Aug 14, 2023·edited Aug 14, 2023Liked by Tim Helm, Cameron Murray
Great piece as usual, love the literal tree analogy being used but from someone else.
I think the issue comes back to not the logic of the natural speed limit irrespective of planning controls, but what happens when land is upzoned. There's this weird idea that value of land and the development sites themselves fall upon mass rezoning (I have yet to see convincing evidence from Auckland or elsewhere). In essence saying, currently your plot of land is restricted to a legal tree density of e.g. 1 tree per 10 m2. But if we relaxed that to say 1 per 5 m2 or even abolished the legal limit, what would it do land values and the market for trees?
Of course, the same interest rate logic applies. And as pointed out, you can permit no trees on some land and more trees on other land. End result should be the same. Because it's never been about the amount of trees per plot, but rather the economic constraint binds the quantity of trees in total. And such density constraints could only be effective if they were so strong as to be binding and prevent the quantity of trees reaching the equilibrium with interest rates across the entire market.
Perhaps that's what they believe is happening? But that's the dubious part... That planning is a binding constraint on citywide housing supply.
Interesting piece and it is painful how many in the housing debate do not even seem to want to acknowledge this mechanism. One thing that I wonder about though is whether it matter that the housing market, specifically those who choose to build, are diverse. Big developers certainly landbank, however, there are also 'mum-and-pap'/'leech-off-your-neighbour' landlords (the richer ones) who build 6-10 unit apartment blocks (at least in Brisbane), and I would expect they do not have the same insights into the market and do operate on the same time scale as large scale developers (e.g. they might delay 1-2 years at most). If correct and there are enough of them, they could create downward price pressure, and in turn disincentive bigger developers to landbank. Kinda like a failure of the commons, but by the land-owning class, and to the benefit of the rest of us. Is anything known about this?
I've wondered the same. For policy purposes, I've wondered if concentration of landownership, which is most prevalent at the urban fringe, matters for the rate of housing supply. Could we induce faster building by dispersing the ownership of land zoned for housing (such as by rezoning more land)?
I've never seen any empirical evidence that concentration matters.
There's also one important factor linking small and large developers together and constraining both of them: the capacity of the construction sector. Lots of arguments about zoning and supply ignore this constraint.
Building involves numerous skilled professions, and the industry can't expand fast. Small and large developers are competing for the same construction resource.
That means if a small and impatient developer wanted to build faster they'd have to bid resources off a more strategically-minded developer. The short run supply of new housing wouldn't change. In the longer run, the pressure by small developers to build faster would inflate construction costs a bit more than otherwise, inducing the industry to expand faster than otherwise. But this pressure would also raise development costs today relative to development costs tomorrow, which would further incentivise delay by strategic developers.
So we actually have two forces that regulate the supply of new housing: (1) house prices, i.e. oversupplying housing today relative to the equilibrium rate means selling at a discount relative to what could be achieved tomorrow, and (2) development costs, i.e. oversupply housing today relative to the equilibrium rate means developing at a premium cost relative to what could be achieved tomorrow.
I presume we're already experiencing an equilibrium where all these forces are in balance.
On your point about small landlords/developers, remember that anyone who owns these sites will also be the most patient owners. The most patient buyers buy property (can pay the most) and impatient owners sell.
Crampton is the chief economist for a right wing lobby group that used to be known as the Business Roundtable.
Is he demonstrating that economic theory can be produced to serve the differing needs of different supporters and members of this lobby group? A sort of market in justification for what supporters and members want to do?
Andrew Riddell, if there are flaws in Eric Crampton's arguments, the challenge is to find them directly, rather than infer them from his employer.
Every single person comes with ideological baggage and is subtly shaped by (or selected for the role by) the incentives of their employer.
Trying to cancel people from debate on the basis that you don't like their employer is, one, a poor basis for constructive debate, and two, an admission you can't find the actual flaws in their argument.
I was of course responding to the article by Cameron Murray setting out two different views by the same person. Pointing out some implications of that is hardly trying to cancel people because I don't like their employer. But whatever.
Andrew, if I'm wrong about something, it's because I'm genuinely wrong. You have all these crazy conspiracy theories. Ask anyone who's ever tried to be my manager. I don't take the kind of direction you're here implying.
Hotelling pricing tells you that the path of a nonrenewable commodity price over time will follow interest rates and expectations about changes in tech/costs. Whenever present prices look high, relative to that path, people bring forward extraction/sale; whenever present prices look low, relative to that path, people shift outward extraction/sale.
And so you don't expect people to do something dumb like dump a pile of the resource onto the market when prices are low if they expect prices to be sufficiently higher in future. NZ's climate commission basically expects forest owners here to do that: to dump a pile of credits on the market in the 2030s, and then everyone faces rapid escalation in carbon prices a little while afterwards when those credits run out. But that doesn't make sense - forest owners would hold credits for future sale instead. Or someone else would buy the credits from them and hold.
Now consider housing. You have a pile of sites in town that are forbidden against intensifying, and a pile of sites out of town that are forbidden against transforming from paddock to subdivision. Land prices across the entire urban gradient are artificially inflated.
If you ease zoning so houses can be built in more places, the scarcity rents built into zoning have to dissipate over time. Because zoning stops being as relevant a constraint. Selling earlier rather than later means capturing at least some of that, before the decline. Paddocks turn into subdivisions; those houses compete with downtown properties; downtown land prices are pushed down. Recall that this is a transition to a new lower price path. Holding in hope of some future price increase won't help.
Isn't the simple direct response to the argument that housing isn't a nonrenewable commodity? You build a house, it pays out imputed rents in perpetuity (with some upkeep of course). The value of a house today is a perpetuity of its rent values today, and the value of it ten years is the same payments but starting 10 years from now. There are some frictions between the land vendors and builder-developers in terms of financing, political savvy, and differing valuations, but you'd expect some coasian bargain to occur so that land wouldn't sit underused since the NPV of the perpetuity today is greater than the NPV starting n years from now.
If you wash up on a desert island with a gallon of oil and a bunch of building supplies, you're going to be careful portioning out your oil over your lifetime, but your not going to hold-off erecting your shelter structures: you want to get as many years out of it as you can.
This all is built on the assumption that there is enough land in the area, enough different owners (no monopoly/cartel), and enough groups that are able/willing to make use of upzoning changes (owner-occupiers have no intention to upgrade) to drive prices down. I do not expect this to be true for all cities. If land is too scarce, prices will still go up, meaning delaying remains a valid option for developpers. If almost anything is owned by owner occupiers, why would they upzone? If you expect owners to be bought out, this would likely mean land prices increased (else why would they leave), in turn providing justification for higher rents/house prices . If there are a few major landlord corporations that influence a significant part of the market, they have the possibility to play the long game again. I see how Murray's analysis might be too simple, but it is still more nuanced than the YIMBY typical story is.
However I think the dichotomy between "investors and land speculators willing to sell their land for development if the price is right" and "owner occupiers unable to be induced to sell at any price" is too stark.
The option to sell up, quit town, retire to the regions, and watch your old villa demolished and replaced by townhouses is perpetually available to owner-occupiers.
When prices rise enough, the capital gains are attractive enough to induce a lot of this.
In NZ's and Australia's highest priced cities (Auckland and Sydney), we see larger net internal migration outflows of citizens than in other cities. Growing pains might explain this. But so might the attractive financial return on leaving the city and buying in the regions.
The upshot?
The idea that the current owners' preferences might effectively constrain the amount of zoned land able to be built on such that zoning more could relax a hard constraint is not, I think, that relevant. When the price is high enough everyone is a seller.
Thanks for the elaboration! As non-economist looking with an interest in this, it is hard to get the nuance without feedback. In regards to the last point, I agree everyone is a seller if the price is high enough. The concern is that if prices rise 'enough', this goes against the goal of the policy, as the goal was to bring down (linked) prices. Combined with intertia (ie. people prefer neighbours they know) and cost of transition (moving house, demolition, construction, etc), the required increase in land prices to free them up land might mean new more denser housing ends up costing the same as the old houses.
With one exception: old wooden trees used as carbon sinks are beautiful while old wooden houses used as carbon sinks are ugly racist symbols of a shameful colonial past.
At least that's what we believe here in Wellington :)
You've assumed your conclusion. That's the flaw in your argument that trees and houses are different.
Take your comment above:
"land prices across the entire urban gradient are artificially inflated..."
"If you ease zoning, the scarcity rents built into zoning..."
Without the assumption of "artificial inflation" and "scarcity rents from zoning" your argument falls over.
You say that SINCE land prices are high because of regulation, removing regulation will make land prices fall, and once people recognise this, they will respond by building more, faster.
Replace the SINCE with IF and your statement is correct.
But it's a big IF. Your argument is based on the premise that regulation is currently making land prices higher than they would otherwise be.
What justifies this premise?
As best I'm aware, you and others believe this solely on the basis of AMM modelling in which population is assumed fixed and all developable land is assumed to be fully developed. In this model world, regulatory restrictions on physical space for building mechanically and necessarily raise land prices. This model world is highly unrealistic as an explanation for current land prices, for many reasons.
Am I wrong? Do you have other theoretical arguments for why you think zoning is increasing land prices?
The key point to recognise about zoning regulations is that they're binding on density at the site level, but not binding on the rate of new supply at the market level.
The state regulates site density, the market regulates new housing supply.
If zoning doesn't bind new supply, it doesn't raise house prices above some counterfactual. The market sets prices. The price of land flows from it. So the price of land is no higher due to zoning.
I see no contradiction in Eric's two articles. I think Tim correctly diagnosed the critical juncture: the claim that excessive land regulation causes urban land prices to be much too high. (Tim asks what those causes are; I'll come back to this.)
If relieving those excessive regulatory constraints causes trend decline in urban land prices towards normal levels, then the real option value to delay gets very small, and investment timing will be brought forward. Or conversely, relieving constraints allows superabundance of land supply, then the real option value reduces, which would induce strong steady real (ie, not speculative financial) investment. This would cause trend decline in urban land prices until they leveled off at normal levels.
Normal levels is farm land price (which is a mere $2 to $4 per m2 to buy, not rent, in New Zealand) plus the opportunity cost of fringe development and infrastructure (including a real option value for irreversible investment), plus a premium for inner city living because of savings on transport costs.) At normal levels then housing investment might look a lot more like the mining example. Note, urban areas are only about 0.9% of NZ land area, and expand at half the rate of population increase (0.7% compared to 1.5% pa) so are hardly "sprawling".
What are the regulatory constraints that cause excessive urban land prices?
Chiefly, constraints on the competitive supply of fringe urban land across all cities as a system. These are: geography, regulation (eg urban growth boundaries, forbidding of leapfrogging to cheap land), a lack of public infrastructure (because we removed revenue bonds in 1989 and 1995); and a lack of long-term planning to reduce future costs since the 1980s (Solly Angel style).
This causes land price appreciation, and there are many feedbacks that cause this to escalate and be sustained for many decades. (Eg, delayed real investment, speculative financial investment, suppressed public infrastructure supply and increasing regulatory restrictions, genuine financial investment (because housing returns become unduly sensitive to interest rates and become less positively correlated with other investments in a portfolio), increasing aversion to local upzoning by homevoters etc.)
For more, see these articles in the public domain (not exhaustive). Note the concepts are not reliant on AMM modelling Tim, although I find it helpful to use the AMM to start to analyse them. I'd like to use real options analysis one day, but it's really hard. I admire Cam's advocacy of real options analysis, which inspired my study of them.
First, I didn’t ask about the CAUSES of land prices being too high. I asked for EVIDENCE that zoning causes land prices to be too high. Evidence means two things: theory linking zoning and prices, and empirics that falsify other theories of prices.
There’s a general methodological point here.
It sounds like your approach to inquiry is to ask “first, are land prices too high, and if so, why?”.
If that’s the case, then your approach puts the cart before the horse. It will lead you to unfalsifiable claims about cause and effect. This is unscientific. Your claims will be unfalsifiable because you’ve already found empirics that prove (to your satisfaction) that these effects are present.
Specifically, one cannot judge whether land prices are “too high” without a theory of what determines land prices.
If you do so via casual empiricism stripped of theory, you’ll be misled. Just as patterns in nature can look like intelligent design when we have no theory of evolution, so too can urban fringe differentials or prices in Houston look like proof of growth boundaries driving up urban prices when we have no clear theory of what urban prices would be without the boundaries.
Searching for causes of why land prices are too high before having a theory of land prices means searching for unfalsifiable claims.
Short version: I’m drilling you and Eric for clear theory of why you claim zoning rules are driving land prices higher. All I see is closed-city AMM models. Do you have any other theory?
*****
Second, I think we agree on the mechanism by which land prices WOULD come down IF current zoning were causing them to be higher than they would be under looser zoning.
That mechanism is expectations and investment timing (a lower value on delay).
Do we agree that the crux of disagreement is the premise in this argument?
1. Tighter zoning causes land prices to be higher than if zoning were looser [premise]
2. Upzoning would change price expectations and investment timing [mechanism]
3. Via mechanism (2) upzoning would lower land prices [conclusion]
*****
Third, we agree that anyone using the argument above to justify the premise in (1) is using circular reasoning. Premise (1) needs to be justified independently. You can’t prove premise (1) using the conclusion of an argument that relies on premise (1).
Real options logic is just a mechanism. It’s interesting but not complete as an argument for conclusion (3). You can only reach conclusion (3) with premise (1). That’s why I ask for your theory to justify premise (1).
Eric’s arguments read as circular. He takes premise (1) as given, talks through mechanism (2), then reaches conclusion (3) without offering independent support for premise (1). (To be clear, he might have independent reasons for premise (1), but hasn’t expressed them in anything I’ve seen.)
Short version, do we agree you can’t prove zoning is driving land prices higher via real options logic?
*****
Fourth, while I generally agree with your real options reasoning (mechanism 2), there’s something you’ve forgotten about when applying that reasoning that derails your claim that a “rush to the door” is possible by developers seeking to sell homes and land before they fall in price after upzoning.
That something is construction capacity.
In the short run, capacity is fixed. In the medium run, it’s flexible. But between short and medium run, any “rush to the door” will alter the relative cost of present and future construction in a way that halts the rush.
Specifically, any developer building more today will have to bid fixed construction resources away from other developers. That will raise construction costs. That will induce entry. That will make tomorrow’s costs lower than today’s elevated costs. And that will increase the real option value to delay.
That means if some developers build more today, other developers will build less today.
Do you understand this argument?
(The same logic applies to the effect of current development on current sales prices relative to future prices).
These two factors cause market forces to “regulate” the rate of new housing supply. Sales prices and construction costs are automatic stabilisers. They affect incentives to develop vs delay in ways that regulate/stabilise the rate of construction and sales.
You can alter this single market equilibrium, but you’ll never reach an “escape velocity” in the rate of development that gets you to some more desirable equilibrium. There’s one equilibrium, not two.
BTW I’m yet to see the endogeneity of current and future construction costs embedded in a mathematical real options model of development, but you can see from my verbal reasoning how such a model would work.
*****
Fifth, recall the efficient markets hypothesis?
It says that all information is already contained in the price. It applies to financial asset prices. Land is a financial asset.
If you’re comfortable with the EMH, think about its key implication – that land prices will fall immediately once smart money (e.g. landbankers) think the policy settings are right for sustained decline.
It won’t take years of actual building. Land is infinitely lived, so most of its present value comes from distant future rents. This means credible policy will crash land prices immediately.
Yet this never happens. Tripling zoned capacity in Auckland from 30 years’ to 90 years’ supply didn’t do it. And if a hefty dose of the medicine did not budge the needle why do you suppose more medicine will suddenly work? Do you have theories of tipping points, thresholds?
Also why do developers and even developer LOBBY GROUPS push for this?
You can posit co-ordination failure when individual landowners push for zoning changes in their own interest but not the collective landowner interest. But you can’t plausibly claim co-ordination failure when groups existing precisely to co-ordinate interests are pushing for this.
If all the smart money wants zoning change that you claim will destroy their asset values, have you been fooled or have they?
On 1, I agree theory is needed before interpreting data in cities.
Ed Glaeser said this p117 in his 2008 book ‘Cities, Agglomeration and Spatial Equilibrium’: “Understanding cities requires understanding at least three types of decision-making agents: individuals who are choosing where to live and work, firms that are choosing where to operate and how many people to employ and developers who are deciding how much to build. This perspective is part of a broader view that the economic approach to understanding cities can never use empirics without theory. I believe that every correlation or regression investigated by an economist should be justified, if only in the economists’ head, by some kind of formal model that starts with decision-making agents.”
I presented at this year’s 2023 NZ Association of Economists (NZAE) annual conference on the work the Housing Technical Working Group (the Treasury, Reserve Bank of NZ, and the Ministry of Housing and Urban Development) is doing on land efficiency indicators. The gist is market operations free of frictions should lead to continuous prices across all margins of adjustment (net of fixed costs) or close to that if there are lumpy increments such as stories. Those margins of adjustment are numerous (eg, floor area vs garden space; distance from city centre; residential vs business use; rural-urban). However, this won’t maximise social welfare if there are externalities that aren’t internalised through pricing, landownership etc. Assessing allocative efficiency also requires shadow pricing of those wider effects omitted from market price signals. Research papers on the theory and metrics will likely be forthcoming.
On 2, I don’t agree with premise 1 in all cases, but in some cases I do. In my spatial equilibrium theory and modelling of cities, if a city fails to accommodate people well (by imposing high costs to intensify) then fewer people will live in the city, which bids down its differential land rents. This accords with the Henry George Theorem I’m developing for an open city in spatial equilibrium. (I presented on this work at the 2022 NZAE conference.)
On 3, I wouldn’t start with those kinds of premises, as they breach Glaeser’s principle I quote above. I’d start with the objective functions of households (utility maximisation), firms (profit maximisation), and housing developers (profit maximisation), and perhaps (if I’m researching local public policy) local governments (supposing their objective is differential land rents maximisation).
On 4, I agree construction sector capacity is an important constraint on the rate of house building. I think it determines house rents, which are contemporaneous. House prices are future looking, as you outline in your point 5, so they can respond much more abruptly. I think construction sector capacity can evolve differently over the long-run depending on if it faces a steady large flow of work vs volatile low flows of work.
On 5, I agree the credibility of policy is very important, and can drives prices and investment behaviours in the short-term. Be careful with the word “never”. There is some historic literature that land prices dropped precipitously from the late 1800s to the mid-1900s, and this was associated with globalisation and technology revolution. (Eg Louis Winnick’s 1953 PhD thesis “Wealth Estimates for Residential Real Estate, 1890-1950”.) That’s what Thomas Picketty documented too in his 2014 book “Capital in the 21st Century”: massive wealth in land (urban and rural) in the 1700s and 1800s, reducing to a minimum in the three decades after WWII, and now urban land prices are ballooning again.
Well I never an economic logic arguing both sides of the same argument. Hang on wait the aerodynamics "theory of lift", and the theory of everything else....... Perhaps we still need to temper this theorising with practical demonstration to actually achieve any change...
Whether reducing obstacles to building new housing in particular jurisdiction will reduce the (quality adjusted) price of housing is an interesting issue. It is however not central to the policy decision about how to regulate construction. Suppose in the new regime only a lot of "luxury" houses/apartments are built for people who lives elsewhere before and the average price of housing rises. That would still be a desirable outcome for the liberalizing jurisdiction.
While I appreciate the theoretical argument, in the real world which is messy, individual choices will be different. The cheap criteria is the ability to carry assets based on current cash flows and amount of debt being carried.
As you can see from current interest rises, first home buyers who bought in the last ~2 years are starting to dump property. While property owners with multiple properties and/or with much lower LVRs are largely unconcerned (so far).
While this is tangential to the main thrust of the article, individual circumstances matter. Also future expected values are volatile (more so now because the expectation itself of property values always goes up is no longer so true) so there is a bias to what’s in it for me now. Far easier to sell a pill to lose weight than a 12month program for diet, exercise and bring a boring person! 😜
Great piece as usual, love the literal tree analogy being used but from someone else.
I think the issue comes back to not the logic of the natural speed limit irrespective of planning controls, but what happens when land is upzoned. There's this weird idea that value of land and the development sites themselves fall upon mass rezoning (I have yet to see convincing evidence from Auckland or elsewhere). In essence saying, currently your plot of land is restricted to a legal tree density of e.g. 1 tree per 10 m2. But if we relaxed that to say 1 per 5 m2 or even abolished the legal limit, what would it do land values and the market for trees?
Of course, the same interest rate logic applies. And as pointed out, you can permit no trees on some land and more trees on other land. End result should be the same. Because it's never been about the amount of trees per plot, but rather the economic constraint binds the quantity of trees in total. And such density constraints could only be effective if they were so strong as to be binding and prevent the quantity of trees reaching the equilibrium with interest rates across the entire market.
Perhaps that's what they believe is happening? But that's the dubious part... That planning is a binding constraint on citywide housing supply.
Interesting piece and it is painful how many in the housing debate do not even seem to want to acknowledge this mechanism. One thing that I wonder about though is whether it matter that the housing market, specifically those who choose to build, are diverse. Big developers certainly landbank, however, there are also 'mum-and-pap'/'leech-off-your-neighbour' landlords (the richer ones) who build 6-10 unit apartment blocks (at least in Brisbane), and I would expect they do not have the same insights into the market and do operate on the same time scale as large scale developers (e.g. they might delay 1-2 years at most). If correct and there are enough of them, they could create downward price pressure, and in turn disincentive bigger developers to landbank. Kinda like a failure of the commons, but by the land-owning class, and to the benefit of the rest of us. Is anything known about this?
Great question, MrJP.
I've wondered the same. For policy purposes, I've wondered if concentration of landownership, which is most prevalent at the urban fringe, matters for the rate of housing supply. Could we induce faster building by dispersing the ownership of land zoned for housing (such as by rezoning more land)?
I've never seen any empirical evidence that concentration matters.
There's also one important factor linking small and large developers together and constraining both of them: the capacity of the construction sector. Lots of arguments about zoning and supply ignore this constraint.
Building involves numerous skilled professions, and the industry can't expand fast. Small and large developers are competing for the same construction resource.
That means if a small and impatient developer wanted to build faster they'd have to bid resources off a more strategically-minded developer. The short run supply of new housing wouldn't change. In the longer run, the pressure by small developers to build faster would inflate construction costs a bit more than otherwise, inducing the industry to expand faster than otherwise. But this pressure would also raise development costs today relative to development costs tomorrow, which would further incentivise delay by strategic developers.
So we actually have two forces that regulate the supply of new housing: (1) house prices, i.e. oversupplying housing today relative to the equilibrium rate means selling at a discount relative to what could be achieved tomorrow, and (2) development costs, i.e. oversupply housing today relative to the equilibrium rate means developing at a premium cost relative to what could be achieved tomorrow.
I presume we're already experiencing an equilibrium where all these forces are in balance.
Agree that this mechanism is completely ignored.
On your point about small landlords/developers, remember that anyone who owns these sites will also be the most patient owners. The most patient buyers buy property (can pay the most) and impatient owners sell.
Thank you!
Crampton is the chief economist for a right wing lobby group that used to be known as the Business Roundtable.
Is he demonstrating that economic theory can be produced to serve the differing needs of different supporters and members of this lobby group? A sort of market in justification for what supporters and members want to do?
Andrew Riddell, if there are flaws in Eric Crampton's arguments, the challenge is to find them directly, rather than infer them from his employer.
Every single person comes with ideological baggage and is subtly shaped by (or selected for the role by) the incentives of their employer.
Trying to cancel people from debate on the basis that you don't like their employer is, one, a poor basis for constructive debate, and two, an admission you can't find the actual flaws in their argument.
I was of course responding to the article by Cameron Murray setting out two different views by the same person. Pointing out some implications of that is hardly trying to cancel people because I don't like their employer. But whatever.
Andrew, if I'm wrong about something, it's because I'm genuinely wrong. You have all these crazy conspiracy theories. Ask anyone who's ever tried to be my manager. I don't take the kind of direction you're here implying.
Ok, where to start here.
Hotelling pricing tells you that the path of a nonrenewable commodity price over time will follow interest rates and expectations about changes in tech/costs. Whenever present prices look high, relative to that path, people bring forward extraction/sale; whenever present prices look low, relative to that path, people shift outward extraction/sale.
And so you don't expect people to do something dumb like dump a pile of the resource onto the market when prices are low if they expect prices to be sufficiently higher in future. NZ's climate commission basically expects forest owners here to do that: to dump a pile of credits on the market in the 2030s, and then everyone faces rapid escalation in carbon prices a little while afterwards when those credits run out. But that doesn't make sense - forest owners would hold credits for future sale instead. Or someone else would buy the credits from them and hold.
Now consider housing. You have a pile of sites in town that are forbidden against intensifying, and a pile of sites out of town that are forbidden against transforming from paddock to subdivision. Land prices across the entire urban gradient are artificially inflated.
If you ease zoning so houses can be built in more places, the scarcity rents built into zoning have to dissipate over time. Because zoning stops being as relevant a constraint. Selling earlier rather than later means capturing at least some of that, before the decline. Paddocks turn into subdivisions; those houses compete with downtown properties; downtown land prices are pushed down. Recall that this is a transition to a new lower price path. Holding in hope of some future price increase won't help.
Isn't the simple direct response to the argument that housing isn't a nonrenewable commodity? You build a house, it pays out imputed rents in perpetuity (with some upkeep of course). The value of a house today is a perpetuity of its rent values today, and the value of it ten years is the same payments but starting 10 years from now. There are some frictions between the land vendors and builder-developers in terms of financing, political savvy, and differing valuations, but you'd expect some coasian bargain to occur so that land wouldn't sit underused since the NPV of the perpetuity today is greater than the NPV starting n years from now.
If you wash up on a desert island with a gallon of oil and a bunch of building supplies, you're going to be careful portioning out your oil over your lifetime, but your not going to hold-off erecting your shelter structures: you want to get as many years out of it as you can.
It’s the development site that is not renewable (irreversible) once you’ve built a building (at least for the economic life of the building)
This all is built on the assumption that there is enough land in the area, enough different owners (no monopoly/cartel), and enough groups that are able/willing to make use of upzoning changes (owner-occupiers have no intention to upgrade) to drive prices down. I do not expect this to be true for all cities. If land is too scarce, prices will still go up, meaning delaying remains a valid option for developpers. If almost anything is owned by owner occupiers, why would they upzone? If you expect owners to be bought out, this would likely mean land prices increased (else why would they leave), in turn providing justification for higher rents/house prices . If there are a few major landlord corporations that influence a significant part of the market, they have the possibility to play the long game again. I see how Murray's analysis might be too simple, but it is still more nuanced than the YIMBY typical story is.
Nice arguments.
However I think the dichotomy between "investors and land speculators willing to sell their land for development if the price is right" and "owner occupiers unable to be induced to sell at any price" is too stark.
The option to sell up, quit town, retire to the regions, and watch your old villa demolished and replaced by townhouses is perpetually available to owner-occupiers.
When prices rise enough, the capital gains are attractive enough to induce a lot of this.
In NZ's and Australia's highest priced cities (Auckland and Sydney), we see larger net internal migration outflows of citizens than in other cities. Growing pains might explain this. But so might the attractive financial return on leaving the city and buying in the regions.
The upshot?
The idea that the current owners' preferences might effectively constrain the amount of zoned land able to be built on such that zoning more could relax a hard constraint is not, I think, that relevant. When the price is high enough everyone is a seller.
Thanks for the elaboration! As non-economist looking with an interest in this, it is hard to get the nuance without feedback. In regards to the last point, I agree everyone is a seller if the price is high enough. The concern is that if prices rise 'enough', this goes against the goal of the policy, as the goal was to bring down (linked) prices. Combined with intertia (ie. people prefer neighbours they know) and cost of transition (moving house, demolition, construction, etc), the required increase in land prices to free them up land might mean new more denser housing ends up costing the same as the old houses.
Eric, trees and housing are clearly the same.
With one exception: old wooden trees used as carbon sinks are beautiful while old wooden houses used as carbon sinks are ugly racist symbols of a shameful colonial past.
At least that's what we believe here in Wellington :)
In seriousness...
You've assumed your conclusion. That's the flaw in your argument that trees and houses are different.
Take your comment above:
"land prices across the entire urban gradient are artificially inflated..."
"If you ease zoning, the scarcity rents built into zoning..."
Without the assumption of "artificial inflation" and "scarcity rents from zoning" your argument falls over.
You say that SINCE land prices are high because of regulation, removing regulation will make land prices fall, and once people recognise this, they will respond by building more, faster.
Replace the SINCE with IF and your statement is correct.
But it's a big IF. Your argument is based on the premise that regulation is currently making land prices higher than they would otherwise be.
What justifies this premise?
As best I'm aware, you and others believe this solely on the basis of AMM modelling in which population is assumed fixed and all developable land is assumed to be fully developed. In this model world, regulatory restrictions on physical space for building mechanically and necessarily raise land prices. This model world is highly unrealistic as an explanation for current land prices, for many reasons.
Am I wrong? Do you have other theoretical arguments for why you think zoning is increasing land prices?
The key point to recognise about zoning regulations is that they're binding on density at the site level, but not binding on the rate of new supply at the market level.
The state regulates site density, the market regulates new housing supply.
If zoning doesn't bind new supply, it doesn't raise house prices above some counterfactual. The market sets prices. The price of land flows from it. So the price of land is no higher due to zoning.
Hi Tim, Eric, Cam
I see no contradiction in Eric's two articles. I think Tim correctly diagnosed the critical juncture: the claim that excessive land regulation causes urban land prices to be much too high. (Tim asks what those causes are; I'll come back to this.)
If relieving those excessive regulatory constraints causes trend decline in urban land prices towards normal levels, then the real option value to delay gets very small, and investment timing will be brought forward. Or conversely, relieving constraints allows superabundance of land supply, then the real option value reduces, which would induce strong steady real (ie, not speculative financial) investment. This would cause trend decline in urban land prices until they leveled off at normal levels.
Normal levels is farm land price (which is a mere $2 to $4 per m2 to buy, not rent, in New Zealand) plus the opportunity cost of fringe development and infrastructure (including a real option value for irreversible investment), plus a premium for inner city living because of savings on transport costs.) At normal levels then housing investment might look a lot more like the mining example. Note, urban areas are only about 0.9% of NZ land area, and expand at half the rate of population increase (0.7% compared to 1.5% pa) so are hardly "sprawling".
There is evidence urban land prices are excessive. Eg see https://tewaihanga.govt.nz/our-work/research-insights/urban-land-prices-a-progress-report . They were nearly $1300 per m2 more at the urban fringe of Auckland after accounting for infrastructure costs.
What are the regulatory constraints that cause excessive urban land prices?
Chiefly, constraints on the competitive supply of fringe urban land across all cities as a system. These are: geography, regulation (eg urban growth boundaries, forbidding of leapfrogging to cheap land), a lack of public infrastructure (because we removed revenue bonds in 1989 and 1995); and a lack of long-term planning to reduce future costs since the 1980s (Solly Angel style).
This causes land price appreciation, and there are many feedbacks that cause this to escalate and be sustained for many decades. (Eg, delayed real investment, speculative financial investment, suppressed public infrastructure supply and increasing regulatory restrictions, genuine financial investment (because housing returns become unduly sensitive to interest rates and become less positively correlated with other investments in a portfolio), increasing aversion to local upzoning by homevoters etc.)
For more, see these articles in the public domain (not exhaustive). Note the concepts are not reliant on AMM modelling Tim, although I find it helpful to use the AMM to start to analyse them. I'd like to use real options analysis one day, but it's really hard. I admire Cam's advocacy of real options analysis, which inspired my study of them.
https://www.treasury.govt.nz/publications/oia-response/competitive-urban-land-markets-treasury-report-oia-20180476
https://www.treasury.govt.nz/sites/default/files/2022-09/chris-parker-slides-amm-modelling-uncompetitive%20urban-land-markets-jun21.pdf
The two 2021 papers from the Urban Land Markets Group: https://www.auckland.ac.nz/en/business/our-research/research-institutes-centres/economic-policy-centre/research-hubs/urban/publications/policy-papers.html
https://www.treasury.govt.nz/publications/jp/assessment-housing-system-insights-hamilton-waikato-area
https://www.treasury.govt.nz/publications/speech/housing-affordability-aotearoa-new-zealand-importance-urban-land-supply-interest-rates-and-tax
https://www.treasury.govt.nz/news-and-events/news/what-drives-rents-new-zealand
Chris Parker
Morning Chris
Slow reply from me. Five points in response.
*****
First, I didn’t ask about the CAUSES of land prices being too high. I asked for EVIDENCE that zoning causes land prices to be too high. Evidence means two things: theory linking zoning and prices, and empirics that falsify other theories of prices.
There’s a general methodological point here.
It sounds like your approach to inquiry is to ask “first, are land prices too high, and if so, why?”.
If that’s the case, then your approach puts the cart before the horse. It will lead you to unfalsifiable claims about cause and effect. This is unscientific. Your claims will be unfalsifiable because you’ve already found empirics that prove (to your satisfaction) that these effects are present.
Specifically, one cannot judge whether land prices are “too high” without a theory of what determines land prices.
If you do so via casual empiricism stripped of theory, you’ll be misled. Just as patterns in nature can look like intelligent design when we have no theory of evolution, so too can urban fringe differentials or prices in Houston look like proof of growth boundaries driving up urban prices when we have no clear theory of what urban prices would be without the boundaries.
Searching for causes of why land prices are too high before having a theory of land prices means searching for unfalsifiable claims.
Short version: I’m drilling you and Eric for clear theory of why you claim zoning rules are driving land prices higher. All I see is closed-city AMM models. Do you have any other theory?
*****
Second, I think we agree on the mechanism by which land prices WOULD come down IF current zoning were causing them to be higher than they would be under looser zoning.
That mechanism is expectations and investment timing (a lower value on delay).
Do we agree that the crux of disagreement is the premise in this argument?
1. Tighter zoning causes land prices to be higher than if zoning were looser [premise]
2. Upzoning would change price expectations and investment timing [mechanism]
3. Via mechanism (2) upzoning would lower land prices [conclusion]
*****
Third, we agree that anyone using the argument above to justify the premise in (1) is using circular reasoning. Premise (1) needs to be justified independently. You can’t prove premise (1) using the conclusion of an argument that relies on premise (1).
Real options logic is just a mechanism. It’s interesting but not complete as an argument for conclusion (3). You can only reach conclusion (3) with premise (1). That’s why I ask for your theory to justify premise (1).
Eric’s arguments read as circular. He takes premise (1) as given, talks through mechanism (2), then reaches conclusion (3) without offering independent support for premise (1). (To be clear, he might have independent reasons for premise (1), but hasn’t expressed them in anything I’ve seen.)
Short version, do we agree you can’t prove zoning is driving land prices higher via real options logic?
*****
Fourth, while I generally agree with your real options reasoning (mechanism 2), there’s something you’ve forgotten about when applying that reasoning that derails your claim that a “rush to the door” is possible by developers seeking to sell homes and land before they fall in price after upzoning.
That something is construction capacity.
In the short run, capacity is fixed. In the medium run, it’s flexible. But between short and medium run, any “rush to the door” will alter the relative cost of present and future construction in a way that halts the rush.
Specifically, any developer building more today will have to bid fixed construction resources away from other developers. That will raise construction costs. That will induce entry. That will make tomorrow’s costs lower than today’s elevated costs. And that will increase the real option value to delay.
That means if some developers build more today, other developers will build less today.
Do you understand this argument?
(The same logic applies to the effect of current development on current sales prices relative to future prices).
These two factors cause market forces to “regulate” the rate of new housing supply. Sales prices and construction costs are automatic stabilisers. They affect incentives to develop vs delay in ways that regulate/stabilise the rate of construction and sales.
You can alter this single market equilibrium, but you’ll never reach an “escape velocity” in the rate of development that gets you to some more desirable equilibrium. There’s one equilibrium, not two.
BTW I’m yet to see the endogeneity of current and future construction costs embedded in a mathematical real options model of development, but you can see from my verbal reasoning how such a model would work.
*****
Fifth, recall the efficient markets hypothesis?
It says that all information is already contained in the price. It applies to financial asset prices. Land is a financial asset.
If you’re comfortable with the EMH, think about its key implication – that land prices will fall immediately once smart money (e.g. landbankers) think the policy settings are right for sustained decline.
It won’t take years of actual building. Land is infinitely lived, so most of its present value comes from distant future rents. This means credible policy will crash land prices immediately.
Yet this never happens. Tripling zoned capacity in Auckland from 30 years’ to 90 years’ supply didn’t do it. And if a hefty dose of the medicine did not budge the needle why do you suppose more medicine will suddenly work? Do you have theories of tipping points, thresholds?
Also why do developers and even developer LOBBY GROUPS push for this?
You can posit co-ordination failure when individual landowners push for zoning changes in their own interest but not the collective landowner interest. But you can’t plausibly claim co-ordination failure when groups existing precisely to co-ordinate interests are pushing for this.
If all the smart money wants zoning change that you claim will destroy their asset values, have you been fooled or have they?
Tēnā koe Tim
On 1, I agree theory is needed before interpreting data in cities.
Ed Glaeser said this p117 in his 2008 book ‘Cities, Agglomeration and Spatial Equilibrium’: “Understanding cities requires understanding at least three types of decision-making agents: individuals who are choosing where to live and work, firms that are choosing where to operate and how many people to employ and developers who are deciding how much to build. This perspective is part of a broader view that the economic approach to understanding cities can never use empirics without theory. I believe that every correlation or regression investigated by an economist should be justified, if only in the economists’ head, by some kind of formal model that starts with decision-making agents.”
I presented at this year’s 2023 NZ Association of Economists (NZAE) annual conference on the work the Housing Technical Working Group (the Treasury, Reserve Bank of NZ, and the Ministry of Housing and Urban Development) is doing on land efficiency indicators. The gist is market operations free of frictions should lead to continuous prices across all margins of adjustment (net of fixed costs) or close to that if there are lumpy increments such as stories. Those margins of adjustment are numerous (eg, floor area vs garden space; distance from city centre; residential vs business use; rural-urban). However, this won’t maximise social welfare if there are externalities that aren’t internalised through pricing, landownership etc. Assessing allocative efficiency also requires shadow pricing of those wider effects omitted from market price signals. Research papers on the theory and metrics will likely be forthcoming.
On 2, I don’t agree with premise 1 in all cases, but in some cases I do. In my spatial equilibrium theory and modelling of cities, if a city fails to accommodate people well (by imposing high costs to intensify) then fewer people will live in the city, which bids down its differential land rents. This accords with the Henry George Theorem I’m developing for an open city in spatial equilibrium. (I presented on this work at the 2022 NZAE conference.)
On 3, I wouldn’t start with those kinds of premises, as they breach Glaeser’s principle I quote above. I’d start with the objective functions of households (utility maximisation), firms (profit maximisation), and housing developers (profit maximisation), and perhaps (if I’m researching local public policy) local governments (supposing their objective is differential land rents maximisation).
On 4, I agree construction sector capacity is an important constraint on the rate of house building. I think it determines house rents, which are contemporaneous. House prices are future looking, as you outline in your point 5, so they can respond much more abruptly. I think construction sector capacity can evolve differently over the long-run depending on if it faces a steady large flow of work vs volatile low flows of work.
On 5, I agree the credibility of policy is very important, and can drives prices and investment behaviours in the short-term. Be careful with the word “never”. There is some historic literature that land prices dropped precipitously from the late 1800s to the mid-1900s, and this was associated with globalisation and technology revolution. (Eg Louis Winnick’s 1953 PhD thesis “Wealth Estimates for Residential Real Estate, 1890-1950”.) That’s what Thomas Picketty documented too in his 2014 book “Capital in the 21st Century”: massive wealth in land (urban and rural) in the 1700s and 1800s, reducing to a minimum in the three decades after WWII, and now urban land prices are ballooning again.
Well I never an economic logic arguing both sides of the same argument. Hang on wait the aerodynamics "theory of lift", and the theory of everything else....... Perhaps we still need to temper this theorising with practical demonstration to actually achieve any change...
Whether reducing obstacles to building new housing in particular jurisdiction will reduce the (quality adjusted) price of housing is an interesting issue. It is however not central to the policy decision about how to regulate construction. Suppose in the new regime only a lot of "luxury" houses/apartments are built for people who lives elsewhere before and the average price of housing rises. That would still be a desirable outcome for the liberalizing jurisdiction.
While I appreciate the theoretical argument, in the real world which is messy, individual choices will be different. The cheap criteria is the ability to carry assets based on current cash flows and amount of debt being carried.
As you can see from current interest rises, first home buyers who bought in the last ~2 years are starting to dump property. While property owners with multiple properties and/or with much lower LVRs are largely unconcerned (so far).
While this is tangential to the main thrust of the article, individual circumstances matter. Also future expected values are volatile (more so now because the expectation itself of property values always goes up is no longer so true) so there is a bias to what’s in it for me now. Far easier to sell a pill to lose weight than a 12month program for diet, exercise and bring a boring person! 😜
Excuse my typos. Can’t seem to edit. Thought provoking piece. 🙏