single family zoning it displaced” Not true at all, and you and I have disagreed on this point for years. Man I wish OrphanRoad was up right now. Check this link later.

      5. Yes, we had this discussion months ago. I registered a trial copy of RS Means and the numbers in the StreetsBlog piece are totally wrong. When you start to add in all of the finish costs, like stoves and toilets and don’t get caught up in the trap that at certain story breaks there are huge gaps in the software’s pricing it works out pretty close to all the real world construction costs and sale prices of recent construction in Seattle. Can you build barracks style housing in a tall building? Sure, they’re called dorm rooms. Would a developer do this to maximize profit? Probably not. As for apodments they’re actually really expensive on a square basis to rent as are hotel rooms.

      6. Maybe you don’t understand the issues. The price of granite countertops don’t matter. As you build up, your per-unit price goes down. This is because your land costs are fixed.

      7. As you build up, your per-unit price goes down. This is because your land costs are fixed.

        Scroll back up to my original reply. Land costs are not fixed. Land zoned for higher buildings is more expensive. Yes, the more units you cram into a given structure the cheaper you can sell each one but that’s not what happens when developers build something like Bellevue Towers. Really tall buildings are really expensive per square foot. Land the Columbia Tower sits on is appraised at $600/sq-ft. Of course you don’t put a bungalow on property like that. But construction costs are phenomenally high for a 76 story building. It’s a pretty good ROT that the taller the building the more you’re going to pay for comparable space.

      8. “Land the Columbia Tower sits on is appraised at $600/sq-ft. Of course you don’t put a bungalow on property like that.”

        That land has a high value mostly because of its location. A bungalow on 2,500sf of land built there would cost $1.5M for land – sure that’s expensive, but that’s on the order of land values for single family homes built near downtown areas (here‘s one in SF with a land value of $1,410/sf).

        Sure, changing to an in-demand zoning can bump up land values a bit. But that’s because we’re so limited in higher zones in Seattle. That effect drops as you upzone more property.

      9. Where did you find the land value? I don’t even see the lot size listed. Nice home though. Great transit connections and at 25% off the original asking price quite affordable at a multiple of only 287X the median household income := An aside, I was surprised to see (according to Census QuickFacts) that median household income and value of owner-occupied housing are the same in Seattle as SF. The percent of home ownership however is much higher in SF, 57.4% vs 48.9%. I’m guessing prop 13 has something to do with that but I can’t really explain it.

      10. It took me a minute too – scroll down to Public Records, and they have lot size, taxable land value ($9.8M), taxable additions ($4.4M), and 2011 taxes ($166k).

      11. Interesting stuff. The land value is appraised pretty close to the 1/3 of total property value that banks like to see. It’s appraised at $14M, went on the market for $20M and the price was dropped to $17.5M. That’s pretty much in line with residential appraisals in King County. King County however has a record of greatly under assessing commercial property even after a market sale value is recorded. For example, the Silver Platters parcel assessed at $1.8M (virtually entirely land value) sold way back in 2004 for $3.8M.

        For fun I compared the CA taxes to what I pay in Bellevue which has one of if not the lowest rates in King County. I was surprised to see SF is only 0.25% higher; probably close if not lower than Seattle. They have a slightly lower sales tax, 8.5% but then you’re hit with State income tax. The maximum rate is 9.3% but that kicks in at only $6k over the median SF household income. The gas tax is pretty and assuming they have roughly the range of sin taxes and fees as WA I’d pay roughly double in State tax if I lived down there. Yet while WA has had to tighten the belt CA is for all intents and purposes broke.

      12. I’m wasn’t impressed with the CA highways. The stretch of I-5 north of San Francisco is like a test track washboard section. CA does pour a lot of money into higher education and the U of CA state school system is perhaps tops in the country. Other than that I don’t see any state provided service where I’d be measurably better of there than in WA. No surprise NY City is at #2 for highest taxes. I have to question the methodology though when Boston is a #31 and Honolulu comes in at #33. Really, you pay more in Birmingham, Boise, or Des Moines? Detroit is #6 and Louisville is right next to NYC at #3. Mmmm… Seattle comes in at 43, the entire bottom of the list, 43-51 are the 9 States with zero income tax. Of those only Nevada and Texas have a serious budget problem. The State that’s really underwater is Illinios! Oregon’s deficit is the same size as Florida’s in millions of dollars but Portland ranks #18 in highest taxes.

      13. I-5 isn’t in great shape, but you should see the miles and miles of beautiful, often new state highways sprawling throughout the state.

        I’m not sure there’s a connection between handling budgets correctly and the level of taxation. You can balance your budget with a tiny amount of tax, you just won’t provide any service.

      14. Fair enough, the drive through northern CA to Lake Tahoe is spectacular but so is North Cascades. CA has many many more miles of roads but it’s a huge state both in area and size of it’s economy.

        You can balance your budget with a tiny amount of tax, you just won’t provide any service.

        But that’s not what happens. What are these great services that Illinois, New Jersey and California are providing that have pushed them so far into the red all the while collect higher taxes than Washington. The tendency of most state and certainly the federal government is to look at increased tax revenue as an increase in the limit on their credit card. When I was in college the sales tax was around 6-7% and nobody was complaining about lack of funding for higher education.

      15. “What are these great services…” Schools. Growing up in CA they provided driver’s ed, art clases, wood shop, computer labs (this was the late ’80s), and plenty of sports teams. Many cities even have full-day kindergarden and have been considering universal preschool. You’ve mentioned the colleges – they range from world-class univerities to almost-free community colleges.

        It looks like they spend 41% of their huge budget on schools. We spend more percentage-wise (52%), but we’ve got to be far, far behind on spending per student. Education doesn’t translate dollar-to-knowledge, but money sure doesn’t hurt.

      16. Turns out WA and CA are almost dead even in spending per student on primary and secondary education. The US Census published a report. WA is ever so slightly higher in spending, salaries and amount that actually goes to instruction. Both are pretty average; on par with the mid-west. The northeast not surprisingly is near the top. NY spends as much per student as WA and CA combined. A big surprise to me was to see WY and AK in the top tier.

      17. Interesting. I wonder if CA’s large immigrant population has something to do with that – they certainly collect more taxes than us, and the percentage spent on schools isn’t that different. AK isn’t a mystery at all – they have negative taxes from all of the oil, and labor is really expensive there. I have a friend that’s considering moving there, but is hesitating because the schools aren’t good.

        I have no idea what’s happening in WY.

      18. I have no idea what’s happening in WY.

        I think their situation is similar to AK in that they get huge revenues from the coal industry. And they have considerable natural gas and oil. Interesting that your friend thought the AK schools weren’t very good. They rank 42nd in HS graduation rate. Pretty bad but at least they’re throwing money at the problem. CA isn’t much better at 39th. WA is nothing to write home about at 38th. Wischeeskin is #1. They spend more but not a lot more. About the same as LA which is 48th. Nevada comes in dead last; they’re an odd duck as is their Senator Harry Reid. SD spends less than we do and ranks 6th.

      19. Regarding property taxes in CA (and specifically in the neighborhood of that mansion), see here.

        This is a pretty large part of California’s budget issues.

        Land costs are not fixed. Land zoned for higher buildings is more expensive.

        This is only true, I think, if there’s a very limited supply of land zoned for higher densities. If the zoning suddenly allowed higher densities over a large swath of the city, it wouldn’t change prices significantly.

  8. This is a pretty simplistic approach that doesn’t take into account real world factors that affect the marketplace, absorption being the first and foremost.

Comments are closed.