Feasible development, intensification and the PAUP

In my blog of the 22 August 2016 I discussed the concept of feasible development capacity. This concept features in the proposed National Policy Statement on urban development capacity and the deliberations of the independent hearings panel that considered the Proposed Auckland Unitary Plan (PAUP). I questioned the use of the concept to determine zoning envelopes, especially longer term estimates of required capacity. I acknowledge the need to consider how much of the theoretical capacity enabled by zonings may be feasible or realistic, but am unsure if 'commercially feasible development capacity' is the way to go. 

In this blog I want to look at whether the idea of commercially feasible capacity may help determine what type of zoning should be applied, if not how much zoning.

The data in Table 1 is from the Auckland independent hearing panel's report on the recommended PAUP. It shows the total plan enabled capacity by the four main residential zones used by the PAUP, and the proportion of that enabled capacity, that was deemed commercially feasible by the council's model.

Table 1: Number of dwellings

Zone	Plan enabled capacity	Estimated feasible capacity	Proportion capacity "feasible"
Single House	34,445	11,259	33%
Mixed Housing Suburban	312,627	50,966	16%
Mixed Housing Urban	297,834	53,750	18%
Terrace Housing and Apartment Building	325,648	45,929	14%
Total	970,554	161,904	17%

So, not much difference in feasibility across the zones. 

If we take the land area involved in each of the 4 zones and the number of dwellings feasible per hectare of land, we start to see the efficiency of the more intensive zones.

Table 2: Feasible dwellings per ha.

	Area (ha)	Feasible dwellings per ha
Single House	8,761	1.29
Mixed Housing Suburban	12,497	4.08
Mixed Housing Urban	8,211	6.55
Terrace Housing and Apartment Building	2,384	19.27
Total	31,853	5.08

The data gives a hint of the issues at play when considering feasibility and location.

The Single House Zone feasible capacity of 33% is actually not much different to my estimates done 20 years ago for similar types of zones(as outlined in my 22 August 2016 blog), albeit by a different methodology.  Estimating realistic capacity of zoning that involves adding a further dwelling to an existing section is not too difficult to calculate. All you need to know is the number of sections above a minimum size.

Up to about the end of the 1990s, most development in Auckland was by way of greenfields expansion and urban infill.  The infill bit was mostly adding a further dwelling to a site, and was often landowner driven, or undertaken by a development-orientated builder. However, the amount of development realised per hectare of land is not that great.

Redevelopment is now the most common form of urban intensification. This is in the sense of existing housing on a section having to be demolished, removed or shifted to accommodate new housing.  This is a more complex process than infill,  requiring more skills and 'deeper pockets'.   Suitable sites are harder to find, with many sites subject to cross-lease or unit title development. Development timeframes are longer so market ups and downs are more of a risk. The resulting product may not be affordable , for the developer or the purchaser. It is the type of development undertaken by specialists rather than 'part timers'. Access to finance is critical.  These factors are not taken into account by the feasible development model, and arguably further reduce what may be 'feasible'.

Stepping away from the Single House Zone, for the other three zones, feasible capacity drops by close to half, to be in the order of 14 to 18% of total enabled capacity.  This reflects the extra 'costs' involved and the finer balance between costs and returns. This is a fairly tight band of feasibility across the three 'intensive zones'. There is no real advantage of one zoning type over another in terms of feasibility. So in the future, could we run off some sort of 15 to 20% figure to determine how much redevelopment capacity is realistic in the short term?

The number of commercially feasible dwellings per hectare of land does vary by type of zone. But while the number of feasible dwellings per ha of land zoned terrace housing and apartments is much higher than the other zones, the pool of land is much smaller. Arguably, finding development sites will be even harder in this smaller pool.

So do we focus on making the pool of potential development opportunities as big as possible? This is a pretty simple way of looking at the issue - just make the whole pie big enough so that hopefully there will be enough to satisfy everyones needs.  Ideally there would not be  a large  gap between enabled and feasible capacity. A big gap suggests room for unintended outcomes, especially if actual development patterns do not match the estimate of feasible development capacity. Development may end up concentrated in one area, rather than nicely spread around, and make it hard for infrastructure providers to work out where to put their effort. Equally though, plenty of 'slop' in the system should enable more room for market forces.

I do think the concept of development feasibility has some validity in determining where different types of intensity should be located, and hence plans being able to better correlate enabled capacity with feasible capacity.

The graph below groups the urban Wards in Auckland into three rings: Inner, middle and outer. These rings roughly accord with the growth of the city up to the mid 20th century (inner); post war to about the 1980s (middle); and last 30 years or so (outer).  The feasibility of different house types is then grouped by these three areas. The data provided is based on the zoning set out in the recommended version of the PAUP, and so the data is not a picture of what the market would do, unencumbered by zoning controls.

Also note that the three rings are not of equal size. The middle ring is by far the largest in terms of the number of existing dwellings.

Not surprisingly in the inner ring of suburbs, it is apartments that are the most viable type of development, by a long shot. In the outer suburbs, there is more of a mix. The way that the council's model is structured may mean that it chooses low rise apartments in these areas, but in reality it is quite possible that it will be terrace housing, duplexes and town houses that will be developed.

Figure 1: Feasible dwelling types by urban location 

The heavy lifting in terms of feasible capacity is undertaken by the middle ring suburbs. Not surprising given that in Auckland's case, so much of the inner ring are heritage areas. But how much capacity needs to be enabled to get to this amount of feasible development? 

At the back of all this data is some fairly simple relationships between land values, construction costs and sale prices. As those parameters change - particularly as land values rise - then the financials of development alter and shift towards apartments away from stand alone houses, even when apartments cost more on a per square metre basis.

This can be seen when feasible capacity is compared to enabled capacity, in the three rings. Using slightly different data from that in the figure above (data based on an earlier version of the PAUP), the graph below shows the existing dwellings (at 2013), the plan enabled capacity and estimated feasible development.

Figure 2: Dwellings by Urban Location 

In the middle ring, only 10% of enable dwellings are feasible, while in the inner ring, that figure is closer to 30%.

Three things spring to mind in terms of planning principles:

1. Does feasible development tend to be scattered development, rather than concentrated development?
2.  Is density around amenity a better principle that density around transport hubs? 
3. The urban system is continually  adjusting - should zoning also be able to adjust automatically as land values alter?

The first and second questions relate to a long-held planning principle that clustered intensification around a town centre or along a transport route is the way to go. There were four main reasons for this. The first relates to the amenity issues associated with intensification - 'saving the suburbs'. Intensification should be grouped or concentrated so as to not see existing suburban fabric substantially modified. The second and third relate to the transport effects of intensification. The inevitable effect of more people is more traffic - locate the people close to shops and transport hubs and hopefully that impact is lessened as rates of walking and cycling and public transport use will be higher. The extra population close to a centre would also help with sustaining the viability of that centre. More local services means a more attractive place to live and less need for travel out of the centre. Finally, infrastructure providers (particularly Watercare) had a fair idea where infrastructure upgrades are likely to be needed. All in all, a winning sounding strategy.

But consideration of feasible development has considerably diluted this strategy.  The amount of development around nodes that is 'feasible' is less than what may have been hoped for. So to find capacity, either the nodes need to be a lot bigger, or the suburbs need to be opened up for more growth. The end result is that in its search of capacity for 30 year time frame, the PAUP has tended to spread growth across the urban area and while the intensification enabled by the Terrace Housing and Apartment Buildings zone is important for capacity, it is no longer the main source of feasible capacity. You could say that this is the major change between the notified version of the PAUP and the decisions version:  concentrated versus spread development.  This is particularly so for the middle ring suburbs.

The consideration of commercially feasible capacity also tends to push consideration of upzoning to areas of higher land values - coastal areas and inner city areas. If we take the urban Wards and look at where the highest proportion of enabled dwellings are feasible dwellings, then the four top Wards are:

 

Devonport - Takapuna	36% of enabled capacity  is feasible
Waitemata	34%
Hibiscus&Bays	26%
Orakei	20%

Amenity drives land values, which in turns drives intensification. But these areas are constrained by a number of factors (not all of which are to do with the environment), and so to get the capacity figures up, the middle ring suburbs have had to 'expand'  But the middle ring is the least efficient in terms of turning enabled development into commercially feasible development.

On the last question - things change, so should zoning - how can zoning change more easily? 'Fungible' zoning (the ability for zoning to adjust automatically) is something to look at in more detail in a later blog.