In a previous article, I responded to Dan Bertolet’s rebuttal of my essay outlining why urbanists must support linkage fees. My primary argument, that land is inelastic, means that land supply doesn’t increase when prices go up or decrease when prices lower.

Land was sometimes defined in classical and neoclassical economics as the “original and indestructible powers of the soil.”[1] Georgists hold that this implies a perfectly inelastic supply curve (i.e., zero elasticity), suggesting that a land value tax that recovers the rent of land for public purposes would not affect the opportunity cost of using land, but would instead only decrease the value of owning it. This view is supported by evidence that although land can come on and off the market, market inventories of land show if anything an inverse relationship to price (i.e., negative elasticity).

Inelastic land supply is incompatible with the theory that landowners will hold land off the market at lower prices. While the debate about the elasticity of land is important, it’s actually unnecessary. Even in a world with elastic land supply, the linkage fee wouldn’t reduce housing supply.

Connecting The Supply Of Land To Housing Supply

Overall, the rebuttal of the linkage fee did four things. It asserted that land is elastic. It then told a story about why landowners would hold their land off the market. It went on to explain why evidence showing the success of inclusionary zoning is insufficient. Finally it suggested there are better alternatives to linkage fees.

To engage further, I’ll entertain the assertion that land is elastic. With that assumption, I will examine if landowners will wait to sell until they get the price they want.  I will look at an example property going through multiple sales. Then I will examine if holding some land off the market could affect housing supply. Finally, I’ll run through the stories being told about the market and implore urbanists to refocus on housing limits, rather than regulatory costs.

The Role Of Development Capacity

The rebuttal article begins by drawing a line between land values and development capacity. To be clear, development capacity doesn’t cause development. Everyone should repeat this to Lesser Seattleites claiming, “Seattle has plenty of development capacity.” This is also the most logical response to Bertolet’s initial points regarding development capacity.

Land value is determined by calculating the potential revenue a property could earn and subtracting the cost to reach that potential. Improvement value on the other hand is the assessed value of current improvements.

How fees affect land value.
How fees affect land value.

Bertolet’s rebuttal begins by describing the common model for determining redevelopable land:

the ratio of the value of the improvements on the land, to the value of the land itself. The larger the ratio, the less likely a property will redevelop.

He shows that less land is measured as ‘redevelopable’ when property values drop without improvement value changing. This analysis is useful for understanding development capacity but he then says:

we can see that because a linkage fee reduces land value but not improvement value, it skews that ratio towards predicting no redevelopment.

 

To put a finer point on it:  Pickford’s position contradicts the standard methodology used by the City of Seattle, King County and countless other municipalities to estimate “buildable land,” that is, land that can be redeveloped. This alone should be enough to cast serious doubt.

Suggesting linkage fees won’t affect housing supply can’t contradict the standard methodology because the methodology doesn’t make predictions about housing supply. Development capacity models are primarily used to guide development to particular areas and predict if development is possible. Comparing land value to improvement value is a proxy for understanding if parcels are at their maximum use. But a drop in land value due to a fee doesn’t necessarily reduce the potential usage of a property.

It even turns out that there are different methods for determining redevelopable land. As an example, the Snohomish County Buildable Lands Report prepared by EconoNorthwest says:

Note that redevelopable land, as it is typically defined, deals primarily with parcels with developed structures that are judged as likely to be demolished and new buildings constructed in their place. The standard approach to identifying redevelopable land is to compare improvement value to land value. Many analyses assume that tax lots where improvement value falls below land value (a 1:1 improvement to land value ratio) are redevelopable. Not all, or even a majority of parcels that meet this criterion for redevelopment potential will be actually redeveloped during the planning period. The issue of how much of the potentially redevelopable land will be assumed to redevelop over the planning period needs to be considered.

 

An alternative approach to estimating redevelopment potential is to analyze the relationship of parcels to other surrounding parcels. For example, some jurisdictions define redevelopment potential as parcels that have improvement values significantly lower than surrounding parcels in similar designations. This approach, however, requires a property-by-property analysis using advanced GIS tools.

 

Another approach to estimating redevelopment potential is to analyze land value as a function of parcel size. In general, one would expect larger parcels with lower improvement values to have higher redevelopment potential. The distribution allows analysis of the relationship between improvement value and parcel size, and shows clear breakpoints in that distribution.”

The second method doesn’t even consider land values. Instead it just uses relative improvement value. But just like land value, improvement value isn’t a cause of development. Overall, the important point is that the models aren’t predicting development and neither improvement or land value cause development.

At What Price Are Profit-Maximizing Landowners Willing To Sell

It’s often stated that the city can’t simply reduce the value of someone’s land by fifty percent and expect them to sell. But what does this mean? The claim seems to rest on the assertion there is a tipping point for landowners to sell. A price too low and the landowner won’t sell. When the price is high enough, development will happen. Bertolet points to the development capacity model as a way to understand this but I don’t think that’s the best way. To represent this tipping point, I think the profit-maximizing landowner who received a bid from a developer would ask, “over the long run, is it more profitable to keep my property at the existing use or to sell it right now.” There are a lot of reasons selling might be better:

  1. Rents should be higher in a newer building because it’s new and often has more units;
  2. Operational costs may be increasing and income decreasing due to building age;
  3. Landowners may want to cash out and enjoy retirement; and
  4. A dollar now is worth more than a dollar in the future due to interest, investments and risk.

On the other side, there are costs to development. Obvious costs include things like; loans, labor, materials, demolition and regulation. But the known costs are shared by nearly all developers and are largely handled during bidding for land. The big cost is the risk of lower than expected demand, like a recession. This is why many developers went bankrupt during the last recession. They sunk costs into building and then their income evaporated. Everyone should keep this in mind when skewering developers as evil or taking insane profits.

Looking At An Example

To understand whether or not a landowner will sell, I’ll discuss a real example. Outsider guesses at the impact of linkage fees on land values vary widely from 5% to 40%. In previous comments, Bertolet indicated 25% and he’s pretty close. If we look at the history of the parcel containing the Sunset Electric, we can see a linkage fee would’ve reduced land value by 20%.  The 2012 buyer of this property would’ve had to consider, at worst, a $22 per square foot linkage fee because of their plans to redevelop. The building that now occupies that lot would’ve garnered about $1.35 million for affordable housing, or 20% of the $6.7 million price tag they paid for the property.

So if the bid for land in 2012 was 20% lower, would this have been enough to sell the property? The previous owners, 11th and Pine Associates, bought the property in 2006 for $3.275 million in 2012 dollars. They left the property empty the entire time they owned it, indicating nearly any price would’ve been greater than the income from the existing use.

Why did I choose this property, which didn’t already have income? It actually tells a decent story about how development happens. Developers typically target under-utilized properties. The property also appears to be an example of land speculation. It’s possible the 2006 buyers intended to develop the property and the recession got in the way but even considering that, the property was unused for at least a decade. During that time the property value increased more than 300%. It’s hard to believe a 20% lower land value, garnering over a million dollars for affordable housing, would’ve prevented or slowed a sale according to the logic of comparing existing uses to potential income.

Instead this evidence indicates landowners trying to cash out at the market peak, an entirely rational way to treat an asset like land. Unlike the developers that built the new Sunset Electric building, the money speculators received wasn’t from value they created. Furthermore, this development didn’t become possible because the land value hit the right level. The development became possible because Capitol Hill became a more desirable neighborhood, in a wealthy, growing city; the same reason the land value increased. The linkage fee would simply capture some of this value so that those with less money can continue to afford housing.

Anecdotal evidence can be misleading and no one should put too much weight on this example. But it’s more useful than theoretical stories. Linkage fee opponents will likely show an example of a project that wouldn’t ‘pencil out’ if linkage fees existed. This strategy is similar to the businesses suggesting a higher minimum wage would require laying off workers. It doesn’t acknowledge the fundamental economic shifts that will apply to everyone. Like the business that can raise their prices, because everyone else in the city also raises their prices, developers can lower their land bids because all other developers have to as well. Since we can’t trust anecdotes though, the reason I feel relatively confident is because this has been done in other cities. It turns out the reliable research show basically no affect on housing supply. This is more important than stories or anecdotal evidence.

But What If Fewer Landowners Are Willing To Sell

Under the worst scenario, some landowners are unwilling to sell for some period of time. This doesn’t even consider that upzones could increase the value of land. It also ignores all the arguments about inelastic land supply. Remarkably, even if this happens the fees should still work.

In order for housing supply to be affected through land, there has to be too few redevelopable parcels. If the development capacity model is related to the amount of redevelopable parcels, or the number of property owners willing to sell (something Bertolet is at least implying) this conclusion is unrealistic. The development capacity model estimates capacity for 223,713 housing units. While this isn’t a perfect model, Seattle isn’t lacking capacity. I’m not familiar with a clear explanation of how development capacity affects housing supply but opponents to linkage fees haven’t presented this either.

With that said, the assertion that fewer willing landowners mean less housing supply relies on some reduction in redevelopable capacity. In fact the reduction would have to be enough to prevent the amount of development that would happen. Last year developers only built about 8,000 units, or 3.5% of the redevelopable capacity. Would there have been less development if capacity had been 150,000, 100,000 or 10,000 units? Is it reasonable to think that a 20% decrease in land values would result in a 10, 20, 50 or 96.5% reduction in capacity? If capacity actually decreased drastically, would the Department of Planning and Development do nothing to increase this capacity? These are important questions for linkage fee opponents but they remain unanswered. Instead opponents focused on hypothetical stories about how market actors behave.

Stories We Tell About ‘The Market’

First, I found the retiree narrative confusing. The story goes that a potential retiree will hold their land off the market due to lower bids:

Slap on a million dollar linkage fee, and maybe they’ll wait it out through the next development cycle—perhaps 7 years or more—until market rents rise enough to offset that million dollar hit to their nest egg caused by the linkage fee.

Where did the million dollar fee come from? This would likely require a sale price in the millions of dollars. Apparently, adding millions to their retirement nest egg isn’t enough to retire, nevermind the fact that waiting 7 years might actually result in less money due to a recession. Or the fact that a retiree will likely want money sooner so they can spend it before they die.

Development Capacity Report from the City of Seattle.
Development Capacity Report from the City of Seattle.

Another confusing argument is that the fee makes development less likely in areas needing development, such as Northgate and Rainier Valley. It’s arguable these places ‘need development,’ just ask those worried about displacement. Overlooking this concern, my interpretation of this argument is that some people think these areas aren’t economically feasible:

A linkage fee’s level of impact on these hold/sell decisions will depend on the value of the existing uses relative to the land. And where it will have the greatest tendency to hinder sales for redevelopment is in areas with lower land values where feasibility is currently marginal, such as the Rainier Valley, Northgate, and the International Districtin other words, the very locations where the City most needs new housing to meet its goals for sustainable development.

But the model cited by Bertolet – not the model used by developers – shows these areas have lots of development capacity. This again gets at the tenuous relationship between land value, improvement value and actual development.

I think a more compelling explanation for why these areas aren’t seeing development is that they are relatively less profitable. Contrary to the claim he makes, his logic suggests the fee would likely increase the appeal of Rainier Valley and Northgate. The proposed fee is tiered from $5 to $22 dollars and these areas have the lowest fees. If a high fee discourages development a relatively lower fee should relatively increase development.

Moving on, readers find another archetype, the incompetent government bureaucrat. He says:

Given that every development project is unique, and given how the market varies over both geography and time, it’s delusional to believe that the City could set a linkage fee rate in some kind of “sweet spot” that wouldn’t end up sabotaging land transactions for redevelopment.

Immediately following this he goes on to say the evidence showing no affect on supply is weak because some municipalities enacted density bonuses with their inclusionary zoning programs.

If an IZ program is “revenue neutral” it means that it would not cause any reduction of land value, and therefore no impacts on housing production would be expected! In contrast, Seattle’s proposed linkage fee offers zero in the way of cost-offsets. Thus the results of all of the California IZ studies should be presumed to have minimal relevance to linkage fees.

In other words, inclusionary zoning was successfully implemented in nearly every municipality studied. Even though bureaucrats are incompetent they somehow managed to find both an appropriate fee level and an appropriate density bonus. Perhaps most importantly, if density bonuses work, there’s no reason the upzones we’ve already done wouldn’t work with some level of fee over the long run.

Linkage Fee feasibility.
Linkage Fee feasibility.

Again, the most charitable way to interpret these objections requires accepting there could be a level at which the fee would work. If that’s the case, it makes sense to trust the analysis provided by the city that actually worked through the math indicating development could handle the proposed fee. This analysis estimates that the proposed linkage fee would be equivalent to a 3-5% set aside of affordable units. This is both modest and cautious compared to the 10-20% set aside seen in most areas studied with successful programs.

Let’s Increase Development (Not Just Capacity)

Getting to the end result of more development is complicated. Increasing development capacity would be a good start. This fits perfectly with my call for urbanists to focus on housing limits, rather than regulatory costs.

No urbanist wants to reduce development capacity and most, including myself, want to increase it. But most urbanists know less development capacity doesn’t equal less development or vice versa. Pursuing efforts to improve the quality of the city, such as ensuring mixed income neighborhoods and equitable access to urban benefits, shouldn’t be avoided because of affects on development capacity. We shouldn’t get rid of Seattle City Light’s efforts to make buildings more energy efficient just because it increases improvement value.

What this tells me is that we need to focus on housing limits. Good urbanists have spent a lot of time and energy on reducing housing limits. I made this argument by providing evidence that regulatory costs probably don’t affect development. There is compelling evidence this approach is accurate. But even if it’s not, it doesn’t mean urbanists should focus on land values, improvement values or regulatory costs. There is not a straight line between regulatory costs and housing supply. There is a much clearer line between housing limits and supply.

requiring-bike-parking
Smart Growth Seattle telling a Cascade representative we don’t need bicycle parking.

To drive the point home, many regulatory costs improve the quality of life in Seattle. This is why affordable housing and social justice advocates support linkage fees, the fees will help achieve integrated neighborhoods in a way other affordable housing programs can’t. But the battle against regulation goes beyond linkage fees. In fact, opposing regulations aimed at making the city nicer draws a line between urbanists and many other advocacy groups since quality of life is so closely related to development. Affordable housing advocates read Bertolet’s article and perceive a battle with urbanists. Is a battle with the neighborhood greenways and transit advocates next? Or should we just move on to fighting education activists? If regulatory costs don’t affect supply or even have a marginal impact, this is extremely counter-productive.

It’s also counter productive if efforts to remove housing limits will overwhelm other factors. Political outcomes are unpredictable and can have unintended consequences. As an example, I believe Bertolet’s argument inadvertently provides evidence linkage fees will cause upzones. The city is required by law to plan for growth by expanding development capacity when it drops below 20 years of growth. If we see declining capacity due to lower land values, the city will likely pursue upzones to meet GMA requirements. By definition lower land values would create the impetus for further upzones.

There are great urbanists that disagree with my view; reasonable people can disagree. I don’t expect them to change their minds; people rarely do. But my take is that the best we can do when trying to understand economic outcomes is to look at empirical research. I’d be happy to continue looking at research regarding land elasticity or the effects of linkage fees but I don’t think Bertolet’s rebuttal presented any new empirical evidence. Instead it focused on a narrative story. Furthermore, after dismissing the evidence provided he claimed the burden of proof is on The Urbanist. I disagree and believe the burden of proof should be on those that have not presented supporting evidence. Most importantly though, I believe a successful urbanist political effort will require focusing on things that most clearly harm housing supply, housing limits. It will also require building coalitions with natural allies that want regulations making the city more livable. If no one changes their mind about linkage fees, urbanists could still refocus on housing limits. In the end, if urbanism is going to be a viable political force, I believe urbanists must distinguish between the efforts to eliminate regulations and the fight to eliminate housing limits.

We hope you loved this article. If so, please consider subscribing or donating. The Urbanist is a non-profit that depends on donations from readers like you.

23 COMMENTS

  1. Owen,

    Thank you for developing a very thoughtful and well documented analysis. You do really thorough work and I am glad you are out there working hard to understand the market consequences of these policy proposals. I’ve also read through David Rosen’s “Seattle Affordable Housing Nexus Study” and understand that your conclusion on supply inelasticity may be derived from that work.

    I agree that the supply of physical land is absolutely inelastic. As a wise old developer once said “you can build more buildings and plant more trees, but God isn’t making any more dirt”. What trades in a “land sale” isn’t exactly the land itself, but a bundle of rights temporarily held by an estate, “property rights”. Land itself is perpetual, no more or less will ever exist. Landowners are time limited, we are born, do some work, pay some taxes, and die.

    Even though a specific narrowly defined location may have a very limited availability, on a city-wide or regional basis, the vast number of participants in the market nullifies the premise that the supply of property rights is perfectly inelastic. Each individual owner has a different basis, zoning rules create a variety of legally allowable uses, physical conditions vary from site to site influencing design options, and the market is constantly changing creating uncertainty and risk. These variables mean landowners and estate trustees face a cloud of potential decision drivers influencing sale decision and timing.

    Generally speaking, as prices rise, barriers to selling are removed and more owners are willing to transact their property rights. This distinction is pretty much the definition of supply elasticity as far as a developer is concerned.

    One could argue that Government can put a land owner in a headlock and steal their property rights. (Legally called taxes) Fortunately, in our country a landowner is due just compensation in the event an eminent domain action is taken. That protection insures a reasonable level of willingness to negotiate and trade property rights rather than ending up with a culture of monarchy ownership of all lands.

    I completely support the idea that our City needs to find a creative way to develop more and better affordable housing solutions across the board so don’t take my next statement the wrong way.

    Given David Rosen’s assumption that land is perfectly inelastic is not a reasonable starting place for the analysis of market dynamics in play for developers (let alone the proposed linkage fee/tax(possibly illegal taking); would you consider revisiting your analysis of the impact of linkage taxes? Have you considered looking into the constitutional legality of creating a tax on new construction? I’m not a lawyer so not sure if I am completely correct but it seems to me that the application of this concept could be challenged as a form of “taking” that may impact any landowner in the city who owns anything other than SFR zoned land. If true, wouldn’t this proposed linkage fee be very tempting bait for some law firm to put together a class action suit against the city? Have you considered writing a piece on that risk?

    You have developed an excellent understanding of the remaining metrics and are an excellent writer. I think you could do a fantastic job analyzing this issue absent the false assumption of land inelasticity.

    One other thing I wonder if you would consider exploring.

    Much is currently being made about developer and landlord profiteering as a result of the rebound in demand we are experiencing regionally. Almost nothing is being reported about the consumer side surplus associated with development projects. From an economics perspective there is always both a producer and consumer surplus. When demand is less elastic than supply generally the consumer surplus is greater than the producer surplus. Individually, developers get a very concentrated form of producer surplus if they are successful on a project. They also take a tremendous amount of risk. Consumer surplus is spread over the entire population of users who benefit from the project but take considerably less risk when they buy.

    Have you ever considered developing an analysis of the relationship between producer and supplier surpluses in our housing market on a risk adjusted basis?

    Thanks for your great work!

    • I’m no lawyer, either. but I remember a similar issue came up with environmental regulation. Basically, someone complained that new regulations made their land worth a lot less. One of the justices then asked if the land was worthless — in other words, could he buy it for a dollar. The owner said no, and long story short, the government can regulate things as much as they want (as long as it doesn’t completely destroy the value of the land).

      In this case, the city could do the same thing, either with zoning or with fees. The city could restrict all new development, if they wanted. A lot of owners would be upset, but the land would retain value (in its old use) so they couldn’t do much from a legal standpoint. So far as I know, there is nothing special with regards to state law on this issue, either.

      • Rossb,

        Of course we can regulate. The cost of the regulations is what I am questioning.

        If this particular legislation is deemed unconstitutional, the cost will be significant in both failing to achieve the stated goal of creating a reserve for building affordable housing as well has costing tax payers hundreds of millions of dollars in defending against suits, and rising rents resulting from further supply constraints. The law will create a wind fall for existing landlords and create additional pressures to move toward rent controls (also illegal) because of limited new supply and continued increasing demand.

        I’m trying to look at the “holistic costs” not zeroing in too closely on the right of government to regulate. Eminent domain is very expensive.

        • and what I’m saying is that based on my understanding of prior court cases, it is very unlikely that this would be ruled unconstitutional (in federal court). Simply changing the zoning to not allow additional development would probably not be considered a “taking”, since the owner could still develop the land with a smaller building (a house). Yes theat is more a taking than this. Taxing development in this manner has been applied in other states. Someone could sue, but they would probably be unsuccessful (unless there is some other angle that I’m not thinking of and the opponents in those other states didn’t think of). But you could say that about any change in the law. Someone could always sue, it is just that most of the time, they would lose.

          If there is some state law or state constitutional restriction, then that is different, but I don’t know of one, and I assume the folks who wrote this don’t believe there is one either. This makes it different than rent control, which is clearly illegal by state law.

          So, yes, worse case scenario folks spend a lot of money in court, but to be frank, whether we have a substantial amount of new affordable housing will not rest on the outcome. This will, at best, amount to a very small amount of housing that would otherwise not be built. Changing the zoning laws, on the other hand, could result in a lot more.

          • There is a long history of rulings — Wikipedia lists a bunch: https://en.wikipedia.org/wiki/Regulatory_taking
            I can’t find the one I was thinking of, probably because it wasn’t that big of a case. Basically, the court has ruled many times in favor of the state, and many times in favor of the land owner. But based on those cases, I just don’t see this being challenged in federal court (especially since other owners in other jurisdictions haven’t successfully challenged it). If we were the first to try this, then I would be worried — but this has been going on for a really long time.

            As far as my other statement goes, I mean that changes in the zoning law — changed in what is allowed, would do a lot more for affordable middle class housing than this. To be clear, this will provide some very low income housing to a handful of folks (who sit through the permit process long enough to get their unit). But changing the zoning law to permit more construction is the best way to lower prices. It is why we are in this mess in the first place, and why this won’t have that detrimental an effect on construction. The demand is so high, and the amount of land that allows development is so low that this fee is rather minor in the big scheme of things. But if the city changed the zoning for low rise, or ADUs, or got rid of the parking requirements (for everything) it would have a bigger effect. If the city just changed the ADU laws (made it more like Vancouver) then it would result in a lot more affordable units than this policy. Ultimately, this is a very interesting argument (whether linkage fees are good or bad public policy) but a side show to what is really important, which is reforming zoning laws (for example, this: http://daily.sightline.org/blog_series/legalizing-inexpensive-housing/).

          • Sounds like we are on the same page with supply increases being the best way to counter skyrocketing market-rate prices.

            Your conclusion that the linkage fee won’t be challenged is a little shaky given the number of dollars of consequence on the line and the probability of identifying a class of folks willing to put up retainer fees being quite good.

  2. I’m glad you referenced the minimum wage, because I think it is an excellent analogy. In general, most economists (and they have plenty of studies to support them) believe that raising the minimum wage “just enough” is great for the economy. You basically redistribute wealth that would otherwise simply go to owners. Every increase leads to a decrease in hiring, but if you get it right, this decrease is small, and offset by the sharing of wealth that occurs. At the same time, I don’t think there is an economist out there that believes that we should raise the minimum wage to a million dollars an hour. That would result in a lot of unemployment.

    With linkage fees it is the same thing. With a hot market, you are basically redistributing wealth, and the small decrease in construction is made up for the new construction that is done with the money you raised by the fees. This is money that would otherwise just go to the owners. Set the number right, and it works out right. You have also made a very good case that the city did just that — they set the rate right, and the result will be a net increase in properties.

    But that assumes no change in the market. You correctly pointed out that construction ground to a halt during the recession. But why? Demand was still there (rent didn’t drop to zero) but it lowered to the point where the construction cost exceeded the value added cost. Property value took a big hit (as they will with linkage fees) but construction costs were still relatively high. If the city subsidized construction (covered half the cost, for example) then I’m sure those buildings would have been built a long time ago. So basically a drop in demand (but not a drop to zero) was enough to slow construction to a crawl, given construction costs (which didn’t drop that much).

    I could see the same thing happening in Seattle. There are many who have predicted a housing bubble in this town, and plenty who would love to see one. This would mean that rent prices would actually go down. Are linkage fees flexible enough to drop when that happens?

    The city could also create a much better market for renters by changing our zoning laws. If Seattle liberalized its ADU laws (as well as many other rules) I think you would see a potential flood of new, affordable units on the market. This would put downward pressure on the market, and blow a hole in the numbers that they used to determine the “just right” level for linkage fees.

    In other words, unless the linkage fees are flexible — unless they are set every year, I think liberalizing the zoning laws will be counteracted by the fees, and the effect they have on construction. Allow a bunch of new ADUs, and next thing you know, that warehouse that was going to be replaced by an apartment stays as a warehouse. That might not be that bad — the linkage fees aren’t applied to ADU or low rise (I assume) and the small construction market might overwhelm the rest of the market to the point where bigger buildings aren’t even needed. But I personally would want to make sure that this is flexible, so that we can continue to see a lot of construction (a lot more construction) and actually see a lowering of rental prices.

    • I would be willing to consider looking at linkage fees in a way to make them more flexible. I think your summary here is pretty good. I also support liberalized ADU laws.

      The only point of disagreement though, I think, is regarding the recession. I would say demand dried up during the recession. It’s also important to note that construction lags demand. So if construction stopped during or after the recession, the demand it was reacting to was after that. Demand for rental housing increased following the recession due to foreclosures and I think we are only now, finally seeing the uptick in construction.

      Theoretically, I have a hard time imagining a world with declining rents and increasing demand due to the lag in construction. I think this would be necessary in order to make the fee problematic.

      On an additional note, I also think the fee is set really, really low so it has a lot of breathing room before being problematic.

      • I agree that demand dried up during the recession, but it didn’t dry up to zero, did it? Are you claiming that no one would buy a skyscraper in Seattle, at any price? Lower the price to $100 and I would have paid cash for it.

        It was demand relative to cost that dried up. Folks would pay ten million for the building, but it costs 15 million to build it, so it sat as an empty lot (or, if they were lucky, a parking lot). Before the recession, they would have got 20 million. So absolutely, it was demand (a drop from 20 million to ten million) that caused construction to fail. But ultimately, this is just shorthand for “demand relative to cost”. Its that way with almost all products (software being an exception).

        I am pleased to here that the fees will be really low. That suggests that it won’t be like raising minimum wage $100, but more like raising it a quarter. That being said, I’m not sure how much money it will raise. It seems to me like a simply property tax would be a much better way to raise money. But I guess it is easier to tax one small group than it is everyone.

  3. ” I’m not familiar with a clear explanation of how development capacity affects housing supply but opponents to linkage fees haven’t presented this either.”

    I think you understand some of this but I’ll dive in. First be very wary of any development capacity analysis, as it’s a 1-dimensional number. It’s the same kind of simplification as trying to say what the demand is for housing, when in reality it’s a curve not a point.

    Development capacity determines the price of housing in a growing market. Note this is a bold statement, and an important one. Developers build when they predict that there is demand for housing at the price point they can build it plus profit. However, as development capacity decreases we go from building on empty lots to building on parking lots and single story structures to tearing down housing to build housing, etc. Looking at the extreme, you need to tear down a 39 story apartment building to build a new 40 story apartment building. Housing prices would have to be insanely expensive to justify that.

    Let’s take a less extreme example. Take a 3-story apartment building with $2k/month rents and 5 units per floor. That’s $360,000 a year you have to throw away while you’re building your new apartments. Then, when you’re done building a 6-story apartment building you’re only adding rent to half of the units – you’ve just torn the other 3 floors down and rebuilt them. This means all of that construction cost needs to be made up in income from only half your units.

    This building would not be built until rents are high enough to justify this construction. Of course if this is typical of the building stock available, then rents will rise to this level as there’s otherwise no new supply to meet this demand. And note that if this case was typical of the building stock available you could have a development capacity equal to your entire city’s building square footage.

    My point is that development capacity is important, and we need way, way more of it than many people assume. You can’t say “Seattle has plenty fo development capacity” just because the number looks big.

    • I agree with everything you are saying but people who dismiss the development capacity numbers as insufficient have not provided a clear example of how adding redevelopable land would reduce costs. I get that adding supply is important but redevelopable land doesn’t equal more development. That’s kind of the point of this article.

      Costs are a function of the rents people are willing to pay because landowners will charge up to the maximum they can for their land based on the rents of the land. If we open up development capacity in areas in which demolition is less expensive, we don’t necessarily get less expensive rents. In other words, downtown office towers built on surface parking lots don’t have lower rents than office towers that replaced 10 story buildings.

      Overall supply is critical but supply and demand is hyper-local so the affect of increasing capacity on the level or rents is tenuous at worst and unclear at best.

      • I don’t see how you can agree with everything I said and not think that adding redevelopable land can decrease rents. Assuming this land is nearby in a similar market*, dropping the cost to redevelop property absolutely drops the rents of this building and everything nearby. Remember, development capacity determines the price of housing in a growing market.

        “downtown office towers built on surface parking lots don’t have lower rents than office towers that replaced 10 story buildings” Zoom out a step. Opening up that surface lot for development lowers the rent for all nearby units, if the market was at a point where it made economic sense to tear down 10 story buildings and rebuild. You’ve added supply without touching demand. If it was just over that point then you end up with twice the number of units available, dropping rents. If it was just under the point where a 10-story teardown was economically feasible then you’ve added a building’s worth of units without having to reach that point.

        * And by similar market I mean anything a potential tenant would consider as an alternative. For potential homeowners or renters, this may be a significant area of the city and beyond. Of course there are secondary effects throughout the market – as people choose the new location the unit they would have rented frees up, and this cascades through the market.

        • Your argument is plausible and my previous comment is a little over simplified. But I would reiterate the cost of the construction is nearly irrelevant. If the downtown skyscraper adds enough capacity to actually affect demand, it’s not because the construction costs were lower due to it being built on a surface parking lot.

          As we zoom further out and suggest that there is some level at which development isn’t feasible because of construction costs, what we are really saying is that there isn’t enough demand for the costs.

          But zooming out too far distorts how the housing market actually works. For example, cheap housing in Texas likely doesn’t affect housing costs in Seattle. So what is the appropriate scale to look at? I personally think demand is completely related to neighborhood urban benefits. In other words, housing in Ballard likely isn’t competing with Shoreline. These areas are unique because of their urban benefits. That’s why building in shoreline would likely have no affect on rents in Ballard. This argument slowly spirals into the argument I was making about hierarchical price levels in neighborhoods and segregation.

          In the end, if construction of a specific area aren’t what’s driving prices, and instead demand, like the downtown skyscraper example, further costs can be passed on as lower land prices. This is why a linkage fee on a downtown sky scraper wouldn’t affect supply.

          • “As we zoom further out and suggest that there is some level at which development isn’t feasible because of construction costs, what we are really saying is that there isn’t enough demand for the costs.”

            Yes. Developers are all working with the same land and markets. As demand goes up, rents go up until new redevelopments begin to pencil out. Construction starts, supply is added, and demand is satisfied for a while (unless it’s risen enough while that cycle of building happened to justify the next cycle). So it’s the cost of these redevelopments that sets rents.

            “housing in Ballard likely isn’t competing with Shoreline” I disagree completely – every unit built in Seattle means we don’t build a unit out in sprawl. But even with your urban benefits argument Ballard is certainly competing with West Seattle or at least Fremont. If a whole section of the city is what you mean by hyper-local, then at least I can understand your point.

          • I agree that building units in Seattle likely means less sprawl and this article strongly advocates for eliminating housing limits. The point I’m making is that if you are a shopper willing to spend luxury rents on new units in Ballard, there is a hierarchy of where you are looking. Vacancy rates for new units in Ballard are high enough you’ll likely find a unit there. But if you can’t, you aren’t going to go to shoreline. I would say Bellevue is more likely, or as you say West Seattle or Fremont. If higher bidding renters start to go to those places they will outbid existing renters, and ultimately raise the value of land. This is where the extra value is being captured.

            This increase in land value was at no effort of the existing landowners. Reducing the return for landowners doesn’t change developers calculations and or renters. If we think of this another way, if we increase the housing in Ballard so this person doesn’t go to West Seattle or Fremont, it may very well put downward pressure on individual rents but it still increases land values because the absolute amount that can be earned from an individual parcel is higher.

          • As a side note, I really appreciate this conversation Matt. Even though we disagree I think you do an honest and thoughtful job challenging the assumptions and throught process I’ve had. You’ve made me think harder about this issue.

        • “But I would reiterate the cost of the construction is nearly irrelevant”

          No, no it’s not: http://www.theurbanist.org/2014/09/02/85-foot-and-125-foot-height-limits-are-a-missed-opportunity/

          Builders don’t always build right up to the limit, even though they are legally allowed to, because it is too expensive. The same is true on the margins — for example the 3 story building replaced by a six story one. If it was free to make that conversion, then most landlords would, and you would see a flood of new units, which in turn would lead to lower prices.

          As to larger argument, I’ll post that below, because I would rather not type too much in this little box (I want a bigger box).

          • This is a different argument than the one I am making. What you are talking about are skewing affects, in which costs are better or worse under various situations. Different levels of costs can create different incentives.

            What I’m talking about are whether or not costs applied universally across developments will affect supply. In this situation, there isn’t a skewing affect.

            Stepping back from the theory into the actual proposal, there are differences across geographical boundaries but those were specifically designed to accommodate the different circumstances of each area in the city. Overall, I think skewing affects would be minimal but one could definitely argue otherwise.

            Also, in the example provided in the link, just because a developer didn’t use their capital to build out to capacity doesn’t mean that capital wasn’t used elsewhere to build more housing. You’d have to take a really macro level view to understand the effects. This isn’t to say that not building to capacity is ok. I think it sucks. I want more housing in SLU. But this could’ve been solved with increasing height limits, something affordable housing advocates likely could’ve gotten behind if they could be shown the increased revenue. A political coalition between these two groups might’ve been enough to overcome the opposition to more height.

          • How is that different?

            You claimed that “the cost of the construction is nearly irrelevant”

            Irrelevant to whom, or to what?

            It is obvious it is relevant to developers, as I found an example of an entire class of buildings that aren’t build as high as they would otherwise be built only because of the cost of construction. That isn’t theory — that is the stated reason given by the developer.

            I’m sure if you asked the developers during the recession why they stopped building, they would give the same answer, but it would take a while.

            Q: Why did you stop?
            A: It isn’t worth it — rent went down.
            Q: But it didn’t go to zero — your building would still be worth more than the parking lot.
            A: Right, but building a building is expensive — dude, I thought that was obvious.
            Q: So you would build it was much cheaper to build?
            A; Of course I would, but it would have to be a lot cheaper — like if the city subsidized it or something.

            The second case is macro — very macro, and it was repeated throughout the country. Construction costs dropped, but they didn’t drop as much as the value of the new units. If it did, then construction would have continued.

          • I apologize. As I mentioned in my previous comment, that sentence is not precise and I can see why it would pass the wrong interpretation. I think construction costs are important.

            What I was trying to say was that costs applied the same to everyone will be absorbed in land values, with no skewing affects.

            The recession is a bad example for a lot of reasons but especially because it was a credit crunch followed by a decrease in demand. Who is going to risk capital when there is already enough supply on the market? What developer is going to get a loan when other developers are going bankrupt. Who can get capital when there is credit crunch?

            The recession ultimately saw prices decrease. Decreasing prices are a strong signal not to build more.

          • But everything you mentioned is a cost. Borrowing money is a cost. There is a cost to build, otherwise it would be easy to borrow money or you wouldn’t need to borrow at all (just use your savings). So basically, banks wouldn’t loan money because the cost didn’t match the expected gain. Again, this is because the cost of construction can’t drop that low. Labor costs dropped, but not that low. Material costs dropped, but not that low. It never hit zero (it never could) which is why you have a housing recession (along with a fiscal crisis) that spread throughout the entire economy.

            The same loss of demand happened in other markets as well. The demand for electronics (e. g. smart phones) dropped a huge amount. Thousands of people out of work — who wants the next iPhone? But they kept making them because it is so ridiculously cheap to make them — the prices just went down.

            Construction costs have a floor — a fairly high floor. Material costs are expensive, especially with world wide demand that remained relatively high. Labor costs hit a floor, because labor doesn’t want to work for nothing (they are better off waiting it out). This floor is why a recession in housing is fairly common. The fact that it took out the rest of the economy was really due to the financial shenanigans that banks performed.

            Like I said in my little Q and A, that is basically what happened. No builder would ever say that he doesn’t build because the cost is too high. That is because the costs actually went down. But ultimately, that was the cause of the empty lots that sat for a few years. Demand dropped (the obvious reason) but if construction costs went down far enough, then you would have seen new construction. It just couldn’t, absent government intervention.

            Smart phones were made despite greatly reduced demand. Software kept being sold, despite greatly reduced demand. This happened because the cost to build were so low that it didn’t matter. This didn’t happen with construction because the cost is too high.

          • I’m honestly unsure I understand the point you are making. If the following points aren’t addressing what you are saying, I apologize.

            I’m making a point about demand. If no one is moving to Seattle and instead people are moving out of Seattle, it doesn’t matter whether or not rents are high enough to pay off new construction; generally speaking there won’t be new construction.

            More generally, costs are accounted for in land values. If construction costs no longer pencil out, landowners can’t sell at their old price and the value of land drops. This means if there is actually demand for new housing it can still be built.

Comments are closed.