For those unfamiliar with property development, the following lessons may be of interest.
Imagine a town zoned for 5 storey buildings up to its new greenbelt. At the boundary of this town there will be a situation where one block of land greatly differs in value to its neighbour by virtue of being inside the town while the neighbouring block is in the greenbelt. This difference is the price gradient.
At the other side of the greenbelt one could expect nearby towns to experience a similar price gradient at the edge of the greenbelt. But most importantly there will be a price gradient between the main town and neighbouring town that partly reflects the costs of traversing the greenbelt. These two towns are substitutes. Lesson 1: There are always substitutes.
If the greenbelt didn't exist, homes would be built in that area at an intermediate price giving a smooth price gradient, but fewer homes would be built inside and outside the greenbelt. The most interesting thing is that if you reduce the cost of traversing the greenbelt, by extending the rail network across it or expanding highway capacity for example, this price gradient between towns will decrease – it will decrease the price premium for living within the greenbelt, and increase the prices of home outside the greenbelt to reflect these reduced costs - the symmetry of the price gradient ensures that the general price level is unaffected.
This gives us Lesson 2: Location premiums work in both directions.
What about the rate of delivering new supply over time? Surely planning impacts prices by affecting the rate of development? The short answer is no.
Lesson 3: The rate of land and housing supply is determined by the rate of sales of new stock (known as the absorption rate). It has nothing to do with the rate of development approvals or even the price level. Describing the market in classic economic terms (although a poor representation generally), the supply curve for residential land and housing is vertical, and shifts to the right at a rate commensurate with sales of new land and housing stock (until a theoretical limit imposed by planning which is reached when is it not possible to develop a single extra dwelling - hence theoretical). The image below is the best representation of the market using supply and demand curves.
This is in stark contrast to the common position of elastifying the supply response to price increases. Such talk is clear evidence of a misunderstanding about how land supply reaches market. The following example demonstrates.
A new land subdivision has 20 lots in an existing suburb of 100 homes. This subdivision keeps 3 new lots for sale at any given time and achieves a sale about once every 2 months. Also in the suburb is a farmer who wants to sell his land, which could be subdivided into 20 lots under current planning rules.
A developer is assessing the feasibility of subdividing the farmers land. One important thing he needs to know is how long it will take to sell the 20 lots. Given that the existing subdivision is only able to sell 6 lots per year and he will have to compete for the same level of demand, the developer estimates he will only be able to sell one block every 4 months, thus taking 6 years to sell the whole development. At the same time, the existing subdivision will take twice as long to sell due to the new competition.
Given this huge time delay in receiving revenue, the developer may find that the highest price he can pay and still make a profit is below the value of the land as a farm. Even if his price does buy the property, it doesn’t matter how many blocks he puts on the market, the number of sales in the area will be constant.
Indeed the rate of sales also affects how long financing can take, given the requirement for presales prior to lending for construction. While it may be theoretically possible for discounting to bring forward demand enough to increase returns, but this requires an extreme degree of price elasticity of demand and unrealistic cost structures - it is not what happens in the market.
Indeed the rate of sales also affects how long financing can take, given the requirement for presales prior to lending for construction. While it may be theoretically possible for discounting to bring forward demand enough to increase returns, but this requires an extreme degree of price elasticity of demand and unrealistic cost structures - it is not what happens in the market.
We can use this example to demonstrate Lesson 4: There is no price competition in land markets. Value is derived from the highest and best use, and there is no incentive to compete on price.
Our developer is still assessing the feasibility of the farm land. He knows he can increase his rate of sales and decrease his risk if he undercuts the prices at the existing subdivision. However, there is no incentive to do so unless it increases his return. If he drops his price by 10%, he may sell blocks twice as fast, but is losing almost 50% of his profit.
But the only way he can decrease his costs is to offer a lower price to the farmer. But this would not be enough to compete with the price other developers would offer knowing the returns from selling at market prices. It never adds up to compete on price.
Imagine that the developer buys the block and demand increases in our hypothetical suburb. Prices increase and the rate of sales of land doubles to 12 lots per year.
Under these circumstances the developers could choose to maintain their prices and attempt to undercut the market, selling all their blocks much quicker. However, this simply provides a gift to the buyers who can on-sell for a profit.
In a market with rising prices there may even be a slight incentive to slow the rate of blocks offered to the market and get higher prices for more of the stock.
Even in a market with low demand leading to falling prices there is no incentive to withhold supply. A developer with a stock of land has the choice to either sell at the market price and rate, or withhold supply to wait for a recovery. But withholding supply in a falling market is the worst choice. Revenue evaporates while the value of remaining stock falls – a recipe for a date with the receivers.
Now imagine the council is comes pressure due to unaffordable home prices, and upon advice from Economists-For-Hire consultants they change the planning code to allow more density.
The two current subdivisions combined have 25 lots remaining unsold. They apply to council to change the configuration of their remaining land to accommodate 35 lots. Because the price per square metre is higher for smaller lots they are now both on track to make greater profits than anticipated. Yet they continue to supply at market prices with sales rates determined by demand. . Lesson 5: Relaxing planning controls does not change the rate of supply of dwellings, but provides windfall profits to land owners.
Such windfall gains can also lead to gaming by landholders. Knowing that the council is very keen to increase density, the next landholder with development potential has an incentive to wait and see if the council will boost allowable densities once again. Why sell now when it is likely that the council will provide you with massive increase in land value in the near future? This why developers land-bank in fringe areas.
Similar reasoning applies to reducing upfront capital contributions from developers. Since the value of development sites is the residual of revenue minus costs, and reduction in costs increases the site value. This gives us Lesson 6: Benefits from decreasing costs of development accrue to current the landholder, not the final purchaser.
In fact, recent examples show that when the State Government has intervened and ordered council to stop increasing planning density there has been a flood of development sites on the market. Lesson 7: Markets have political expectations.
The final lesson is Lesson 8: Planning rules are not strict. In Brisbane for example, the Cityplan has a three methods for approving development applications depending on the scale of your proposal – self-assessable, code-assessable, and impact assessable. If the scale of your development is within the boundaries of self-assessable limits in your area you can develop without a development application! This means that it is actually impossible for councils to delay such developments.
In Brisbane, it is only when a developer applies for an impact-assessable approval where public notification is required and where long delays from NIMBYs and State Government referrals can occur. This approval process is for proposals that fall outside the scope of the plan, such as requests for greater heights and densities.
The reason that this happens so often is that in order to buy a development site you must outbid other developers who have factored in better outcomes from impact-assessable approvals. It is only the foolishness and overconfidence of developers as a group that increases the number of these more risky, time-consuming and expensive approvals.
The property market works within planning controls to develop land to met demand. But these planning controls merely change the location and type of development to what might have otherwise occurred, but are absolutely necessary to coordinate infrastructure and city planning. While plans can pick winners and losers, most importantly they cannot impact the general price level of housing.
Any questions?
Not sure if you're factoring in the time value of money for lesson 4? If a developer undercuts his competitors, he may sell more, pay less interest and still come out better for it. If he held, he may lose all his gain to the bank. Unless he is big enough to control supply...?
ReplyDeleteAlso if a developer increases density, he may sell more because that's the price point that will sell, without him making a loss on the site (albeit he may have paid too much to start with) Although, cost of infrastructure may be reduced with greater density, allowing development to occur where it mightn't have previously i.e. remained a farm - perhaps a farm is the better use?? for planners to decide I guess....
Re: Lesson 4, the problem with discounting is that you are giving the buyer a gift - the ability to resell for a profit at market prices.
ReplyDeleteI can only imagine that increasing the NPV of the development by decreasing prices would require an impossible combination of extremely elastic demand and high interest rates.
Comment from email (response in bold)
ReplyDeleteI think you are using a theoretical model to assess a real problem. The reality of the Brisbane, and other housing markets, is there is no equal substitute town. The Sunshine Coast, Gold Coast or Ipswich is not a suitable substitute for Brisbane. Brisbane is where the work is and therefore, people want to live close to work. Yet train loads and highways full of people commute from the Gold Coast, Logan, Caboolture, Ipswich and fringe development areas (lakes developments)– looks like a substitute to me. Mostly the limit is about 1 hour travel from work. As you indicate the dwellings closer to Brisbane CBD are more expensive and get cheaper as you get further away from the CBD. There is a limited amount of land within 1 hours travel of the CBD. The Federal, State and Local Governments all interfere with the amount of land available for housing. Not amount. Location. These Governments do release 'new' land however, this land is not a substitute for dwellings close to the CBD.
Dwellings close to the CBD are regulated by the Governments in a variety of ways. You say: “There is no price competition in land markets. Value is derived from the highest and best use, and there is no incentive to compete on price.” I would say the Government intervention in markets do not allow the land to be put towards its highest and best use Government decide the limits of development on a site – that limit is typically the highest and best use. Just because nearby site are allowed a different limit, does not mean that that site should also have that limit. The Government does not allow freedom to divide land into smaller blocks or build more dwellings on the same land. Old houses on 800 sqm blocks could be developed to make way for town houses or very large residential high rise. This would greatly increase the supply of dwellings close to work. However, the Government has vast amounts of Brisbane land restricted to only allow improvement on 100 year old houses. Therefore, the price of the individuals house grows while the rest of the people wanting a dwelling are force to move ever further away from the CBD. I agree that planning does force development at different locations to where it would otherwise occur. But note the symmetry of this price premium.
The cost is to the many and the benefits to the few. The price of a dwelling within 3km of the CBD is extremely high and you still only have a 2 bedroom house made out of wood. Everyone else has to pay higher taxes to build more roads and trains to travel further and further. The environment suffers through urban sprawl and increase pollution from driving to the CBD. Therefore, I cannot understand your reasons for saying “Relaxing planning controls does not change the supply of dwellings, but provides windfall profits to land owners.” You are still talking about the location, not the supply. If there is sprawl there is new supply.
continued...
ReplyDeleteBrisbane planning rules are extremely strict! The fact that the limits set in planning are enforced does not make it strict. It is almost impossible to build a new house within 5km of the CBD. You try and purchase a new house in that distance of the CBD and you will come up wanting They exist, but they are expensive of course. Unless you want a small unit in the Valley for over $500k. There are entire suburbs in Brisbane that you are not allowed (Government policy) to develop. Yes, but doesn’t mean that there are other areas that development is allowed. This is a large restriction on supply of dwellings in Brisbane and reduces the standard of living of every person living in them. Look around. There are thousands of apartments proposed or being constructed in West End, Bowen Hills, Woolloongabba, Kangaroo Point, you name it. There would be even more constructed if people would buy them.
So your point that the speed of approval may not affect the house price is not the problem. The problem is the other areas of Government that interfere with the housing market. The above example is only one. The others would include housing commission close to the CBD, Government development departments competing with private, Government developments competing for limited development funds etc.
I think your definition of supply and demand is a bit off. They are ex-ante concepts, not ex-poste. The potential development of the farmer’s block counts as supply and influences price. The 17 blocks that the developer is keeping up her sleeve by only releasing 3 a month also count as supply and influence price.
ReplyDeleteThe land that is allocated to its highest value use only got to that use by other uses competing for it. The next-best use was still competitive, it just was able to meet the market-clearing price. Developers will compete for land, and reduced development costs will be shared by the property buyer and property seller, depending on market power.
Anyway, I wholeheartedly agree with the sentiment that the focus on development delays and costs is way overblown. From my recent experience, the primary change in the housing market has been the availability of credit. Banks would drive to your house on a Sunday 3 years ago to talk about loans. Now they won’t return your calls. Development costs, housing shortage, employment, income, population, demography and every other supposed ‘driver’ is just background noise. So I guess I’m a begrudging housing bubble convert.
Do you really take developable land as supply? By my estimate there must be millions of such blocks in Australia (given that I can think of about 200,000 in SEQ alone off the top of my head). If that was the case prices would be near zero!
ReplyDeleteI say he keeps it up his sleeve to give the impression of urgency to buyers of the 3 block that are for sale. Not that it would accelerate the rate of sales if he put them all on the market.
Second, land goes to its highest value use not by competing with other uses, but by competing buyers who want to use it at the highest and best use. There doesn’t need to be some secondary use to create value. Indeed, planning rules usually constrain residential block to single residential dwellings – what’s the secondary use there?
What do you mean by this – “Developers will compete for land, and reduced development costs...”? If developers compete for land they bid up the price of raw land – an increase in development costs.
There are no technical innovations in land developments. Essentially you buy a site, put in the planning application, and sell the blocks. You cannot reduce costs by being more efficient. They are all fixed (the site component being fixed by the market). You cannot beat the market in property through innovation, only from gifts by councils – Lessons 5 & 6.
Dear Cameron
ReplyDeleteThanks for drawing my attention to this.
I agree that complaints about planning restrictions largely miss the point. No amount of deregulation will force land owners to build housing and offer it for sale or rent. Private decisions concerning whether to develop, whether to build, what to build, and what to do with existing buildings, are heavily influenced by taxation -- e.g. whether taxes represent holding costs, transaction costs, or building/development costs, and whether values of buildings are included in the tax base. Hence I have always been more interested in tax reform than in planning reform (see e.g. the debate on stamp duty, "land tax" and CGT at http://is.gd/psd_strike , which I am not necessarily winning).
But I would never have gone so far as to claim that planning restrictions make NO difference. As long as there are people wanting to go all the way -- i.e. build housing and offer it for sale or rent -- and governments saying "No you can't," then governments will be reducing the supply of housing, forcing people to compete more keenly for it and consequently to pay more.
--- Gavin R. Putland.
Thanks for the comments Gavin. While I might not have made it clear, I do believe that governments COULD influence the general price level of housing through planning controls if the result was effectively in a quota system ie. they would only approve a certain number of dwellings per period over quite a large area, say greater SEQ or greater Melbourne. But I have never seen such a situation, hence my lessons reflect how I see development happening on the ground (where quantity of sales limits dwelling construction, rather than development approvals)
ReplyDeleteI will chase up some data on the number of development approvals in different local government areas as I would be very interested to see just how many more approvals are granted than dwellings constructed.
Thanks again.
Cameron
ReplyDelete>I will chase up some data on the number of development approvals in different local government areas as I would be very interested to see just how many more approvals are granted than dwellings constructed.<
I'll keep an eye out for that
You say:
ReplyDelete".....Imagine a town zoned for 5 storey buildings up to its new greenbelt. At the boundary of this town there will be a situation where one block of land greatly differs in value to its neighbour by virtue of being inside the town while the neighbouring block is in the greenbelt. This difference is the price gradient...."
This is not a "price gradient", it is a "price discontinuity", i.e. a serious "step" on a curve that naturally would have a smooth slope.
You correctly point this out:
".....If the greenbelt didn't exist, homes would be built in that area at an intermediate price giving a smooth price gradient......."
But you go on to say:
".......but fewer homes would be built inside and outside the greenbelt......"
No, this is quite wrong, both in real life and in land economics theory. More will be built inside the greenbelt, and fewer outside, if an urban land market is functioning freely without regulatory distortion. More thorough explanation later.
You say:
ReplyDelete".....The most interesting thing is that if you reduce the cost of traversing the greenbelt, by extending the rail network across it or expanding highway capacity for example, this price gradient between towns will decrease – it will decrease the price premium for living within the greenbelt, and increase the prices of home outside the greenbelt to reflect these reduced costs - the symmetry of the price gradient ensures that the general price level is unaffected...."
The price of CBD land will rise considerably because of the increased access to it by greater numbers of people. This usually has an effect on the price of land in the innermost suburbs as well, partly due to pressures from potential CBD expansion.
One thing that is quite crucial, is how "controlled" development is outside the green belt. You are right to suggest that land rents inside a green belt can be eased over the long term, PROVIDED that the development OUTSIDE the green belt is sufficiently "mixed" as to allow for decentralisation of employment as well as residential. This implies that some activity will be displaced from the existing CBD due to lower land prices elsewhere being a major factor in the decision. Ironically, this incentive becomes greater for households as well as businesses, the higher land prices throughout the city are driven.
Good planning for growth, especially rights-of-way for roading and other public infrastructure, enables rapid low-cost provision for growth in comparison to badly-planned existing areas where built-out property needs to be acquired and NIMBYism overcome. It also ensures that congestion is avoided as growth occurs.
ReplyDeleteYou also say:
"....Describing the market in classic economic terms (although a poor representation generally), the supply curve for residential land and housing is vertical, and shifts to the right at a rate commensurate with sales of new land and housing stock (until a theoretical limit imposed by planning which is reached when is it not possible to develop a single extra dwelling - hence theoretical). The image below is the best representation of the market using supply and demand curves......
.....This is in stark contrast to the common position of elastifying the supply response to price increases. Such talk is clear evidence of a misunderstanding about how land supply reaches market....."
This is a very common misinterpretation of classical land rent economic theory, but one that has done colossal damage everywhere the economics profession has intersected with the land use planning profession. Here is how it has been admirably dissected by Prof. Alan W. Evans (University of Reading, UK), in his book "Economics, Real Estate and the Supply of Land" (2004).
We start with Ricardo. When all land is used to grow corn, the price of corn will determine the "rent" of land. However, when there is competition between, say, corn growers and potato growers for the available land, the most profitable use will determine the "rent" of as much land as is required to reach the equilibrium where demand for that product is satisfied; the remainder can be used for the less profitable activity and yeild lower rentals.
The same principle can be extended to all potential land uses; and ultimately, the higher value land uses do not require so much land that the rent of the lowest-value-use land will be raised by very much at all, except in extremely land-constrained nations. (But even then, "factor substitution" in international trade does have an effect)
Moving on with our discussion of theory, urban land rents can be derived similarly to agricultural land rents as described by Van Thunen, according to their advantage of location. However, all the theory described so far, assumes "total land supply" to be fixed, AND "under competition" between the various potential users. The theories relating to urban land rents are simply an extension of those relating to agricultural land rents.
Evans points out the glaring omission on the part of economists who today, parrot the theory without understanding the underlying proviso: that is, that there needs to be free competition between the various potential users of "all land", for the theory to apply at all. Obviously an urban growth boundary invalidates that proviso. The whole point of the "supply is irrelevant" theory, is that the supply is VAST, and rents set purely by the use of it under conditions of competition between users, where higher value uses outbid lower value uses. As soon as conditions of competition for ALL land is removed arbitrarily, and supply for a particular use is LIMITED to a tiny fraction of the total, of COURSE "supply" IS important.
Evans does wonderful graphs.
Evans also suggests that economics students have often confused economic rents and commercial rents in their thinking.
The relationship between permitted densities and land rents is complex (as my comment above about prices per square metre illustrates), but is not an economic game-changer like growth restraints. As a broad principle, in a city with an uninterrupted land price curve and low overall land prices, higher densities occur in redevelopment and infill and so on close to the urban centre(s) and fringe development is invariably much lower density. Households trade off land cost against travel distance.
ReplyDeleteBut as land has escalated in price in growth-constrained cities, this trade-off has been eliminated. Households are now forced to "drive to qualify" for a mortgage. The cost of transport always is reflected in real estate values - a premium needs to be paid for a location that enables transport savings. This effect remains constant regardless of the "base price" of the land, which is determined by the price of the lowest-priced, least conveniently located land. But an elevated "base price" due to a discontinuity at a regulatory boundary, results in the "cost of transport" price signal being swamped in the COMBINED "cost of mortgage plus cost of transport" signal.
Consider this hypothetical case (which DOES reflect real life). Before land price inflation, an average young household might have the following choices at each extreme. Central location: transport costs 5% of income; mortgage costs 80% of income. Total 85% of income. Fringe location: transport costs 15% of income, mortgage costs 40% of income. Total 55% of income. There will be numerous viable choices in between.
AFTER land price inflation: Central location: transport costs 5% of income; mortgage costs 160% of income. Fringe location: transport costs 15% of income, mortgage costs 80% of income. Impossible. So it becomes necessary to trade off space as well as location; that is why fringe development becomes denser. Increase the density to the point where the average young household MIGHT, with a sub-prime mortgage, be able to buy a home at all. They now might have, transport costs, 15% of income, mortgage costs, 60% of income - for a smaller home than they had under the low-land-price scenario. The total strain on their income is much higher. And the "range of choices" they had to live closer to the urban centre(s), is now non-existent. They are forced into the worst option on all counts; cost, location, and space.
wodehouselee, I'd be keen to hear your reasoning behind the claim that fewer homes would be built outside a greenbelt. Surely peopl who would have bought there now have the choice to live in the city for a higher price, or live outside the greenbelt for a lower price. Not everyone will take the first option.
ReplyDeleteAnon, I don't think that decentralisation is necessary or likely to reduce the price premium inside a greenbelt. In fact, when the cost of traversing the greenbelt falls there is a greater incentive for business to shift into the city (to get access to a greater potential workforce and customer base), and far less incentive for people to live inside the greenbelt (since getting into the center is less costly).
Human resource intensive businesses, and retailers, do not shift away from city centres under almost any circumstances. That's why decentralisation policies always require governments to relocate departments. This then becomes a problem because these human resource intensive departments have fewer people within commuting distance and more difficult interactions with other business.
"Anon" is actually me, "Wodehouselee". I don't know why, but after my first comment, your system did not allow me to post, giving me a message that "this ID has been blocked by the systems administrator". I therefore resorted to posting as "Anonymous" but most of my postings, while they successfully appeared at the time, had disappeared a few hours later. I will have a few more tries one way or the other.
ReplyDeleteWodehouselee continuing....
ReplyDeleteIronically, there is every likelihood that it is the roads and cars that have enabled Manhattan to resist the trend to leakage to the surrounding competing nodes. Even so, Manhattan HAS leaked somewhat. Another factor often pointed out, is that SOME cities CBD's are based on "international capital" rather than national and local capital. It is unrealistic to expect to replicate Manhattan too often, when there is a limit to the number of international businesses and wealthy magnates who can locate anywhere else.
One thing you are correct about, and Crane et al point this out, there is nothing to be gained by government departments relocating away from city centres. There is in fact, economic efficiency gains to be made in the decentralisation of almost every other sector. Kang-Rae Ma and various co-authors have produced some good analyses of Seoul's decentralisation following on their implementation of a green belt. And even then, the land supply OUTSIDE the green belt has been rationed, leading to one of the world's worst ever housing bubbles as early as the late 1980's. A unique feature of THAT bubble was that it had nothing at all to do with mortgage finance - it happened IN SPITE OF an almost non-existent mortgage finance industry and cultural traditions to work hard, save hard, and "pay cash" for one's first home.
IF planners manage to thwart this decentralisation process, as well as forcing land prices up within the limits they impose, there are dire consequences for economic and social outcomes in the long term. This is one more reason that a bubble in land prices cannot last forever - it is a choke on economic growth and a creator of inequality and social instability, which effects HAVE to collide with the unaffordability of land sooner or later.
Check out the terms "dense sprawl", or "dysfunctional density". Not that those using those terms have all yet understood that it is "limits" that are the problem, not decentralisation.
Yours is a very common misinterpretation of classical land rent economic theory, but one that has done colossal damage everywhere the economics profession has intersected with the land use planning profession. Here is a summary of how it has been admirably dissected by Prof. Alan W. Evans (University of Reading, UK), in his book "Economics, Real Estate and the Supply of Land" (2004).
ReplyDeleteWe start with Ricardo. When all land is used to grow corn, the price of corn will determine the "rent" of land. However, when there is competition between, say, corn growers and potato growers for the available land, the most profitable use will determine the "rent" of as much land as is required to reach the equilibrium where demand for that product is satisfied; the remainder can be used for the less profitable activity and yeild lower rentals.
The same principle can be extended to all potential land uses; and ultimately, the higher value land uses do not require so much land that the rent of the lowest-value-use land will be raised by very much at all, except in extremely land-constrained nations. (But even then, "factor substitution" in international trade does have an effect)
Moving on with our discussion of classical theory, urban land rents can be derived similarly to agricultural land rents as described by Van Thunen, according to their advantage of location. However, all the theory described so far, assumes "total land supply" to be fixed, AND "under competition" between the various potential users. The theories relating to urban land rents are simply an extension of those relating to agricultural land rents.
Evans points out the glaring omission on the part of economists who today, parrot the theory without understanding the underlying proviso: that is, that there needs to be free competition between the various potential users of "all land", for the theory to apply at all. Obviously an urban growth boundary invalidates that proviso. The whole point of the "supply is irrelevant" theory, is that the supply is VAST, and rents set purely by the use of it under conditions of competition between users, where higher value uses outbid lower value uses. As soon as conditions of competition for ALL land is removed arbitrarily, and supply for a particular use is LIMITED to a tiny fraction of the total, of COURSE "supply" within that limit, IS important.
I have now found a reference to job dispersion in Australian cities; this author really knows his stuff: (Editorial on the "New City" blog)
ReplyDelete"......Writing in a publication of the 2008 9th World Congress of Metropolis, Sydney University’s John Black observed that “apart from some noticeable peaks, employment density is quite uniform across the [Sydney metropolitan] region”. According to the NSW Department of Transport, only 12 per cent of Sydney’s jobs are in the CBD and second tier centres like North Sydney, Chatswood, Parramatta, Hurstville and Penrith have less than 2 per cent each. David McCloskey, Bob Birrell and Rose Yip of Monash University (demographers, not urban planners) report the same about Melbourne. The CBD hosts around 20 per cent of jobs and the rest are scattered all over the metropolitan region.
Platitudes like “we must locate people close to where they work”, or “we must locate jobs close to where people live”, have little basis in reality. They infringe another immovable law of economics, relating to economic rents or bid-rents. This mechanism determines how industries and firms are distributed. Put simply, a parcel of land will go to whichever use delivers the highest profits. Centrally located land (near major transport or infrastructure hubs) commands high prices, and goes to the most profitable uses. Peripheral land goes to less profitable or marginal activities.
Over the last thirty years, economic deregulation, flexible transport, advanced communications and population growth have raised up a sector in the latter category, extracting value from cheap outer-metropolitan land and low rents. It includes industries like transport and distribution, building and construction, food, consumer products, personal services, wholesale and retail. They depend on favourable location costs and proximity to urban markets and labour pools. According to the Greater Western Sydney Economic Development Board, “prime industrial land with direct access to transport infrastructure is 75% cheaper [in GWS] than other areas of Sydney”.
Ultimately, green planning will phase out cheap urban land, undermining this sector and destroying jobs in the process....."
What is happening with my posts? It is as if my subsequent post deletes the previous one. I have referred again and again to a paper, Randall Crane et al, "As jobs disperse, whither the commute"? This always seems to be fated to disappear. The one between 11.44 and 12.21 disappeared, now the one I just posted also disappeared when I posted a further comment.
ReplyDeleteI referred back to this paper in my 12.21 posting; it makes no sense because the earlier comment vanished.
Let's try again, Randall Crane et al, "As Jobs Disperse, Whither the Commute"? Finds that commuting distances generally reduce as jobs disperse. The exception is government jobs. This explains why capital cities, especially where government bureaucracies are notoriously bloated, have a high level of CBD employment eg Wellington, NZ 32%.
ReplyDeleteThe famous Colin Clark way back in a 1982 book, "Regional and Urban Location", discusses the rapid decentralisation of cities that was underway. He suggests that congestion is one cause, and centre city land prices, perhaps due to speculation, eventually "pricing out" many businesses. Because real estate prices are downwards sticky, the CBD prices never fall fast enough to arrest the loss of businesses once it has started. Clark also suggests that CBD's with more road lane-miles are the most resilient to this trend. Ironically, this would mean that Manhattan owes its high density and viable public transit, to the fact that it copes with several times as many cars per sq km than most other cities can. Increasing density NEVER results in an outright drop in car numbers; merely that once extremely high densities have been reached, fewer and fewer cars are added per unit of additional population.
The famous leftwing economist Anthony Downs (now nearly 90) has been saying for decades that planning for rail based travel and high density living, is always doomed to failure by the way real estate markets in free countries work. This is because, as Downs points out, planning and infrastructure provision both drive land prices up; long before as many people as the planners HOPED would live in "x" location, the price of homes at "x" location will have gone up to way beyond what everyone can afford. The result, ironically, is a wealth transfer from the poorest in society, to the incumbent owners of property at the locations favoured by the planners. It is tragically little understood today, how significant a role "save the planet" planning plays in both economic stagnation AND rising inequality.
ReplyDeleteRoads, cars, and sprawl, on the other hand, have been powerful forces for the wide dispersion and sharing of gains from economic progress, as well as the major enablers of that economic progress. Rails and high density go naturally with totalitarian government; roads, cars and sprawl go with free democratic society. Downs himself expresses regret that the people in the free west, while apparently largely convinced about saving the planet via urban growth boundaries, are not politically ready for the totalitarian government.
Attempting to impose "plans" on what is still a free market in Real Estate, actually makes things WORSE than simply leaving the free market alone - except that some outcomes the planners may point to as a "success" will be the result of economic collapse, not people voluntarily changing their behaviour. It is not that large numbers of people have moved closer to their jobs or started catching trains, it is that large numbers of people will now have no job to travel to at all.
"Sprawl" might partly be bad to the extent that it might be forced by minimum section size regulations - I agree that these regulations always were wrong. But "sprawl" is actually largely "decentralisation", and is a free market mechanism by which congestion is reduced, travel times shortened, labour productivity increased, and social mobility increased. Why is the US economy the world's most efficient by a wide margin? "Sprawl" and "automobility" are part of the REASON, not things that the US economy succeeds "in spite of". (I agree that discretionary waste like gas guzzling SUV's are things that the US economy succeeds "in spite of" - yet even this discretionary waste does not neutralise the advantage of sprawl and automobility). Low cost urban land saves far MORE money for the participants in the economy, than can be saved in transport costs, even under conditions of rising oil prices - because the cost of transport, if rising, ends up capitalised into the price of efficiently located land anyway.
In-freakin' credible. Just as well I am now saving everything I post. When I posted the above at "10.22PM", my previous posting vanished. Here it is again. (This is the second time my response to Cameron's argument below, has disappeared - this makes my consecutive postings unfortunately less coherent)
ReplyDelete........One of my disappearing posts, was my response to this:
Cameron Murray said...
"wodehouselee, I'd be keen to hear your reasoning behind the claim that fewer homes would be built outside a greenbelt. Surely people who would have bought there now have the choice to live in the city for a higher price, or live outside the greenbelt for a lower price. Not everyone will take the first option.
Anon, I don't think that decentralisation is necessary or likely to reduce the price premium inside a greenbelt. In fact, when the cost of traversing the greenbelt falls there is a greater incentive for business to shift into the city (to get access to a greater potential workforce and customer base), and far less incentive for people to live inside the greenbelt (since getting into the center is less costly)....."
I meant that FEWER homes would be built outside where the greenbelt would have been, if it was not there; and more inside. This is because land price inflation, due to the greenbelt, makes a home inside it unaffordable to more and more people who would otherwise have chosen to live closer to the CBD. MORE people end up living outside the greenbelt because that is where they can afford the cost of a mortgage, period. This phenomenon has been well aired in the USA, going back as far as the landmark "Costs of Sprawl" study from Rutgers University in 2000. Part of the research examined how housing prices vary with distance from the regional downtown of each metropolitan area. Although only a few areas were analyzed, the study showed consistently that prices of similar homes tended to decline about 1.2 to 1.5 percent per additional mile from the regional downtown, except where proximity to the ocean had more influence on prices—as in Southern California. Meanwhile, longer-distance commutes added to fuel and travel-time costs by about the same amount per mile in every region. The study also found that per-mile housing-cost savings from added commuting distance were much larger in regions with absolutely very high housing costs than in those with absolutely low housing costs. Therefore, it was MORE LIKELY TO BE ECONOMICALLY WORTHWHILE FOR HOUSEHOLDS TO MOVE FURTHER OUT TO GAIN CHEAPER HOUSING in HIGH-HOUSING-COST REGIONS such as the San Francisco Bay and Boston areas, THAN IN LOW-HOUSING-COST areas....."
(My emphasis)
The advocates of urban growth containment have completely failed to "get" this point - thinking that more control, and regional control, are justified. Actually, the Netherlands "got it" decades ago - they use powers of compulsory purchase to keep fringe land prices low, even when strict containment policies are in effect. That way, they have their cake and eat it too - they get urban containment, AND low urban land costs, enabling their economy to stay competitive, and social mobility to be unhindered. Ironically, the States in Australia abuse similar powers to rip everyone off, possibly justifying it by the same shonky reasoning about classical land economics theory as you helpfully outlined above.
Regarding decentralisation and the prices of land, this IS how it works everywhere in real life. Inflated land prices drive decentralisation, along with lack of roading provision. The decentralisation ends up similar to that that occurs in free markets anyway, only with massive costs imposed on society and the economy meanwhile.
April 26, 2011 10:05 PM
Here is an NZ blogger, Phil McDermott, tackling the "decentralisation" issue in the context of the planning hubris infesting Auckland:
ReplyDeletehttp://cities-matter.blogspot.com/2011/04/restricting-retailing-to-save-cbd.html
A few points, in reverse order.
ReplyDeleteThe link to Phil McDermott's blog makes a reasonable point in light of restrictions on retail development. Unlike the housing market, I have seen evidence of governments constraining retail development using the justification of a 'needs assessment'. Put simply, if there is another large shopping centre in the area they are assumed to be able to supply the needs of the community. Lobbying from existing centres protects them from competition.
However, without some good data it is hard to say to what degree this occurs. If there is a large stock of approved retail developments you would be hard pressed to make the supply constraint argument.
You say -
"...planning and infrastructure provision both drive land prices up; long before as many people as the planners HOPED would live in "x" location, the price of homes at "x" location will have gone up to way beyond what everyone can afford. The result, ironically, is a wealth transfer from the poorest in society, to the incumbent owners of property at the locations favoured by the planners.
Roads, cars, and sprawl, on the other hand, have been powerful forces for the wide dispersion and sharing of gains from economic progress, as well as the major enablers of that economic progress. Rails and high density go naturally with totalitarian government; roads, cars and sprawl go with free democratic society. "
I don't quite understand why new roads don't also raise the value of surrounding land and have the same impact you argue that rail has. Also, I would argue that if you look to 19th century Europe you would say that railway was THE great enabler of growth. They are both simply menas of transport that require large capital investment, so I can't see any reason for your differentiation.
Linking these two substitute means of transport to totalitarian governments is verging on ridiculous.
Lastly, you talk about the price gradient but not the overall price level when talking about decentralisation. The flattening the price gradient is unimportant to the general price level.
Cameron,
ReplyDeleteI think we are close to agreement on the above details at least.
Rail infrastructure is always much more localised in its Real Estate price effects. The "dispersion" of gains from new rail infrastructure depends heavily on the road network enabling "feeder" services. But the road network acts as a much stronger disperser of gains in the case of new road infrastructure. There is far LESS incentive for buyers to bid up the price of land right NEXT to the end of a new motorway (regardless of residential or business), than there is for buyers to bid up the prices right NEXT to a railway station. People who are going to drive anyway (a huge majority) do not mind at all about being 1 or 2 km from the new motorway onramp, but people who want to catch the train DO care much more about distance.
This effect is especially strong at the CBD end. Once again, a business located 1 or 2 kms away from the motorway offramp is going to reap SOME value from it, whereas being 1 or 2 km from the railway station is far more critical. This is why strong constituencies form around CBD rail: even if the costs exceed the benefits, conveniently located CBD property owners will support any proposal that sees them carrying $1 of every $10 of cost (with 'everyone else' paying the other $9), and gaining $1.80 of every $2 of benefit.
Rail definitely started the process by which lower cost land could be brought into the urban economy, raising productivity, improving health, and enabling social mobility and equality improvements. But this created "ribbon sprawl", and still depended on horse-drawn transport to disperse the benefits beyond the immediate vicinity of rail lines. Roads and cars enabled much more efficient contiguous "sprawl". According to Robin Best, in "Land Use and Living Space", the spread of urban areas was MORE rapid in the pre-automotive era, driven by public health and equality concerns and enabled by rails, than in the following automotive era, which brought another order of magnitude of economic efficiency gains. THIS is why there is actually a good correlation between the overall "efficiency" of economies around the world, and the "automobility" and "sprawl" in economies. I wager that China and India and other developing countries are going to fail to match our achievement unless they match our automobility - and they can do this far cheaper than we HAVE.
Cameron,
ReplyDeleteI think we are close to agreement on the above details at least.
Rail infrastructure is always much more localised in its Real Estate price effects. The "dispersion" of gains from new rail infrastructure depends heavily on the road network enabling "feeder" services. But the road network acts as a much stronger disperser of gains in the case of new road infrastructure. There is far LESS incentive for buyers to bid up the price of land right NEXT to the end of a new motorway (regardless of residential or business), than there is for buyers to bid up the prices right NEXT to a railway station. People who are going to drive anyway (a huge majority) do not mind at all about being 1 or 2 km from the new motorway onramp, but people who want to catch the train DO care much more about distance.
This effect is especially strong at the CBD end. Once again, a business located 1 or 2 kms away from the motorway offramp is going to reap SOME value from it, whereas being 1 or 2 km from the railway station is far more critical. This is why strong constituencies form around CBD rail: even if the costs exceed the benefits, conveniently located CBD property owners will support any proposal that sees them carrying $1 of every $10 of cost (with 'everyone else' paying the other $9), and gaining $1.80 of every $2 of benefit.
Rail definitely started the process by which lower cost land could be brought into the urban economy, raising productivity, improving health, and enabling social mobility and equality improvements. But this created "ribbon sprawl", and still depended on horse-drawn transport to disperse the benefits beyond the immediate vicinity of rail lines. Roads and cars enabled much more efficient contiguous "sprawl". According to Robin Best, in "Land Use and Living Space", the spread of urban areas was MORE rapid in the pre-automotive era, driven by public health and equality concerns and enabled by rails, than in the following automotive era, which brought another order of magnitude of economic efficiency gains. THIS is why there is actually a good correlation between the overall "efficiency" of economies around the world, and the "automobility" and "sprawl" in economies. I wager that China and India and other developing countries are going to fail to match our achievement unless they match our automobility - and they can do this far cheaper than we HAVE.