Thursday, 21 January 2021

Urban design and reform of the RMA

Some more thoughts on the replacement to the RMA and the possible role of urban design in the new Act. A central feature of the reform panel's report on the RMA is the call for 'urban quality'. Does the call for urban quality not place urban design in the centre of the action? 

First up, urban planning has always responded to the pressures of the day:

  1. The Victorian city and its densely packed housing close to polluting industry. The planning response: separate out land uses and control housing density
  2. The post war city of motorways, suburbia and subsidised sprawl - response: better protect the natural environment (streams, bush, landscapes), develop sub regional hubs, structure planning
  3. The post industrial city of the late 20th Century of infill, redevelopment, gentrification and consumerism - heritage zones, design controls, tree protection, view shafts, mixed uses (which is where urban design started to come into focus).

The interesting thing is how each wave has added layers of response onto previous layers. Is the  panel’s call for urban quality a further iteration of this trend of addition, or is it a replacement of  accumulated layers?

But either way are quality urban environments what we need to respond to today’s urban pressures and issues? Constrained supply of housing and too many road blocks to urban redevelopment are seen to be the pressures of the day, pushing up house prices. In one view, in response market forces should become more prevalent in the determination of land use location, density and mix. This will help ensure sufficient and adequate supply of housing, apparently. The call then goes out to make sure it is quality development - where ever and what ever it is.  This will keep the NIMBYs at bay. But does 'quality urban environments' sound like a very middle class aspiration? 

But why do we think market processes will not deliver quality. And what happened to overall urban functionality, equity and long term efficiency?

I'm  a bit worried that the RMA reforms will play out along a simple dichotomy:

Better protect the natural environment, but loosen up on the urban areas. 

People like simple solutions and this one always sounds fair and balanced. More rules and intervention in one sphere, but less in another.

But loosen up in the urban sphere? Loosen up may be right for some circumstances. It is time to have a look at the accumulated layers, and perhaps remove some, but new layers need to be developed to deal with future issues.

The trick to any RMA reforms will be in defining the limits of planning in an urban context. There are defined principles to limit market processes as they relate to the natural environment (bottom lines, fair allocation within limits, managing use of public resources), but come to the urban environment and there are much less clear cut boundaries. The RMA attempted to control urban planning by limiting intervention to the management of externalities. But that never limited constant calls for more planning to control and promote outcomes. Defining the role of urban planning has to involve some sort of consensus on how far it can go.  The call for ‘quality’ is a simplistic response to this need, and will lock debate about urban planning into a very narrow field.   

In previous posts I have looked at urban design and how it relates to plans developed under the Resource Management Act. Urban design has come along way in the past 20 years, but has struggled to find a ‘home’ under the RMA. This is because the qualities and characteristics of urban environments that support social and economic well being are much wider than just managing negative externalities. They are also much wider than a simple notion of 'quality'. Good urban environments are complex:

  • People’s interactions with the built environment is physical, emotional and cerebral
  • Interactions are both positive and negative.
  • They tend to be small-scale, cumulative and accumulate
  • The public-private interface is critical
  • Open space, mixed uses, housing choices, connectivity are all critical to well functioning urban environments.

If anything, todays urban pressures are all about addressing a narrowing of urban choices (the downsides of a more market - or less government - approach?). Equity is being eroded, long term consequences are being downplayed and choices narrowed as private space takes over from public space. There is a growing concentration of urban resources in fewer and fewer hands and a longer tail of people and neighbourhoods being left behind. The old technique of a strong public sector to counterbalance the pressures of a strong private sector no longer works. Calls for more competition to help rein in the concentration of private interests and improve market-based choices tends to see concentration increase, rather than get dispersed.  More housing supply is not translating into more supply of modestly priced housing, for example. 

 How do these thoughts match up with the proposed replacement RMA? Lets start with the purpose:

(1) The purpose of this Act is to enhance the quality of the environment to support the wellbeing of present and future generations and to recognise the concept of Te Mana o te Taiao. 

Wellbeing is wide ranging: 

 In this Act wellbeing includes the social, economic, environmental and cultural wellbeing of people and communities and their health and safety.

Promotion of quality urban environments sounds like a more comfortable starting point for management of urban environments than avoidance and mitigation of adverse effects.  Depends on the outcomes of course. But can a quality urban environment deliver a healthy and safe urban environment for all, for example? Is the reference to 'quality' too confining? Shouldn't a safe and healthy urban environment that can be enjoyed by all be an objective in itself?   So too with housing choice and affordability.  

To achieve the outcome, the review report on the RMA suggests a shift towards outcomes. 

(2) The purpose of this Act is to be achieved by ensuring that: 

(a) positive outcomes for the environment are identified and promoted; 

(b) the use, development and protection of natural and built environments is within environmental limits and is sustainable; and 

(c) the adverse effects of activities on the environment are avoided, remedied or mitigated. 

The obvious problem is that positive outcomes are to be ‘promoted’ while negative effects are avoided or mitigated. Avoidance and mitigation are a lot more definitive terms than ‘promotion’. How are positive outcomes to be ‘promoted”? Promotion involves some sense of active facilitation, but perhaps no more than some marketing. Are positive outcomes to be promoted by enabling them to occur, but not requiring them? 

But then, what is a positive outcome? Is it an improvement over the current state? Is more choice a positive outcome? 

What is meant by the word 'environment'?

(3) In this Act environment includes– 

(a) ecosystems and their constituent parts; 

(b) people and communities; and 

(c) natural and built environments whether in urban or rural areas. 

So promotion of positive outcomes can include something positive for a community? Is a shared sense of place/identity through the coherent design of the built environment a positive outcome (like a Special Character Area, but then dont these special character areas lock up supply options?). 

To assist in achieving the purpose of this Act, those exercising functions and powers under it must provide for the following ‘outcomes’:

Built environment:

(f) enhancement of features and characteristics that contribute to the quality of the built environment; 

(g) sustainable use and development of the natural and built environment in urban areas including the capacity to respond to growth and change; 

(h) availability of development capacity for housing and business purposes to meet expected demand; 

(i) strategic integration of infrastructure with land use.

So does promotion of environments that support urban well-being get reduced to "lots of  capacity for ‘quality’ development?". 

I think this bit of the possible Act needs serious consideration. 

There needs to be some form of definition or explanation of features that contribute to the functionality of the built environment, of which quality is one aspect. This is very wide ranging, spanning from sense of safety to aesthetics with everything in between. 

The built environment is also a vast space covering public, private and semi public/private spaces. Should the reference to qualities be more confined to the public components of the built environment (streets, open spaces, public buildings) or perhaps to the public areas and their interface with private spaces? 

There is a danger that the more comprehensive attempts to incorporate positive outcomes become, the less defined the outcomes become and less support is provided to their incorporation. The alternative is to take up the language of bottom lines and limits and say that developments in urban environments should incorporate some basic features that support quality public environments and do not detract or reduce these values.   

And by the way,  the strategic integration of infrastructure provision with land use and development has to be  linked to funding arrangements. Structure plans and rezonings should only be advanced after infrastructure needs, costs and funding obligations have been determined and agreed. Otherwise, there is not going to be any 'integration'.  

Sunday, 22 November 2020

Working from Home Part 3

In my previous post on working from home I speculated on the residential locations that might be affected by a large take up of people working from home. There is conjecture that freed from having to travel daily into central workplaces, and the associated long commutes and high house prices, many office-based workers will spread out into the countryside as they look for a better home-work balance. Of course, to make it all work, most people in the household would need to be able to work from home; not much use relocating to the countryside if one member of the household can work remotely, but the other is tied to a desk in a central city workplace. The down side of more working from home could be a quieter central area and reduced passenger transport use. 

Census data suggests that many current ‘work-from-homers’ live in high amenity suburbs in the central Isthmus and along the eastern seaboard of the city, and so may see little point in shifting to somewhere else. But maybe newer households to be formed by people that can work from home will think differently about where to live in the future, especially if housing is cheaper in towns further away from the main urban area? In particular, will small towns within 2 hours drive of Auckland see an upswing in demand? Places like Muriwai, Warkworth, Matakana, Waiuku, Raglan and Waihi Beach? Why two hours? It is a doable trip in a day if you have to go into work one day a week, or for an in-person meeting. 

It is early days and any trend towards the growth of 'satellite' townships may be a few months off.  The question of how big the trend may be remains unclear. 

Census data suggests that outer lying townships have higher rates of working from home than the main urban area. 2018 data suggests that about 12% of workers who live in Warkworth work from home, up from the 9% for the region as a whole.  

Google mobility data (note 1) gives us another perspective on the impact of Covid 19 on the city and people’s movements, and with it some possible clues as to what may happen.

The data shows how visitors to (or time spent in) categorized places have changed over the past 6 months or so, compared to a baseline, being the median value from the 5‑week period Jan 3 – Feb 6, 2020.

The mobility data is created with aggregated, ‘anonymized’ sets of data from users who have turned on the Location History setting on their mobile phones, which is off by default. The data is presented as the percentage change from the baseline value. 

Data is available for the Auckland Region as a whole. No sub regional data is published. 

Data is presented for six types of activities:

  • Retail and ‘recreation’
  • Supermarkets
  • Public reserves
  • Transit stops
  • Workplaces
  • Residences. 

First up, the 'retail and recreation' catrgory is useful to look at to see how the data is presented. The graph below shows the percentage change in visits to retail places in the Auckland Region, from the per covid baseline data. The pattern of a steep fall off in visits to retail premises in the first lockdown is clear, along with a smaller fall in the second lock down. 

(Note: In all the graphs, the data is presented as a % change on the 'baseline' and is for the Auckland Region - as defined by Google)

The google data suggests that visits to retail premises are back up to pre covid levels. However, the data can't tell us about visits to specific retail areas. 

If we compare the Auckland region with the Wellington and Canterbury regions, we can see the dip in the Auckland region in the 2nd lock down, but not Wellington or Canterbury, so the data looks pretty good. 

Before I look at the data on workplaces and residences, spare a thought for Melbourne. The below graph shows Auckland and Melbourne. 

Turning to whether the data shows any trends as to working from home, the data on visits to workplaces and residences tend to be a mirror image of each other. Visits and time spent in workplaces in the Auckland Region took a big dip during the first lockdown, and then another, lesser dip in the second lockdown. 

The one off, downward spike in late October was Labour weekend. 

The baseline data period used in the analysis was over our summer break (January), so the baseline may be an incorrect measuring point for workplace visits. Certainly looking at the graph, the period mid February to mid March (pre lockdown) was about 20% above the January baseline. If we take the mid Feb to mid March period as being the relevant baseline, then mid October is about 20% below the early February period, with workplace visits more akin to a ‘holiday period’.  The period between the two lockdowns is also at this lower level, suggesting a more than temporary reduction in workplace visits. Having said that, it is not clear from the google data if visits to tertiary educational facilities were counted as workplace visits. 

Turning to residences, the baseline data seems a better starting point, with the mid February to mid March period about the same as the baseline period. Time spent in residences was 5 to10% higher than the baseline during the period between the two lockdowns.

The smaller increase in time spent in residences compared to reduced time spent at workplaces is possibly due to the overall larger number of people likely to be at home, once students are added in. 

The big surprise in the Google data is use of ‘transit stops’. Use of transit is well below the baseline, at least 40% down. Is this the flow on effect of the 20% reduction in workplace visits? Perhaps it is reasonable to say that many of the people who would be working from home were public transport users who would likely benefit the most from not having so long to commute in the morning and evening. 

Google says that Transit Stops covers subway stations, ports, taxi stands and highway rest stops. So a mixed bag.

To complete the picture, what about parks and open spaces? Surely a big rise in visits during the lockdowns? But the figures suggest otherwise. Maybe the cancellation of organised sport is reflected in the figures. 

A reduction of 20% in workplace 'visits' as recorded by google may mean that all workers are working from home for one day a week. But it is likely that working from home will be concentrated in certain groups. NZTA (see note 2) suggests that in Auckland, in November, about 18% of workers are mainly working from home, well down on the peaks during the lockdowns, but higher than pre-lockdown (when about 9% of workers were working from home - which is close to the census data).  

So what if an extra 5 to 10% of future workers did think that they could work from home, and therefore live outside the Auckland urban area?  

Forecasting employment is difficult in the current circumstances. Between 2015 and 2020, the Auckland region added about 110,000 jobs. This is a fast rate of growth. If the next 5 years saw a similar growth, then about 10,000 workers may decide to locate elsewhere. This may equal about 6,500 dwellings. This may mean big growth for little towns, but not much of a difference for Auckland.  

But what happens if, say, 8% of existing workers decide that the accpetance of remote working means that they can shift out of the urban area? In 2020, there were just over 807,000 jobs in the region, so lets say 686,000 are in the urban area (85% of total). 8% of this number is 55,000, or perhaps 84,500 people. This  means about 28,000 dwellings may come 'free'.  

As for the Central Area of Auckland, the below graph is from the Heart of the City website (see note 3).  It is based on real time counts of pedestrian traffic near the foot of Shortland Street. Shortland Street has a grouping of high rise office blocks. Data for the period 17 to 21 November suggests an average predestrian count of about 250 people per hour in the mid afternoon, well down from the 700 people over the same time last year.  The twin mountain peaks of lots of people out at lunch time and after work are replaced with a smooth low rise 'hill'.  

The pedestrian data and public transport use data suggests a major shift in land use patterns within the city may be underway. Equally, it may be that the Auckland urban area sees a period of population decline over the next few years. 





Friday, 16 October 2020

Has the housing deficit gone?

The RMA is again being blamed for soaring house prices. There is a ‘shortage’ of houses, with this shortage caused by restrictive zonings and district plans that are slow and cumbersome to change. But is it all a bit of a supply myth? 

For example ASB bank’s recent flyer on housing says that a “second factor we believe is behind the housing resurgence, a more severe housing shortage. It now looks as if the shortage that developed over the past five years was larger than most appreciated. We’ve updated our estimates of such to reflect Stats NZ’s recently upgraded population estimates. The upshot is that NZ’s long running residential construction under-build has produced a shortage of around 60-65k houses, around twice what was previously assumed”.

But as I have pointed out a number of times (as have others), a lot depends on how you calculate the shortage, and what you assume to be a reasonable response between demand (population growth) and supply (a dwelling being built). 

Previously I have looked at issues like:

  • Possible over estimaiton of migration flows and population growth
  • The composition of demand, and the extent to which recent population growth has been driven by migration and uncertainty around the intentions of younger people on worker visas
  • The lag between population growth occurring and houses being built. Evidence is that there is at least a two to three year gap between a building permit being issued and a completion certificate being signed off.

Time to have another go. 

Statistics NZ have updated their yearly estimates of population growth in the light of the 2018 census. So I thought it would be good to re-run the numbers on demand and supply. I have done this for the two ‘hot’ property markets of Auckland and Queenstown.  

Stats NZ provide an annual estimate of resident population for all Council areas. The most recent set for 2020 says that the estimated resident population (ERP) of each area is based on the 2018 Census usually resident population count, with updates for:

  • residents missed or counted more than once by the census (net census undercount)
  • residents temporarily overseas on census night
  • reconciliation with demographic estimates for ages 0-14 years
  • births, deaths, and net migration between census night and the date of the estimate.

The estimated resident population is not directly comparable with the census usually resident population count because of these adjustments.

Stats NZ have completed an intercensal revision for estimates between June 2013 and June 2018, to reconcile with the estimated resident population 2018-base. This has seen some downward revision of previous estimates.

Based on these recent estimates I have attempted to estimate housing demand and supply over the period 1997 to 2020. A long time period is needed, as plannnig is a long term process.  

As previously noted,  a range of assumptions need to be made. 

First up, demand for housing is based on annual population growth divided by the average number of people per occupied dwelling. The average number of people per occupied dwelling can be calculated at the time of each census, as follows: 

Table 1: Auckland and Queenstown dwellings and resident population




AK - occupied dwellings




QLDC - occupied dwellings




AK - resident pop




QLDC resident pop 




This data gives me the following average number of people per occupied dwelling.

Table 2: Auckland and QLDC people per occupiued dwelling 













Some people at this point suggest that the rise in the average number of people per dwelling is not an independent variable. It is a sign of rising unaffordability as people share living spaces due to high housing costs, when the trend should be for fewer people per dwelling (more choice and more space). Others have pointed out that in Auckland at least, other changes like a younger population and possibly more multi-generational households formed by migrants helps explain the increase. 

For example, for Auckland, there has been a dramatic change in the composition of growth, and with it likely demand for housing between the 2000s and 2010s.  The graph below shows the change in the number of residents by age band, from 1996 to 2018. Through the mid 1990s to mid 2010s, the dominant group were the high income earning 40 to 64 year olds, many of whom would be looking for a ‘new home’. From mid 2010s, the largest group by far are the younger 15 to 39 year olds, many of who may be living at home or in flatting type situations, or after a second (or more) hand home. 

Figure 1: Demogrpahic change - Auckland Region 

This change in demographics must change housing markets, reducing demand for new homes to buy, slowing the flow of homes into the stock of housing, while at the same timing raising the demand for existing houses. 

To determine demand, I have used the above data from the census on people per dwelling. Spreading out these rates on an annual basis leads to an estimate of demand for dwellings. 

Supply is based on building permits issued. Building permits issued does not necessarily accord with the number of dwellings that are consented by way of RMA processes. There is no reliable way of counting how many ‘RMA consents” for new dwellings are issued each year. It could be that dwellings enabled by the planning system in each year is way ahead of building consents issued. 

For this exercise, I need to run off building consents. Supply of new houses is based on the number of building consents issued in the two years after the estimated demand. That is, I have assumed that there is a lag between population growth occurring and house plans being drawn up, building permits obtained and houses built. In other words, the number of houses built in any year will likely reflect what demand was at least two years prior. So matching annual population growth with building permits issued in that same year is miss-leading. 

However, some building permits for new dwellings will involve the removal or demolition of an existing dwelling. Comparing the increase in the number of dwellings (occupied and unoccupied) counted by the census with the number of permits issued in the preceding period (but lagged by you years to account for time to complete a building) gives me a guide as to how many permits are for replacement dwellings. My estimate is that in Auckland, 94% of permits are for new dwellings (not replacements). 

Table 3: New dwellings versus building consents for dwelling units


New Dwellings


















New dwellings as %

of permits issued





The data suggests some unders and overs. For example between 2006 and 2013, 33,465 dwellings were added to the housing stock in Auckland, yet in the period 2004 to 2011, over 50,000 building consents were issued. A large number of consents must not have been actioned, or they took much longer to complete than 2 years. 

In my analysis I have used the average for the period 2001 to 2018. 

In Queenstown, the net addition rate is closer to 97%.

Table 4: New dwellings and building consents for dwelling units - QLDC

New Dwellings














Permits as % of new dwellings




Next allowance needs to be made for ‘unoccupied’ dwellings - second houses, holiday homes and the like. In Auckland, census data suggests that of the total housing stock added between 2006 and 2018, 10% were unoccupied dwellings. In QLD it is more like 25%. Unoccupied means clearly not occupied on the night of the census. 

Table 5: Occpied dwellings - Auckland and Queenstown 





Total new dwellings 








% Occupied





Total new dwellings








% Occupied




To put it together, for example in 2016 to 2017, the population of Auckland City grew by 35,300. At a ratio of 3.29 people per household, this equals a demand for 10,740 new dwellings. In 2019, building permits for  13,754 dwelling units were issued. Assuming that 6% involved replacement of an existing dwelling, then 12,929 new dwellings were added to the stock (net growth). Of these 91% were for occupied dwellings, or 11,745 dwellings. This then equals a surplus of 1,005 dwellings. .

Below is the graph of the annual demand and supply from 1997 to 2018, for the Auckland Region. There is a ‘deficit’ over this period of 14,000 dwellings.  The last 10 years (1999 to 2018) has seen a deficit of 3000.This is a lot less than the 30, 000 to 40,000 sometimes bandied about. 

Figure 2: Auckland region: estimated housing demand and supply

The recent upward swing in building permits for new occupied dwellings suggests that the housing market is responding (as has the RMA). The deficit really built up in the post GFC period, and then it did take a while for the new build market to respond to the fast population growth between 2013 and 2018. 

The accumulated surplus/deficit looks quite large, when graphed. See below. Since 2017, substantial inroads have been made.  

Figure 3: Auckland Region housing surplus/deficit

The same exercise can be  undertaken for Queenstown Lakes District. Here there is a closer match between demand and supply, with some divergence over the period. Is this divergence enough to explain the large increases in house prices seen over this period?  

Figure 4: Queenstown Lakes District demand and supply estimate

There is an accumulated surplus of 480 dwellings, using the methodology set out above. 


Figure 5: QLDC housing surplus

Is there a supply myth? And if it is all a bit of a myth, what is the implications for the reform of the RMA and implementation of instruments like the National Policy Statement on Urban Development? 

Planning needs to keep adding new capacity for more dwellings and businesses, but adding more capacity will not necessarily bring forward more affordable houses for households on low to moderate incomes.  Urban quality is also important, and shouldn't be left behind in the dash for more capacity.