There are two things we like to point out here at the Center for Land Economics: the most valuable land is concentrated in city centers, and these are precisely the areas where land is being held out of use, as surface parking lots and the like.
This is a case in which using a simplified model is actually introducing more complexity than the real world, but that's only true if you choose a specific real world example: i.e. the most important thing in deciding whether to build is like deciding whether to buy: location, location, location.
I live in Midtown NYC, which has a historically low 1.5% vacancy rate, but even that low figure is deceptively high because it includes a lot of empty ultra-luxury units (these days, all new NYC construction is considered luxury unless it's specifically built to be affordable: i.e. at Average Median Income of 160% or less). The YIMBY "formula" therefore is just "Build more, everywhere and to the greatest extent possible." And rezoning goes on all the time too, as well as buying air rights, converting existing office to housing...anything to get more floors and units. There's no question it will sell - if condos/co-ops - or rent if rentals, especially if built for moderate income people. If truly affordable, e.g. 80% AMI or less, you'll need a lottery to screen people OUT, not a broker to get people IN. If it's rent-stabilized, it's "Katie, bar the door."
Yet, even in Manhattan, we still have parking lots, not as many but still too many to be chance. Why? Because it's typical to have a parking lot paying 1/10th the property tax as a building right next door, and if a new skyscraper goes up the ratio may be even more extreme. There's an actual real world example highlighted when it was still a parking lot by two members of my local Common Ground chapter: Lindy Davies (deceased) and Mike Curtis, in this short video: https://vimeo.com/67776419?share=copy&fl=sv&fe=ci. A new luxury high rise was finally built on this 28th Street & Park Avenue(!) corner lot: the Prism at Park Avenue South Apartments. I don't know what they pay in taxes, but it's surely more than the 11-story building that used to adjoin the parking lot before adjoining the Prism.
This parking lot was notable to Lindy and Mike because it had been there for decades. In the same video they pointed out how 1-story McDonald's and similar stores were placeholders, basically paying minimal taxes for minimal land use, while profiting well for their low overhead business, until finally making a killing by selling out. This kind of low profit hold-and-sell-later-to-make-a-killing goes on all the time in NYC. Zoning and NIMBYism (some of which may be legitimate like saving our neighborhood 2-story 2-block Kip's Bay theater,have valid arguments about keeping neighborhood amenities, so there is complication that way).
But in a growing, dense city, you don't need a lot of fancy formulas, and I don't believe developers use them anyway. Given the long development timelines, interest rates and whether a developer will get permission, including rezoning, matter much more. If you build it, they will come.
There is another "real world" case I was personally involved with and led: the RiverArch. Hard as it is to believe, I gained traction on a concept arched building to span the East River between Two Bridges, Manhattan and Dumbo, Brooklyn. It would have provided 7,630 units for 19,100 residents, one of the largest residential buildings in the world, outside of a couple of megabuildings in China (they always build everything bigger, but sometimes can't fill the new spaces). No developer I talked to ever thought it was too big to fill, or too costly to construct and they generally agree with my cost projections and even profit potential, even after allowing for 30% affordable units at 60% AMI (virtually the minimum required in NYC now). We never ran fancy formulas determining build now vs. later.
Realistically, I was told by 2 principal architects from AECOM and others, it would take a minimum of three years to build once ground was broken, but it was the permission ladder and unlikelyhood of getting it that scared them off, and the eventual impossibility of acquiring a pier used as an electric field by the local utility that put a hard stop to the project.
You can read the investor summary here: https://bit.ly/RAInvestor I've written a book on the 8-year effort too, pending clarification on whom I'm allowed to quote.
Excellent piece. One other feature worth noting about the "value" of keeping a plot vacant in valuable location is that you can take out a loan against it's value and have that loan just collect interest and pay down your loan with it and any extra...you have some tax-free money. Josh Vincent and I found quite a few of these in NYC when we were doing a bunch of studies a number of years ago. Some of them with Lindy.
So basically what you’re saying is, if the interest rate on your loan + the annual holding costs is less than the expected annual growth of its value, it’s free money?
Pretty much. I forget the exact numbers but we found vacant lots in Bronx, Brooklyn and Queens worth multiple millions of dollars and owned by LLCs from out of state/town. Then Josh was clever and found a way to correlate that the properties were used as collateral on loans that seemed to be doing exactly this.
Just to clarify a bit further (trying to piece together the actual examples) -15 million dollar assessment on a vacant lot in Queens. Take a loan for $7 million. Put 3/4 of it in an account that bears interest and auto-pays off the loan and taxes. That 1/4 left is free of taxes. It’s awful.
Amen. I can walk from here to a parcel that has full infrastructure and great access, but has been held out of a growing market for decades. Zoning would permit at least 35 dwellings. Conversations with the owner show no signs of rationality. It’s just cheap and easier not to think about it. It should be taxed as if those units were there.
The option-value framing maps onto rural parcels too. In county-level listings we screen, the same lots sit for years at aspirational prices - the seller is effectively holding an option with no expiry. What finally moves them is usually the carrying costs: property tax, association dues, and the opportunity cost you describe. A per-acre price only means something against those numbers.
Great work distilling Titman's work into a digestible format. Though, it would be interesting to hear your thoughts on how this model would be applied with more complex tax structures than the Traditional Property Tax and Land Value Tax (LVT) described.
British Columbia was once equally influenced by Henry George's Progress and Poverty and utilized a pure LVT between 1910-1984 as a way to curb rampant speculation. However, it evolved to include taxation of improvements (structures) to increase its revenue. In its current framework, land value is assessed based on highest and best use + current improvement value and a mill rate is applied to that combined value. It thus begs the question of where this type of tax structure would fall between the original binary discussed by Titman and the actual incentive it forms for the land owner.
Add in layered option penalties such as the provincial Speculation and Vacancy Tax and the Empty Homes Tax in Vancouver in addition to the exorbitant Development Cost Charges and Community Amenity Charges (essentially land lift charges) and the once simple Options calculations become quite muddy.
The idea that land speculation isn’t rational because of the opportunity cost of an income stream misses one key point: using the land to generate cash flow requires actually breaking a sweat, whereas speculating allows one to capture appreciation without lifting a finger. Many (most?) people would (rationally) quit a job that pays 120k for 40 hours/week to take a job that pays 100k for under 5 hours/week. Obviously those numbers are hypothetical but they illustrate why speculation can be desirable even at the opportunity cost of immediate profits.
The point about how some idle speculation to wait until the best moment to build is a fair one because teardowns and renovations aren’t free. However, the Georgist prescription of using LVT to reduce taxes on labor/commerce/investment/buildings would significantly reduce the cost of teardowns and renovations.
Very interesting. I suspect there are more steps from here to understand the equilibrium at play.
For example, there must be price adjustments in your last example of switching to a land tax that will negate this effect—the price of land today must be such that, across the market, there is indifference between waiting and developing on marginal sites.
I have two articles that might be of interest to you and your readers on this topic.
Feels like a lot of math that doesn't actually touch on the main arguments for the inefficiency of speculation, which are the prisoner's dilemma/second mover advantage and irrationality. The math does serve the argument that property taxes are distortionary and encourage inefficient speculation, but that's about it.
Also absent in this article are the holdout/assembly problem, Tideman's "winner's curse" brought by short-sale constraints, and any analysis into financial bubbles, credit and collateralization.
Overall this feels like a deep dive analysis of a single model of speculation, rather than a broad discussion of the mechanics behind land speculation that could better elucidate when it is or isn't efficient.
That makes sense. It seemed framed more to me as response to the Less wrong post, so I found it lacking on that front. But it is a good dive into the Titman paper!
Appreciate the detailed breakdown, and the conclusion is really fascinating. I was just talking with a professor who studies inequality and he said he is still suspicious of LVT (were calling it UBE now?) because of the tendency to suppress development of land, so I will be forwarding this on to him for his thoughts.
Question:
Do zoning issues like the fact that it's probably illegal to build a profitable building (or at least massively more expensive than pure construction cost would suggest) factor in here or is zoning a separate issue?
Restrictive zoning suppresses the highest and best use for sure, which can further tilt the bias towards waiting. That said, it’s not hard to find examples of high value land with plenty of zoned capacity its not using at all.
you're making a brutal economic fallacy here confusing substitution effects (which can be about efficiency) with equity effects, which can't be.
as an analogy, imagine a woman has a necklace that is a cherished family heirloom and she refuses to sell it to you. so you impoverish her so that she's so desperate for money but she's willing to sell it. that is not a proven increase in allocative efficiency. the only way to prove an increase in net welfare (efficiency) is through a revealed Pareto improvement.
this is incredibly basic econ 101 but nevertheless I explain it here because for some reason people keep failing to understand it.
The same dynamic shows up in rural vacant land, scaled down. Raw parcels carry tiny property tax bills, often a few hundred dollars a year, so a seller can sit on an aspirational asking price for years at almost no cost. From the buyer side, that is why rural listing sites are full of stale inventory that never moves: the holding cost of waiting is nearly zero, exactly the tax tilt you describe. One difference from the urban case is that many rural owners are absentee heirs rather than investors pricing an option; they are not watching the market at all. Curious whether the option model still fits when the owner is inattentive rather than strategic.
But it was also a complicated way of saying that taxes are, in all cases, an unnatural influence on any given market. LVT pushes it in a way that appeals to many, but as the LessWrong piece describes, not without potential pitfalls.
Speculation gets a bad rap, understandably. But you didn't spend much ink discussing the value of waiting to develop and ending up with seven additional units (using the chart in the piece.) Given that these developments are costly and tend to last close to 100 years on average, locking in a development today of 5 units where, in a few years, one could develop 12 units, seems foolish, right?
I think I could develop that argument better, yes. The argument basically is that the cost to the city (in higher housing prices, greater stagnation, etc), is very bad when everyone does this. And it’s not like the developer would be giving up on profit - just maybe possibly be making less profit, and there are good reasons we shouldn’t be so scared to redevelop, and lowering taxes on buildings will further reduce that burden. I think digging further into this angle might be a good topic for its own piece as I agree I don’t really flesh it out to well here.
"and there are good reasons we shouldn’t be so scared to redevelop"
Bingo. We shouldn't be wasting valuable land out of fear for having to redevelop it later, we should be doing everything we can to lower redevelopment frictions to create a nimble and dynamic economy. That includes reducing taxes on the productive activities that go into redevelopment, reducing unnecessary regulations and procedural hurdles, investing in basic research to discover more efficient construction methods, etc. What we should not do is let land speculation drag down the economy out of inertia.
The only question I have is why vacant city lots are not used as paid parking while their speculators are waiting for better market conditions to sell the land. Paving or simply leveling and laying down gravel is a pretty low cost improvement. It allows the land owner to collect at least some passive income without creating a substantial cost to a future developer who would have to pay a contractor to remove the pavement or gravel. I suspect it’s because land speculators are looking at it purely as investment that will cash out if they wait. And they can afford to wait.
They are! We have lots of lots that become tented pop-up stores, or multiple stores. Whole outdoor seasonal malls spring up this way every year and add color and variety to our otherwise sea of highrises. They're very popular, so it's something NYC would lose if everything got converted to buildings.
Blanderization is a real issue. We lose something when under-utilized lots become steel and glass residential buildings: pop-up stores, green gardens (officially or not), plazas, even sculpture gardens like the one downtown that after years of protest was made into a permanent park instead of an affordable senior housing complex.
But we also have record homelessness and housing unaffordability. So, the tension over what to do with valuable land will always be with us, beyond what LVT and best-and-highest-use arguments can resolve.
This is a case in which using a simplified model is actually introducing more complexity than the real world, but that's only true if you choose a specific real world example: i.e. the most important thing in deciding whether to build is like deciding whether to buy: location, location, location.
I live in Midtown NYC, which has a historically low 1.5% vacancy rate, but even that low figure is deceptively high because it includes a lot of empty ultra-luxury units (these days, all new NYC construction is considered luxury unless it's specifically built to be affordable: i.e. at Average Median Income of 160% or less). The YIMBY "formula" therefore is just "Build more, everywhere and to the greatest extent possible." And rezoning goes on all the time too, as well as buying air rights, converting existing office to housing...anything to get more floors and units. There's no question it will sell - if condos/co-ops - or rent if rentals, especially if built for moderate income people. If truly affordable, e.g. 80% AMI or less, you'll need a lottery to screen people OUT, not a broker to get people IN. If it's rent-stabilized, it's "Katie, bar the door."
Yet, even in Manhattan, we still have parking lots, not as many but still too many to be chance. Why? Because it's typical to have a parking lot paying 1/10th the property tax as a building right next door, and if a new skyscraper goes up the ratio may be even more extreme. There's an actual real world example highlighted when it was still a parking lot by two members of my local Common Ground chapter: Lindy Davies (deceased) and Mike Curtis, in this short video: https://vimeo.com/67776419?share=copy&fl=sv&fe=ci. A new luxury high rise was finally built on this 28th Street & Park Avenue(!) corner lot: the Prism at Park Avenue South Apartments. I don't know what they pay in taxes, but it's surely more than the 11-story building that used to adjoin the parking lot before adjoining the Prism.
This parking lot was notable to Lindy and Mike because it had been there for decades. In the same video they pointed out how 1-story McDonald's and similar stores were placeholders, basically paying minimal taxes for minimal land use, while profiting well for their low overhead business, until finally making a killing by selling out. This kind of low profit hold-and-sell-later-to-make-a-killing goes on all the time in NYC. Zoning and NIMBYism (some of which may be legitimate like saving our neighborhood 2-story 2-block Kip's Bay theater,have valid arguments about keeping neighborhood amenities, so there is complication that way).
But in a growing, dense city, you don't need a lot of fancy formulas, and I don't believe developers use them anyway. Given the long development timelines, interest rates and whether a developer will get permission, including rezoning, matter much more. If you build it, they will come.
There is another "real world" case I was personally involved with and led: the RiverArch. Hard as it is to believe, I gained traction on a concept arched building to span the East River between Two Bridges, Manhattan and Dumbo, Brooklyn. It would have provided 7,630 units for 19,100 residents, one of the largest residential buildings in the world, outside of a couple of megabuildings in China (they always build everything bigger, but sometimes can't fill the new spaces). No developer I talked to ever thought it was too big to fill, or too costly to construct and they generally agree with my cost projections and even profit potential, even after allowing for 30% affordable units at 60% AMI (virtually the minimum required in NYC now). We never ran fancy formulas determining build now vs. later.
Realistically, I was told by 2 principal architects from AECOM and others, it would take a minimum of three years to build once ground was broken, but it was the permission ladder and unlikelyhood of getting it that scared them off, and the eventual impossibility of acquiring a pier used as an electric field by the local utility that put a hard stop to the project.
You can read the investor summary here: https://bit.ly/RAInvestor I've written a book on the 8-year effort too, pending clarification on whom I'm allowed to quote.
Excellent piece. One other feature worth noting about the "value" of keeping a plot vacant in valuable location is that you can take out a loan against it's value and have that loan just collect interest and pay down your loan with it and any extra...you have some tax-free money. Josh Vincent and I found quite a few of these in NYC when we were doing a bunch of studies a number of years ago. Some of them with Lindy.
So basically what you’re saying is, if the interest rate on your loan + the annual holding costs is less than the expected annual growth of its value, it’s free money?
Pretty much. I forget the exact numbers but we found vacant lots in Bronx, Brooklyn and Queens worth multiple millions of dollars and owned by LLCs from out of state/town. Then Josh was clever and found a way to correlate that the properties were used as collateral on loans that seemed to be doing exactly this.
Wow that’s insane 🤯
Just to clarify a bit further (trying to piece together the actual examples) -15 million dollar assessment on a vacant lot in Queens. Take a loan for $7 million. Put 3/4 of it in an account that bears interest and auto-pays off the loan and taxes. That 1/4 left is free of taxes. It’s awful.
Amen. I can walk from here to a parcel that has full infrastructure and great access, but has been held out of a growing market for decades. Zoning would permit at least 35 dwellings. Conversations with the owner show no signs of rationality. It’s just cheap and easier not to think about it. It should be taxed as if those units were there.
The option-value framing maps onto rural parcels too. In county-level listings we screen, the same lots sit for years at aspirational prices - the seller is effectively holding an option with no expiry. What finally moves them is usually the carrying costs: property tax, association dues, and the opportunity cost you describe. A per-acre price only means something against those numbers.
Great work distilling Titman's work into a digestible format. Though, it would be interesting to hear your thoughts on how this model would be applied with more complex tax structures than the Traditional Property Tax and Land Value Tax (LVT) described.
British Columbia was once equally influenced by Henry George's Progress and Poverty and utilized a pure LVT between 1910-1984 as a way to curb rampant speculation. However, it evolved to include taxation of improvements (structures) to increase its revenue. In its current framework, land value is assessed based on highest and best use + current improvement value and a mill rate is applied to that combined value. It thus begs the question of where this type of tax structure would fall between the original binary discussed by Titman and the actual incentive it forms for the land owner.
Add in layered option penalties such as the provincial Speculation and Vacancy Tax and the Empty Homes Tax in Vancouver in addition to the exorbitant Development Cost Charges and Community Amenity Charges (essentially land lift charges) and the once simple Options calculations become quite muddy.
The idea that land speculation isn’t rational because of the opportunity cost of an income stream misses one key point: using the land to generate cash flow requires actually breaking a sweat, whereas speculating allows one to capture appreciation without lifting a finger. Many (most?) people would (rationally) quit a job that pays 120k for 40 hours/week to take a job that pays 100k for under 5 hours/week. Obviously those numbers are hypothetical but they illustrate why speculation can be desirable even at the opportunity cost of immediate profits.
The point about how some idle speculation to wait until the best moment to build is a fair one because teardowns and renovations aren’t free. However, the Georgist prescription of using LVT to reduce taxes on labor/commerce/investment/buildings would significantly reduce the cost of teardowns and renovations.
Very interesting. I suspect there are more steps from here to understand the equilibrium at play.
For example, there must be price adjustments in your last example of switching to a land tax that will negate this effect—the price of land today must be such that, across the market, there is indifference between waiting and developing on marginal sites.
I have two articles that might be of interest to you and your readers on this topic.
First, on switching property tax to land tax
https://www.fresheconomicthinking.com/p/was-henry-george-wrong-about-land?utm_source=publication-search
Second, on the economic rationale of landbanking (holding vacant land)
https://www.fresheconomicthinking.com/p/what-is-landbanking-and-why-does
Feels like a lot of math that doesn't actually touch on the main arguments for the inefficiency of speculation, which are the prisoner's dilemma/second mover advantage and irrationality. The math does serve the argument that property taxes are distortionary and encourage inefficient speculation, but that's about it.
Also absent in this article are the holdout/assembly problem, Tideman's "winner's curse" brought by short-sale constraints, and any analysis into financial bubbles, credit and collateralization.
Overall this feels like a deep dive analysis of a single model of speculation, rather than a broad discussion of the mechanics behind land speculation that could better elucidate when it is or isn't efficient.
This was originally conceived as a deep dive into the Titman paper, but I could always write a follow up article
That makes sense. It seemed framed more to me as response to the Less wrong post, so I found it lacking on that front. But it is a good dive into the Titman paper!
Appreciate the detailed breakdown, and the conclusion is really fascinating. I was just talking with a professor who studies inequality and he said he is still suspicious of LVT (were calling it UBE now?) because of the tendency to suppress development of land, so I will be forwarding this on to him for his thoughts.
Question:
Do zoning issues like the fact that it's probably illegal to build a profitable building (or at least massively more expensive than pure construction cost would suggest) factor in here or is zoning a separate issue?
Restrictive zoning suppresses the highest and best use for sure, which can further tilt the bias towards waiting. That said, it’s not hard to find examples of high value land with plenty of zoned capacity its not using at all.
makes sense, thank you
you're making a brutal economic fallacy here confusing substitution effects (which can be about efficiency) with equity effects, which can't be.
as an analogy, imagine a woman has a necklace that is a cherished family heirloom and she refuses to sell it to you. so you impoverish her so that she's so desperate for money but she's willing to sell it. that is not a proven increase in allocative efficiency. the only way to prove an increase in net welfare (efficiency) is through a revealed Pareto improvement.
this is incredibly basic econ 101 but nevertheless I explain it here because for some reason people keep failing to understand it.
https://clayshentrup.medium.com/more-buildings-is-not-more-productive-185274062040
The same dynamic shows up in rural vacant land, scaled down. Raw parcels carry tiny property tax bills, often a few hundred dollars a year, so a seller can sit on an aspirational asking price for years at almost no cost. From the buyer side, that is why rural listing sites are full of stale inventory that never moves: the holding cost of waiting is nearly zero, exactly the tax tilt you describe. One difference from the urban case is that many rural owners are absentee heirs rather than investors pricing an option; they are not watching the market at all. Curious whether the option model still fits when the owner is inattentive rather than strategic.
Okay, that was a great piece.
But it was also a complicated way of saying that taxes are, in all cases, an unnatural influence on any given market. LVT pushes it in a way that appeals to many, but as the LessWrong piece describes, not without potential pitfalls.
Speculation gets a bad rap, understandably. But you didn't spend much ink discussing the value of waiting to develop and ending up with seven additional units (using the chart in the piece.) Given that these developments are costly and tend to last close to 100 years on average, locking in a development today of 5 units where, in a few years, one could develop 12 units, seems foolish, right?
I think I could develop that argument better, yes. The argument basically is that the cost to the city (in higher housing prices, greater stagnation, etc), is very bad when everyone does this. And it’s not like the developer would be giving up on profit - just maybe possibly be making less profit, and there are good reasons we shouldn’t be so scared to redevelop, and lowering taxes on buildings will further reduce that burden. I think digging further into this angle might be a good topic for its own piece as I agree I don’t really flesh it out to well here.
"and there are good reasons we shouldn’t be so scared to redevelop"
Bingo. We shouldn't be wasting valuable land out of fear for having to redevelop it later, we should be doing everything we can to lower redevelopment frictions to create a nimble and dynamic economy. That includes reducing taxes on the productive activities that go into redevelopment, reducing unnecessary regulations and procedural hurdles, investing in basic research to discover more efficient construction methods, etc. What we should not do is let land speculation drag down the economy out of inertia.
The only question I have is why vacant city lots are not used as paid parking while their speculators are waiting for better market conditions to sell the land. Paving or simply leveling and laying down gravel is a pretty low cost improvement. It allows the land owner to collect at least some passive income without creating a substantial cost to a future developer who would have to pay a contractor to remove the pavement or gravel. I suspect it’s because land speculators are looking at it purely as investment that will cash out if they wait. And they can afford to wait.
I mean, most of them are doing exactly this, at least in city centers
They are! We have lots of lots that become tented pop-up stores, or multiple stores. Whole outdoor seasonal malls spring up this way every year and add color and variety to our otherwise sea of highrises. They're very popular, so it's something NYC would lose if everything got converted to buildings.
Blanderization is a real issue. We lose something when under-utilized lots become steel and glass residential buildings: pop-up stores, green gardens (officially or not), plazas, even sculpture gardens like the one downtown that after years of protest was made into a permanent park instead of an affordable senior housing complex.
But we also have record homelessness and housing unaffordability. So, the tension over what to do with valuable land will always be with us, beyond what LVT and best-and-highest-use arguments can resolve.