In my last article on linkage fees I made the case that the typical supply and demand dynamics we are familiar with from buying gasoline or coffee do not work with land development. Land is different from nearly all other types of goods because when demand goes up we can’t make more of it and when demand goes down we can’t ship it somewhere else or break it into its components and turn it into something else. Land supply is perfectly inelastic–a change in land price doesn’t change the amount of land–and a linkage fee, therefore, will not change the supply of land. But a significant question remains: won’t landowners simply hold back land from the market until prices rise again, effectively (if not technically) lowering supply?

If this is true it could be a major problem for linkage fees. If landowners refuse to sell until prices return to their pre-linkage fee levels, developers are no longer able to recover the cost of the linkage fee by paying lower land prices. Development costs go up and we are back in the familiar supply-and-demand negotiation where renters lose, developers lose, or both lose.

Naturally, we can’t directly test how linkage fees will affect how much property is brought onto the market because it is merely an outline of proposed legislation at this point (although two developers in an anonymous survey conducted by Vulcan suspected anticipation of linkage fees becoming law would push land prices down). We can, however, indirectly test the effect of linkage fees by examining market data more generally.

After reviewing the available evidence, I can say with about as much confidence as one can honestly muster from the messy business of real estate research that land supplied to the market is not diminished during periods of falling land prices and, conversely, land supplied to the market does not increase during periods of rising land prices. Property supplied to the market does change month-to-month but it does not correlate with trends in the market value of properties. That is, taking it one step further, development fees that lower land values would not restrict land from being made available to the market.

Why Landowners Sell

Before getting into the evidence, it’s worth just thinking about what motivates landowners to sell or hold their land. A rational landowner — the Econ 101 financially self-interested market actor — holds or sells land based on anticipated future income of that land compared to all the available alternatives. If anticipated income is greater than all the financial alternatives of selling the land and putting that money to some other use, then the rational landowner holds on to the land. If it is less, she sells it and invests in Blue Chip stocks or tulip futures. What doesn’t influence the rational landowner is past losses. The rational landowner is always trying to maximize future returns, not recover past losses.

The typical landowner is not a financially self-interested market actor, of course. Most of Seattle’s landowners live in the homes they own. For these relatively risk-averse landowners, holding may be the favored choice because owning the land and profiting off its use provides more stability (if less income) than selling the land and investing it somewhere more lucrative (and having to find a new place to live). What makes landowners offer their property to the market? It probably has a lot to do with life cycle changes: a new job, marriage, divorce, retirement, or the death of an owner and the transfer of land into a trust to be sold and the proceeds divvied up as inheritance (this is often how farms become residential subdivisions).

I mention these two types of landowners for two reasons. The first is that despite having different motivations about when they may choose to sell property, neither motivation is guided by comparing past prices to present prices. The second reason is that a good set of historic commercial sales data is expensive and, additionally, commercial land is so heterogeneous due to size, location, zoning, etc. to make comparison to other variables often unsuitable.

Market Evidence

In order to test the criticism that landowners sell less when prices decline I first examined how stock market trading volumes change in relation to price. Undoubtedly, the stock market is an imperfect analogy for the real estate market but it is easier data to get and offers a much larger set of data points to test the claim made by some opponents of linkage fees.

DJIA volume price

Yahoo! Finance

Here we can step back and look at the relationship between price and volume over three booms and two busts in the last 20 years of the Dow Jones Industrial Average. There is no correlation or, rather, there is no positive correlation between price (line) and volume (bars) in the stock market. If anything, volumes are highest around the bottoms of the two busts and have declined gradually since the market recovery began in 2009.

The correlation between price and volume has been studied in decades of financial research with varying conclusions. A short overview of this research can be found in Chapter 3 of this MIT course document (pdf file). A 2011 study (pdf file) attributes historically high trading volume in recent years to the declining cost of making stock trades. Generally, though, the research has found that absolute price change is correlated with trading volume but not necessarily positive price change.

If linkage fees restrict real estate market activity, and the reason it does this is based on bedrock economic principles so as to be generalizable to the stock market, we should not see what we see in the chart above. Trading volume and stock price are marching to two different drummers.

But that’s stock market behavior and we’re just interested in real estate market behavior. In order to get closer to the question at hand–do price changes in real estate affect real estate available to purchase?–I examined monthly reports from the local multiple-listing service (NWMLS) and compared active listings (properties for sale) to sale price averages for homes and condos. This data (Excel file) is neatly compiled and provided by Seattle Bubble, a local real estate blog. I have placed 12-month moving average lines on each set of data to control for seasonality. The data runs from January 2000 to March 2015, the most recent update.

Listings v Price
Data is courtesy of Seattle Bubble

The chart reveals that although there are periods where price and active listings appear to move in tandem there are also periods where they appear to move in opposite directions, the last three years being especially dramatic as the real estate market recovered while listings declined to their lowest levels in at least 15 years. Again, if some linkage fee critics are right, we should expect more active listings as sale prices go up and less active listings as prices decline. The evidence does not support that claim.

Linkage Fees Are Fair, Effective, and Feasible

As an initial response to the criticism that landowners would withhold properties from the market should linkage fee legislation pass, I attempt to challenge the claim using presently available public data, the best we can hope for with untested legislation. The evidence shows that declining property values do not decrease supply available to the market and, conversely, increasing values do not bring more properties onto the market.

Without a doubt, increasing supply by making it easier to build multi-family housing in Seattle is an important and worthy policy goal. We need to do it. We need to make design review more effective. But we don’t need to legislate two sinks in aPodments. And we need more multi-family options all over the city. In addition to these efforts we need to work towards capturing windfall land profits which are primarily responsible for the increased cost of housing in Seattle and many other growing cities.

The Economist article that quickly sped around the internet a few weeks ago concludes that a land tax is effective and fair but politically and logistically difficult. The Linkage Fee, though, is easy to administer and a politically achievable policy that works much like a land tax. It will produce lots of new affordable housing supply by stuffing the housing trust fund with the proceeds of what is effectively a capital gains tax on land sales.

Photo credit: i don’t live here anymore (detail), Christine Hou

 

44 COMMENTS

  1. I think you’re looking at the wrong things. Here’s a short list as I step through your article.
    1. Although you can’t make more land, you can make more housing on each piece of land. That’s why one of the best ways to limit housing price increases is by upzoning.
    2. You’re looking at a linkage fee as if it’s a proxy for a land value tax. It’s not. You aren’t taxing based on land value, you’re taxing based on square footage of new construction. This is very important and ignored in your article.
    3. You can’t say that land prices don’t act like anything else, then use the stock market as an example of what will happen to land prices.
    4. Your stock also analysis ignores #2. It’s not the price of land going up and down that will affect sales volumes, it’s the tax on development. This would be analogous to adding a large tax per stock you trade (not even per trade, but per stock).
    5. This is also what’s wrong with your look at real estate prices and volumes.

    Remember, this isn’t a land value tax that applies per square footage of land based on value. This is a development tax, based on square footage of construction. Leave a lot as a parking lot and you don’t pay a penny. A 10 story building is taxed 10x that of a 1 story building. A 100 story building is taxed 100x. You have to see that this will decrease construction – it’s always a good idea to tax that you want less of, not more of.

    • The linkage fee is tiered to different sub-markets and based on building sq ft. This formula is effectively a proxy for land value. That’s the clever part about it.

      • That makes it nothing like land value. You aren’t taxing land on a continuing basis, only when someone tries to build on it. And when they do you charge them based on how much they build!

        • Hi Matt,

          Your objection seems to be whether or not the fee it is passed on to land owners. Michael is making the assumption that it is passed on to land owners and you can see why in his first article. If you don’t buy that assumption then this article doesn’t make much sense.

          Because Michael is making that assumption, this article starts from there and asks, ‘if land value decreases will less land be sold.’

          Also, with that assumption the comparison to a capital gains tax on land is pretty apt. The value of land increases as one can build more on it. If the fee didn’t increase, it wouldn’t scale with the value and be similar to a capital gains tax on land.

          But again, this all starts from the assumption that the fee is passed on to landowners.

          • And what I’m saying is that “if land value decreases will less land be sold” is not a useful question. Even giving Michael all of the assumptions, I care only for the number of units built, not the number of times property changes hands.

          • You’re right it’s not a useful question if you disagree with the premise that the cost is passed on to landowners. If you believe the cost is eaten by the developers it’s possible fewer units would be produced.

            On the other hand, if you agree that the cost is passed on to land owners, this question is very important. In fact, this was the objection raised by Dan Bertolet and Roger Valdez regarding the last post which indicated the cost would be passed on to land owners. They somewhat agreed with this argument but the suggested less land would be sold.

            If the cost is passed on to land owners and land doesn’t come off the market, the total number of units produced won’t lessen. In fact, it might even increase because the wealth captured by land owners would go to a public fund to build more units, instead of being taken and used for other purposes.

          • You guys are apparently claiming that the price offered for an income generating property has no relationship to the owner’s willingness to sell. I don’t understand how that could be. Help me out here. Is there any limit? What if the linkage fee was equivalent to 90% of the current land price?

          • With any asset the buy/sell decision is not based on the current price alone. It’s based on the relationship between the current price and the expected future price. Are you making the argument that there are more sales when prices are higher? If so the evidence and theory suggests otherwise.

          • The evidence in the post is for private homes and condos. It seems to me that most people make decisions to sell homes and condos for very different reasons than owners of income generating property do. Do you have any other evidence?

            My anecdotal observation is that when land values rise, you tend to get a lot more redevelopment of underutilized commercial property into higher intensity multifamily, as is happening in hot neighborhoods in Seattle right now. So I think that means there must be more sales, yes.

            Are you saying that if a linkage fee had been imposed, the previous owners of all those redeveloped properties would have sold just the same even though they would have been offered less for their property? And that’s because the relationship between current and future price changes under a linkage fee? If so, I honestly would appreciate more explanation, because I don’t get it, and it’s too late to think that hard.

          • First private homes and condos have imputed rent. But I agree, this case could be made stronger by including other types of property. Michael might look into doing that.

            To your second point, anecdotal evidence, keep in mind that rising values have a chicken and egg relationship to new development and by no means indicate causation. To your question of whether or not there are more sales, look at the 13 through 16 on that graph. You are seeing some of the highest values and the least number of sales. I’m inclined to believe theis graph until someone provides evidence showing prices are positively associated with the number of sales.

            To your last point, I think Michael would qualify his answer. I would just repeat that the value of land would not affect the amount of land sold. The evidence above points in this direction and rational economic actor theory does too.

            If you don’t understand the theory, I’ll give it one more go. Basically, the current selling price is not the sole determinant influencing whether or not someone sells their property. Instead, the property owner is either selling for personal reasons (the non-rational market actor) or is trying to maximize their properties value, waiting until the market peaks before selling (the rational econ).

            In the example you keep using, you are imagining a property owner who is waiting for the property value to exceed the rents they are earning and then they will sell. This would not be rational for two reasons. First, the value from selling is always higher than the rents someone earns. You can extend those rents out over a long period of time, months or even years and the one-time lump sum would still be larger. The asset value is not liquid. Asset values aren’t compared to rents to determine if they are greater. This is true of stocks that earn dividends. This is because the buyer is paying for rents over a very long period of time. Second, the rational property owner doesn’t care what the buyer is offering at the moment. They care what the current offer is in comparison to future offers. That is because if they wait to sell and get less because the market tanked they will be in a worse position. Someone holding an asset doesn’t know whether the future value will exceed their current price. It’s not as simple as waiting until their property is worth more. Furthermore, there is a cost for holding on to property.

            It would make me feel better if you indicated in some way that I’m being clear. I can’t tell if you just disagree or if I’m not explaining this well enough.

          • Let’s try a thought experiment in which the linkage fee was set high enough such that a developer could only make a project pencil by paying zero dollars for land. Since no land is free, the developer could not acquire land and no development would happen. Or considering it from the seller side, any property that is generating net income is a better investment than selling the property for zero cash in return.

            The linkage fee would never be set that high, but the proposed numbers put it at somewhere in the range of 25 percent of land price for a typical project. Given we know that the extreme case would result in zero land transactions for redevelopment of housing, how can you claim that there would be no effect on land transactions under the proposed fees?

          • I appreciate what you are saying and find your argument compelling. I’m confident I fully understand the concern you are voicing.

            I’m not saying that there would be zero affect on development with linkage fees implemented. What Michael wrote indicates that evidence shows sales volume is not associated with sales price. We help ourselves if we remember that the goal of theories and narratives are to illuminate what the evidence shows. If the evidence we’re seeing gives an accurate picture, let’s figure out what narrative or theory explains this evidence.

            What I’ve been doing is trying to provide reasons for why this happens. I think Austin tells a compelling story connecting this evidence to linkage fees so I’ll quote him here:

            “As you say this is likely more of a philosophical point as it is likely to be a fairly marginal effect at the levels that are proposed. And the total effect of linkage fees seem clear to increase development as the difference between fee revenue and landowner profits is likely far greater than the amount of discouraged private development.”

            There are other compelling stories explaining the connection to this evidence and linkage fees. Perhaps the fact that linkage fees would be used to create more properties is a net positive. Perhaps before linkage fees are adopted there will be a rush to develop. Perhaps linkage fees align politically active groups, resulting in more upzones. Perhaps properties being redeveloped are always properties with large differences between current utilization and potential utilization. Perhaps the vast amount of properties being sold are sold for reasons other than those held by rational economic actors. There are many possible explanations of the evidence.

          • I don’t support linkage fees for the reasons mentioned (scattered throughout a bunch of comments here). As I see it, linkage fees are only harmless if the assumptions are correct. They may be correct most of the time, but they aren’t correct 100% of the time (as already shown). These assumptions are:

            1) The fees are reflected 100% in lowered property values.
            2) Owners of property never develop by themselves, but sell to someone else to develop.
            3) The new property owners (the ones that will develop the property) don’t see a drop in rent from the time they buy the property, to the time they actually pay the linkage fee (or develop the property).

            To me those are some pretty big exceptions that could have a negative effect on new development.

            But the linkage fees will probably have a small effect on development, and the possible benefits you quoted are reasonable. So, while I don’t like it for the stated reasons, that isn’t what bothers me most about linkage fees.

            What bothers me most about linkage fees is that it implies that the blame for high rent should be put at the feet of the developers. This is exactly the wording that the city council used when they passed this legislation. In a city that is having trouble dealing with the issue of zoning — that largely ignores the issue of zoning — this is not a good thing. All too often we have politicians or ordinary citizens who want to point blame at individuals, rather than confront errors or trade-offs in legislation (that we, the people, created). The phrase “greedy developer” is common in this city, and it completely misses the point. If I’m concerned about the cost of rent, then I want developers (greedy or not) to develop as many units as possible, with as few requirements as possible (I don’t want them to worry about providing parking). To get out of the rent crisis we are currently in, we need to build a lot more units. At best, the money raised by the linkage fees will be used to build a net handful of properties. This is good. But we have thousands and thousands of units (some in the backyards of houses) that could be built without government help if the city simply changed their zoning code.

          • Dan is being conservative in saying that the incentive fees wipe out 25% of land value. At $22/SF downtown that would be more like 40% of value. Similarly in Northgate at $9/SF. That certainly is enough of a diminution of value in residual of redevelopment to tip the scales for HBU back in favor of the existing use. One could argue that is less of an issue downtown and more of an issue in all of the urban villages.

            As for transaction volume, you are looking at the wrong data set. Pull the commercial sales (redevelopment land) over the 2005-2015 decade and you’ll see that transactions basically stopped between 2009-2011.

          • Regarding the % of land value, I’m inclined to believe the city’s study until someone presents a more rigorous examination. We’d love to publish a detailed look at this if someone is willing.

            Regarding the data set we provided, it would surprise me if the commercial property acted different from both residential and stock assets. I suppose anything is possible though. We would also be willing to publish an examination of this as well if someone can put together the data.

            One thing to keep in mind with your observations from the recession is that indicating price affects the volume of sales is different than indicating demand affects the volume of sales. A recession is by definition a period of less demand but this says nothing about the affects of price on sales. In fact, during many recessions prices drop and development doesn’t increase for sometime, usually not until market actors think demand is increasing. This is what’s seen in the graph above with residential sales.

          • I would love for someone to send me a CBA password. Or even someone to post the listing database for public research. I don’t have access to it presently.

          • The problem here is you are trying to “figure out a narrative” to explain evidence that is flimsy at best, as Chris points out.

            Michael’s conclusion in the post is this: “development fees that lower land values would not restrict land from being made available to the market.”

            But you are “not saying that there would be zero affect on development with linkage fees implemented.” Sounds like you two don’t agree. How much of an effect do you think it would have? On what is Austin basing his claim that up to a 40% hit in land value would have a “marginal” effect?

            Regarding creating more net supply, for every one multifamily project that doesn’t happen because of a linkage fee, you would need something like 25 similar projects to get built for the number of affordable units resulting from the fee to match the number of units lost in that one project. And that’s not considering the loss of production efficiency involved in non-profit housing development with funds administered by the city.

          • I personally wouldn’t call 20 years of evidence from the stock market and 16 years from the King County residential market (going through two booms and busts) ‘flimsy.’ As I mentioned before, there’s no reason to believe the commercial market behaves differently than other markets.

            With that said, I could be convinced if the commercial market evidence shows what you are saying. I would encourage you, or someone who believes you to produce the evidence showing the story you are telling. I would even be happy to put together the evidence if someone provided access. We would also be eager to publish a story with this as well.

            Until then it makes sense for a narrative to explain the available evidence, not explain-away the evidence.

          • There probably is a limit that could noticeably affect private development production. But I honestly don’t know what a 90% capital gains land tax would do. It’s so outside normality I just don’t think we have a clue how other variables would behave in that scenario. Decommissioned gas stations might be an analogy to start working through, although those often have remediation costs exceeding the “scrubbed” land value.

            But as to the limit, I think it’s related to the typical net present value of typical interim uses. Interim uses like parking lots where multi-story development is feasible have a net present value that is just a fraction of the land value (let’s say 50%). The linkage fee as proposed has two options which have been estimated to make up 3-5% of total production costs and indicates a 15-25% cost on land alone (if you buy the idea that the cost ultimately comes out of the land). If the goal is to maximize the affordable housing production without hurting market rate supply then you want linkage fee cost plus interim use net present value to be less than 100% of pre-linkage fee land cost. In the scenario I described above, 50% + ~20% is less than 100% so it’s still in the interest of the landowner to sell for development — if they are acting rationally.

            There are other interim uses that could have an 80-100% net present value of land value but these are, in my experience, rare. Improvements do not usually gracefully decline into interim use. It’s a holding strategy to minimize expense and effort, or the result of some radical physical change like a roof failure or burst pipe in an already aging property, or an upzone from single-family to multi-family. I don’t deny there are such gracefully declining properties that have quietly entered into an interim use phase out there but I doubt that they are significant.

            It’s definitely an area worth exploring and I appreciate you bringing up this point because it’s a good one but the admittedly-not-ideal market evidence available shows that if this is significant, it’s lost in the much larger churn of forces that affect property listings.

          • Michael – If I understand you correctly, based on what you are claiming about net present value of interim uses vs. land value, pretty much every interim use in the city on land zoned for high density would have already been sold for redevelopment, unless the owners were not acting rationally. But it doesn’t seem very likely to me that there are so many irrational owners.

        • Land value scales with building potential. I did not claim linkage fees are taxing land on a continuing basis, but during a sale. Thus the last sentence.

          • So, I admit, I’m confused about the linkage fees. The linkage fees are applied only when property is sold from one person to another, and is based on the potential development of the land, is that correct? In other words, what I am a developer, but also a land owner (pretty common, I would imagine). I own plenty of properties, some as parking lots, some as skyscrapers. So I buy someones parking lot that is zoned for six stories, but it is still a parking lot. I pay a linkage fee, right?

          • I think that is incorrect. The proposed fee is based on an application for development and the amount of square footage in the application. When a developer submits the land use application the square footage in the application would be used to calculate and charge the fee. Right now the legislation is only preliminary though. Here’s the official direction from the council: http://clerk.seattle.gov/~scripts/nph-brs.exe?s1=&s3=31551&s2=&s4=&Sect4=AND&l=20&Sect5=RESNY&Sect6=HITOFF&d=RESF&p=1&u=%2F~public%2Fresny.htm&r=1&f=G

          • Fair enough. Then Matt’s analysis is correct, and the question of “who pays the fee” is meaningless. Look at this way:

            I’m sitting on a parking lot. It makes good money. A builder says “Hey, you should build a six story apartment”. I do the math and determine that building that apartment only makes sense if building can be done for X and rent is Y.

            It really doesn’t matter what X is. X could be cost of lumber, or the cost of labor, or the paperwork involved. No matter what, I’m not building, unless Y matches that. So, ultimately, these fees add to X.

            Unless Y (rent) is high, I’m not building, and I keep my parking lot, which is making good money. If rent goes up, then I build, if not, then I wait. Therefore, this fee adds to the cost of rent.

            I could care less what property costs. I only want to know what rent costs. It costs too much, and is one (small) reason why.

          • If the cost to build (X) decreases, then wouldn’t projects that earn less rent become profitable, or pencil out?

            To give you an idea of where I’m going with this, the cost of land is the biggest single input for development cost in cities. So let’s say we had a way to reduce land costs across the city, wouldn’t more projects pencil out at lower rents?

          • If I’m sitting on land that is earning decent money (as a parking lot or single family home) then this won’t have any effect. If my parking lot makes sense as a parking lot until rent reaches $1,000, then it remains a parking lot (since rent is only $950). This fee won’t have any effect. But once rent reaches one grand (at the very point that I want would consider developing) it kicks in. Thus I wait until rent reaches 1,050 (which means rent really does reach $1,050).

            This will be a meaningless tax on places that would build anyway. It will be a meaningful tax on places that are right on the edge — which is where we need it most. Land may be the biggest single input for development, but there are still plenty of areas (e. g. Rainier Valley) where parking lots sit, waiting for rent to ride high enough to justify the high cost of development.

            I know this is supposed to adjust automatically, so that it doesn’t discourage growth in those edge areas, I just don’t think it can possibly do that.

          • There’s a lot of reasons why what you are describing doesn’t happen but for now I’ll say that we’ll be digging into this deeper. The evidence suggests that programs like this don’t reduce development.

          • If you are decreasing the land cost with a linkage fee then the total cost to build doesn’t change because the builder pays the linkage fee. You have some other way in mind to reduce land cost?

          • Parking lot owners don’t also tend to be developers. Developers don’t want to lock capital up that could be used for operations. So landowners and developers tend to be different groups of people. There is, however, one notable development company that is also a major landowner–Vulcan–but they have the resources (Paul Allen) to sit on large amounts of urban land.

            Other developers don’t stand to lose with linkage fees because they buy land and make something with it, not hold onto it for decades. The proposed phase-in period for linkage fees allows for short holds to not hurt developers’ bottom line.

          • So what if I buy land right before a bubble bursts? Because of the fee, the land I bought was a bit cheaper than it would be otherwise. The owner, in effect paid the fee. Fair enough, but I haven’t built yet. Right before I’m about to build, we get a rent collapse (a burst bubble). So now I have a perfectly good house/parking lot/warehouse/etc. and it no longer makes sense, financially, to build. If it was cheaper to build, or if rent was higher, I would build. As rent starts climbing up again, I consider building, but the property just sits until rent is high enough (or the fee is low enough). If it hangs in that area where rent would be high enough to justify building without the fee (but not high enough with the fee) then the building sits undeveloped for years. So this could happen with a bubble slowly deflating, as opposed to bursting.

            That is just one scenario. I described another (and you confirmed that it exists, although it isn’t common). Basically, the fee assumes a certain type of process (owner sells underutilized land to developer, who then immediately builds) when it is pretty easy to think of alternatives.

          • I think it would be helpful if you repeated back what you think I’m trying to say. I’m not 100% sure I’m being clear and you understand. So to answer your question….

            Keep in mind that what you are describing is a dynamic that could happen without a fee. A lower land value could happen simply based on market forces. That is what the graph above is showing. The value of homes are fluctuating over two booms and busts. If the story you are telling is true, evidence would show the most sales when values were the highest because the most properties would pencil out for development.

            In fact, this isn’t what you see at all. Look at the graph again and you’ll see that January 06 – 07 & January 13 – January 16 you have the highest prices on the graph and the least number of listings. Have you asked yourself why this is occurring?

            The assertion that Michael is making is that the current selling price is not the sole determinant for a property owner. Instead, the property owner is either selling for personal reason (the non-rational market actor) or is trying to maximize their properties values, waiting until the market peaks before selling (the rational econ).

            In the example you keep using, you are imagining a property owner who is waiting for the property value to exceed the rents they are earning and then they will sell. This would not be rational for two reasons. First, the value from selling is always higher than the rents someone earns. You can extend those rents out over a long period of time, months or even years and the one-time lump sum is still larger. This is because the buyer is paying for rents over a very long period of time. Second, the rational property owner is doesn’t care what the buyer is offering at the moment. They care what the current offer is in comparison to future offers. That is because if they wait to sell and get less because the market has tanked they will get a loss. Furthermore, there is a cost for holding on to property.

            The evidence supports this story. If your story was what happened you would see the blue and red line moving together.

            It would make me feel better if you indicated in some way that I’m being clear. I can’t tell if you just disagree or if I’m not explaining this well enough.

  2. Good article. I agree that the availability of housing in the market would not be drastically changed by a change in housing prices. But it seems like that is not necessarily what is at issue with linkage fees. As a tax on redevelopment what would be at issue is how many properties that are on the market would be attractive to higher density redevelopment.

    This tax will suppress some development of properties. Maybe properties will enter the market at a similar rate, but linkage fees will tilt the market towards keeping some number of those properties as single family homes rather than redeveloping them at a higher density.

    I do not presume to know if this will lead to less housing being developed in total, but it does seem seem like it will suppress some amount of private development.

    • Hi Austin,

      You bring up important points. There are essentially three options for a developer facing fees, pass on the fees to renters, pass the fees back to land owners by bidding less on land or eating the costs.

      In a previous article linked at the top of this article, Michael made the assertion that fees would be passed backwards to land owners. This article starts from that assumption.

      It sounds like you are making the argument that developers will eat the cost, making some development not pencil out. This departs from Michael’s initial assumption but is a legitimate concern, addressed in the previous article. I would suggest reading that and then looking through the comments.

      • Basically what I am saying is that linkage fees are an asymmetric tax that effects different buyers differently.

        Hypothetically say there is a large single family house for sale for $1,000,000 and the linkage fees associated with redevelopment would be $200,000. A family looking to buy a home would see the cost as $1m, a developer looking to redevelop would see the cost as $1.2m.

        The cost of the property is seen differently by a buyer that is looking to occupy a single family home. I agree with the previous article that the cost here is in the aggregate borne by the landowner, but I think additionally it will bias the market away from development.

        • Ah, I think I understand what you are saying. So a family looking to occupy a single family home would be willing to bid more for land than someone looking to redevelop the land. The land owner would be more willing to sell to the person bidding higher, who would keep the current use, discouraging redevelopment.

          The way the fee was initially laid out though was meant to take this into account. The goal is to have a fee that captures additional value above and beyond current zoning. This way it will always be marginally better to redevelop a property with higher potential utilization. The difference in land values between SF zoning and LR3 is huge, indicating there is lots of room for capture before a lower utilization becomes more valuable.

          Digging into the details of this would make another great post.

          • “The way the fee was initially laid out though was meant to take this into account. ” How so? I think you’re taking Austin’s point too narrowly. The same thing would happen with a parking lot, warehouse, 2 story building vs. 20 story building, etc. There’s a built-in incentive to build less.

          • No, there’s not a built-in incentive to build less. All types of development are subject to the same fees. Building less won’t save any money because land is priced for the best and highest use.

            The only example in which incentives could be skewed for less development is leaving the current use versus redeveloping. This is a virtually non-existent situation. Austin suggests a SF home versus an apartment but this would never happen. The only cases in which the current use would be close to a redevelopment use is when a property is already being used at or near its maximum potential. In that situation it’s very unlikely the site would be redeveloped anyhow since these aren’t typically sites that are chosen for redevelopment.

          • I think that a tax on new development necessarily will make some properties uneconomic to redevelop. It bends the curve of ‘maximum potential’ towards the existing improvements. While (an admittedly politically infeasible) land tax would bend the curve in the other direction.

            As you say this is likely more of a philosophical point as it is likely to be a fairly marginal effect at the levels that are proposed. And the total effect of linkage fees seem clear to increase development as the difference between fee revenue and landowner profits is likely far greater than the amount of discouraged private development.

    • Certainly interesting. I’m familiar with the Case Shiller Index but not with the other data set (is this REIT investment?). It looks like annual multifamily investment (if I’m reading that right) tracks the Case Shiller Index but beyond that I’m not sure if it implies a causal relationship. It would be more telling to look at home mortgage investment compared to home prices and if we saw the same relationship we could start talking about causal relationships (which drives which) but I don’t know how that relates to a homeowner’s or apartment owners’ willingness to sell. This seems to be more about the willingness to buy (or invest) than willingness to sell.

      Did you have something in mind?

      • It is an interesting chart but I agree with you Michael that is seems much more like it is tracking willingness to invest than availability of inventory to redevelop.

        Note how steeply the multifamily investment crashed in 2008/2009 when there was a large inventory of property available but little credit or willingness to invest.

Comments are closed.