In Auckland, land use regulations are said to account for 56% of the cost of a house. So says the report “Quantifying the impact of land use regulation: evidence from New Zealand” (see note 1).
The 56% equals about $530,000. It is admittedly only one measure of the costs of land use regulation that the study uses, but it is the number that the media has picked up on. Not surprising, given its size!
In this blog I want to try to unpack the numbers and understand how the 56% was estimated.
The percentage is derived by taking the median house price, subtracting the cost of building a house and then subtracting the assumed cost of land based on an estimate of what land may cost if it was freely available. The residual is the land use regulation bit. At least that is what I think is going on.
I think the basic idea of the study is that the cost of a good (in this case a house) should always tend towards the cost of production in a competitive market. If there is a big enough profit being made over and above production costs, then business people (i.e. landowners and developers ) will spy the super profits and leap into the market, produce more or better or cheaper products, helping to bring down prices. So if there is persistently a big gap between the costs of producing a good and the sale price then there must be either a blockage to new entrants or some form of tax that pushes up prices, or both.
The study assumes that the development market is competitive, so the gap between the costs of production and sale values must be a tax. The point that there may be a speculative element in the value of houses doesn’t seem to figure.
Where developer profit fits into the equations is also hard to work out. Even Edward Glaser, who the study refers to a number of times, accepts that production costs should include an allowance for reasonable profits. His estimate is that builders in an unregulated market should expect to earn a 17% over purely physical costs of construction (note 3). The 17% is close to the 20% used in the analysis of feasible development capacity that drove so much of the Unitary Plan rezonings.
Where developer profit fits into the equations is also hard to work out. Even Edward Glaser, who the study refers to a number of times, accepts that production costs should include an allowance for reasonable profits. His estimate is that builders in an unregulated market should expect to earn a 17% over purely physical costs of construction (note 3). The 17% is close to the 20% used in the analysis of feasible development capacity that drove so much of the Unitary Plan rezonings.
Even where there is a gap between costs and prices that can be attributed as a tax, this is not necessarily a bad thing. Having a tax sounds bad. But a tax is not always a bad thing. That tax may pay for useful goods and services, or it may reflect costs passed onto others - like the environment or future generations. As typical of these costs of regulation studies, it is always the costs side that gets estimated, not the benefits side. But a tax of 56%???
What is in this ‘tax’? Here the study gets a bit hard to follow. The term ‘tax’ seems to apply to all of the gap between costs and prices. The study mentions that the gap may be the result of geographic restrictions that limit land availability, time lags between demand and supply (ie a demand shock) that push up prices in the short term, infrastructure availability and Building Code matters, as well as land use controls. So quite a mix of different things, not all of which are related to land use regulations. Nevertheless, these all get lumped together under the heading “land use regulation tax’.
How was the $530K calculated? In round numbers, the study uses an Auckland median house price of $950k. Subtracted from this is the estimated cost of constructing a house of $360k, then less $60k for the land. This equals an unexplained ‘residual’ of $530k. This is the estimated cost of regulation, or 56%.
Now the big assumption here is the $60k for the land, but there is also an assumption around build costs and developer's profit. Let’s start with the land cost.
I don’t know what size section the study assumed. I think it was about 700m2, or about $86 per sqm. Sounds very cheap. QV says the median section price in Auckland is around $500k, but there is a big range (note 2). If we assume a median section size of 500m2, then we get a median value of around $1,000 per sqm. How come the difference - $86 per sqm versus $1000 per sqm? Is the difference all down to regulation?
The study attempts to estimate the ‘basic’ cost of land by means I can’t fully understand. The study uses regression analysis to strip out different factors that influence land prices, including distance from CBD, slope and views and size and age of housing stock. What is left over is the $86 per sqm as a ‘average’ land value estimate. The only way I can makes sense of the $60,000 figure for a lot is to assume it relates to undeveloped land. If you get out of urban Auckland to rural land away from the urban edge, then land might be $80,000 to $100,000 per hectare. Divide that land up into 10 lots and you would get a 1,000m2 section for around $9,000 or $9 per sqm, or if you could get them in with no roads, say $6,500 for a 700 sqm lot. Add to that the costs of providing basic access and services (at say $70,000 per lot) then the cost of a lot is $76,500. This is a bit more than the $60,000 used in the study, but not too far off.
But you might say, where is the bulk infrastructure? This is where the study of the cost of land use regulations gets a bit odd. The following sentence is tucked away in the body of the report:
Moreover, these regulations are often a function of the broader urban planning system, including infrastructure funding.
So the cost of providing bulk infrastructure is a land use regulation? Perhaps it is in one conception of what is a ‘regulation’. You do not need water, wastewater and stormwater reticulation nor an arterial road or motorway nearby to live in a house, provided you (and your neighbours) are happy to live with the inconvenience involved. Perhaps out in the country, but in a city? Network infrastructure is needed for cities to function safely, something that the Victorians worked out (even the Romans seemed to understand this point). So can we separate out the infrastructure bit from the land use regulation bit?
My previous post on the costs of greenfields growth suggested that local and bulk infrastructure costs for a greenfields subdivision could be in the order of $200K per lot, being about $80k for costs within a subdivision and $105k for network infrastructure extensions. Currently only some of that $105k of network extensions gets recouped directly from the user. The rest is a subsidy from the public. This subsidy must get reflected in the land price. Add the $105k onto the section price of $76,500k and you get $181,500. This number is not too far off what sections costs in less busy places like Napier and Palmerston North, although in those areas it is probably easier and cheaper to extend network infrastructure.
Back to Auckland, add build costs of $360K to $181k and we have a total of $541k.
Back to Auckland, add build costs of $360K to $181k and we have a total of $541k.
So is the actual land use regulation bit (assuming the infrastructure is necessary for society to function) really more like $950K-$360K-$180K = a residual of $410K. This is still 43%, quite a hefty amount.
Let’s look at the build costs. The study says $360k for the house. How big is the house? This is a bit hard to work out too, but I think it is 200 sqm. So build costs are roughly $1,800 per sqm. But are these all the build costs? If we go to MBIE’s trusty development feasibility calculator and look at the numbers for “building development’, then my quick add up of costs per square metre gets to closer to a total cost of $2,200 per sqm. This includes site preparation, building costs, driveways, design fees and project management. Council costs like building consent fees, development contributions and the like are not included. Neither is developer profit.
Now some of that might be regulation. You do need to design to the Building Code, for example. Need a driveway? But these costs are either pretty marginal - most people want a driveway - or necessary to keep people safe. So if we run with $2,200 per sqm, then our median house costs $440,000 to build. As a cross check, using Stats NZ data on Building Consents issued for houses in Auckland over the 12 months to June 2017, 5,251 houses with an average floor area of 232 sqm cost on average $455,308 to build. These figures are not for townhouses or apartments, but the study of the cost of land use regulation seems to be focused on the costs associated with building a house.
If we take $440,000 as the average build cost, plus the $180k section, then the land use regulation ‘bit’ falls back to $330k, or 35% of the median house price. The other option is to drop the size of the house down to keep the unexplained part of the cost higher. It is easy to manipulate things one way or another.
But is it fair to compare the cost of a new build with the median price across a large city where the median is set in part by many expensive houses sold in pricier inner city suburbs? And what about normal developer profit?
If we look at greenfields areas out on the edge, then average house prices are more like $800k (with section prices about $400,000 to $450,000). Take off the construction costs and the (full serviced) section price, then the residual is $180k, or 22% of the value of the house. If we allow for 17% developer profit in the mix, then costs rise to $695k for a median house, and the residual falls back to $105k, or 13%. Is this then the cost of the actual land use regulation bit, or an inefficient market, or geographic constraints? If it is all down to a regulatory tax, is it worth the benefits? Would you pay an extra $100,000 to live in a suburb where you knew that zoning would protect your amenity and the natural environment was sustained?
I know that house prices are too high, but saying that land use regulation adds 56% to prices is a distraction, at best.
I know that house prices are too high, but saying that land use regulation adds 56% to prices is a distraction, at best.
- http://www.superu.govt.nz/publication/quantifying-impact-land-use-regulation-evidence-new-zealand
- https://www.qv.co.nz/property-trends/residential-sales-prices
- https://www.brookings.edu/research/reforming-land-use-regulations/
