This is a follow up guest post from Geoff Cooper, Director of Economics at PwC, on work he’s done looking at viewshafts. You can read the first post here.
I’ve had a lot of feedback on the results of our study into Viewshaft E10. A common one, including from readers of this blog, asks what the full cost of E10 might be if we include the city fringe and wider economic benefits. I eluded to this on Nine to Noon a few weeks ago, when I speculated the full cost to the area could easily be double. In this post I am going to show how incorporating these issues could lead to a policy cost as high as $3.1 billion.
A. Estimating the impact of E10 on the City Fringe
Roughly half the land impacted by E10 falls outside of the city centre. Much of this is in Newton and Mount Eden. In a perfect world, we could include this land in the cost estimate by multiplying the area by the policy effect (already calculated at -2,419 per square meter). However, this requires an important assumption, so here it is clearly stated: We need to assume E10 is the binding constraint to development.
Is this a credible assumption?
There are basically three zones under E10 in the city fringe:
- Town Centre zone: Provides for “buildings of between four and eight storeys, although there may be special circumstances where other building heights are appropriate”.
- Mixed Use zone: Provides for “a range of possible building heights depending on the context. Provisions typically enable heights up to four storeys. Greater height may be enabled in areas close to the city centre, metropolitan centres and larger town centres”.
- Residential zone: Provides for low density residential.
The first two zones provide ample discretion, which leaves us in the hands of individual planners. But the wording gives good reason to think E10 is in play, especially since Newton is so close to the city centre (note: if E10 is not binding, it is a policy attribution issue). The residential zone is clearly more binding than E10 however. If we remove E10, we won’t see more housing in Mount Eden – so we exclude all land in the residential zone.
Accounting for other veiwshafts
Four other viewshafts interacting with E10 in Newton are an additional complication. To account for this, we split the city fringe into two cost categories:
- Exclusively constrained by E10: ~30% of affected land
- Overlapping viewshafts: ~70% of affected land. Cost attributable to E10 or another viewshaft.
Stripping out public land and low density residential leaves an additional 366,144 square meters under E10. This equates to a cost of $886 million with 30% attributable to E10 (270m) and 70% attributable to either E10 or one of the other four viewshafts (620m).
Do we have the E10 land impact right?
At this point you might be sceptical about the E10 policy impact on land ($2,419), which came from our study. Do we really see this impact in Newton’s land markets? There is no clean slicing of the city in Newton because the boundary of E10 merges with Viewshaft E16, which in turn merges with a residential zoning area. That means no clear, decisive discontinuity to evaluate. At least not a first glance …
Life without E10 on the City Fringe
If you take a close look at the view of Mount Eden as it was designed, you’ll notice a building piercing E10 (see Figure 1). It’s the newly renovated SKHY building on 5 Hohipere Street, built before E10. It’s probably a terrible annoyance for E10 advocates, but for the rest of us, it’s a unique and rare glimpse into urban life without E10 (see Figure 2). We can think of land under this building as a control (unconstrained from E10). The land around it is the treatment (constrained by E10).
The land value of 5 Hohipere Street is about $18.4m and its land plot size is 2,083 square meters. So land per square meter is $8,831 (see Figure 3). In comparison, adjacent properties are between $4,330-$5,313, a difference of around $4,000. It seems reasonable to attribute this to E10 because the land plots are otherwise similar: both are in the mixed use zone, with largely identical locational amenities. The difference is 70% higher than our study (which we use in the fringe cost estimate) and is striking evidence of the effect of E10 in the city fringe. The policy is clearly preventing people from accessing the city fringe. Land values are lower as a result.
B. Wider Economic Benefits: The Economic Impact of Urban Exclusion
Next we consider Wider Economic Benefits (WEBs). I spent much time debating the inclusion of WEBs in the City Rail Link business case review back in 2011. While the City Rail Link brings people into the city though, E10 locks them out. It’s time to start applying this widely agreed and tested source of benefits to our city’s regulatory policy.
Description of WEBs (and agglomeration) in the CRL business case:
Wider Economic Benefits, CRL Business Case, 2015
Agglomeration economies have been well documented over the past century. The continuing intensification of cities proves the increasing relevance of agglomeration economies in the 21st Century. Agglomeration economies flow from business being able to interact easily with each other. They learn from each other and innovate to develop new products and open up new export markets. Workers presented with a large pool of jobs should be better matched to their roles and hence become more productive. There will be informal transfer of knowledge between skilled workers further boosting innovation and productivity in the long run.
As the economy of Auckland becomes more and more ‘knowledge-intensive,’ agglomeration economies should become increasingly important. The CBD is the preferred location of these knowledge-intensive industries. Effectively, growing the Auckland CBD will help generate more income for businesses located there and more taxation revenue for Central Government.
The problem we have is WEBs are not a straightforward appendage to land value changes. According to the UK Department for Transport:
“Land value uplift will capture any impacts which are capitalised into land values. It could potentially capture any of the following impacts: user benefits, land market distortions and other wider economic impacts, such as agglomeration economies that occur within that development.”
U.K Department for Transport TAG UNIT A2.1
So we can’t add WEBs for risk of double counting. How much double counting? We don’t know. When employment increases in the city, it raises the productivity of all employees. That’s what makes it an externality. So relaxing E10 could lift land values everywhere in the city centre and city fringe – but our evaluation only considers uplift under E10, making it a partial land value assessment only. This is an analytical area in desperate need of attention. Land use regulation (and section 32 analysis) is not held to the same standard as transport appraisal, despite the glaring similarities.
“Removing E10 so a greater number of people can live and work in the city is the transport equivalent of reducing travel times to a few minutes, but without the infrastructure bill.”
What might be the magnitude of WEBs?
We can use the CRL business case to get a sense, since the CRL addresses accessibility and productivity in the very same area that E10 excludes people (these two policies actually oppose one another because E10 directly impacts the CRL stations). WEBs are 37% of CRL conventional benefits. By comparison, other urban transportation projects vary from 16% (Melbourne East West Rail Package) to 56% (Crossrail). Using the CRL as the point estimate, WEBs from removing E10 could be up to $901m. Because of the double counting issue, we include this as a range, 0 – 901m.
Three Billion Dollars Later …
After taking account of sightline benefits, the net local policy cost is between $1.6-3.1 billion. Methodological gaps increase the uncertainty of our estimate (the lower end is not dissimilar to our original estimate) – and highlight an area of future work. In a perfect world, we would have a much larger sample size to use, a more established methodology for including WEBs into regulatory appraisal and a sharper focus on policy attribution. But this is why the total cost of E10 could be far higher than $1.4bn. As I have said before, I cannot think of another single policy that is constraining our city centre more.
|Cost to City Centre (original study, bns)|
|City Fringe: Exclusively constrained by E10|
|City Fringe: Overlapping viewshafts|
0 – .62
|Wider economic benefits (incl agglomeration)|
0 – .9
|Total E10 cost|
|Less: Sightline benefits|
|Total net cost (bns)|
1.6 – 3.1