This morning, we have seen the National Party launch their new housing and urban development plan titled ‘Going for Housing Growth’, in part a response to their previously signalled shift away from the Medium Density Residential Standards (MDRS).
Good write up brother, think the policy is a bit of a utopian view that we can density well. On the theory side it is great to really encourage higher density in key spots, but it doesn’t take into what market can/wants to deliver
That’s right, constraining medium density in suburbs won’t necessarily translate into greater demand for higher density in centres. We need both to meet demand and variety of choices.
Of course, this all depends on the outcome of the election which is far from certain, and the mechanics of MMP may well have an impact. But as you so eloquently express, a mixture of good, bad and plain odd stuff in National’s Housing Policy.
Also missing is effective design standards, recognising the psychological need for human scale and developing/reinforcing communities, really serious central government support for physical, social and environmental infrastructure. Time for much more form-based zoning? Time to re-visit 'A Pattern Language' by Alexander and others?
Isn't National's proposal BETTER for densification than the current system?
You say: "[a 30 year capacity rule] does not mean we will actually get the capacity where we need it, which is in the most central locations close to jobs, education and services."
Where do you think "we need it"? Do we need more housing concentrated in central locations? Or do we need more housing dispersed around the city?
The NPS-UD drafters thought we need it concentrated near metropolitan centres and high-quality public transport. That's why they imposed upzoning that improves the feasibility of development in these locations.
The MDRS drafters thought we needed development to be dispersed everywhere throughout the city, so that urban form would be less dense and infrastructure per new dwelling would cost more. That is why they reduced the incentive to build near centres and public transport routes by upzoning to improve the feasibility of developing land a long way from these places. (Perhaps more likely is that they didn't think about it at all).
The MDRS cancelled out the effect of the NPS-UD, in other words. The MDRS is not good for density or emissions. If you like the NPS-UD on density and emissions grounds you should not like the MDRS.
Since the MDRS makes it profitable to pepper-pot 2-3 dwelling and 2-3 story developments miles away from public transport, it will encourage development to occur in those places, which will preclude development that would otherwise have occurred closer to metropolitan centres and PT. That is not good for urban density. It will lock in a car dependent urban form.
Some people don't understand that allowing medium density on sites a long way from where the population is most dense (MDRS) does not help promote density over the city. If you want a higher-density city, you need to put more people where the people already are (NPS-UD).
I have started wondering whether people are confusing site density with urban density. These are completely different concepts. Higher site density in the wrong places means lower urban density, a la the MDRS.
National's infrastructure contributions proposal will also significantly reduce the feasibility of greenfields development vis-a-vis infill development. That will promote a more dense urban form.
So in conclusion, I really don't understand why density advocates are treating the (possible) death of the MDRS as anything but a good thing. Apart from tribal/political affiliations being triggered, what is the actual problem? It's clearly an improvement on the status quo in terms of density and emissions.
(To state the obvious: the MDRS on top of the NPS-UD will not lead to MORE housing being built overall or any improvement in housing affordability. It will just move development around and change how it looks. Developers don't build out a certain percentage of zoned capacity each year. If that was the case, increasing housing supply would be easy: Councils could just increase zoned capacity and watch development grow proportionally. A cursory glance at the numbers shows that this is simply not what happens. A deeper look at the economic theory relating to new housing supply confirms that)
Hi Tim, apologies but somehow I missed the notification for this comment at the time.
I support the MDRS as it enables 3 storeys across the entire urban area. Not just the 3 house x 3 storey rules, but enabling terraced housing and walk-up apartments of three storeys in all neighbourhoods. This is critical for enabling housing choice, as I have wrote about previously. https://betterthingsarepossible.substack.com/p/we-should-allow-three-storeys-everywhere
I don’t consider this to be completing competing with densification around the rapid transit stations or major centres. In those areas, we should be seeking real density of 6+ storeys. Probably at least 10 storeys in central areas of high access and high amenity. People seeking a terraced house or walk-up apartment in a suburb, will not necessarily choose an apartment in higher density developments if we constrain supply in suburban areas. I often hear people talking about how they wouldn’t mind living in apartment but don't want to be in high density developments surrounded by busy and heavily trafficked streets. They would however, be happy to live in a low to medium density suburban apartment. If that is not available then they will look towards a stand-alone house instead, including in greenfield areas. I think Ockham's development in Auckland showcase the demand for suburban apartment living. Close to good public transport but not always on busy main roads or in town centres. This is still a better outcome than greenfield growth.
I short, I think its overly simple to think demand for one typology or area will easily switch in either direction. We should enable housing choice and then use price levers where appropriate to encourage and discourage growth. E.g. the infrastructure contributions proposal for greenfield areas.
Do you not think the outcomes from NPS-UD+MDRS vis-a-vis NPS-UD alone will involve substitution effects?
In my understanding, the difference in outcomes between those two policy packages is pure substitution.
NPS-UD alone will lead to a higher share of development around centres and transit than NPS-UD+MDRS. The latter will have a higher share of dispersed development than the former.
Am I to understand you don't see that happening? I don't understand why.
As for the actual typologies, I agree, anecdotally many people are comfortable with medium-density (3 storey) stuff and would not contemplate living in the high-density (6 storey) stuff.
Offering them low or high density without the medium option inhibits choice, I agree.
The MDRS achieves that but at the cost of inefficient land use patterns, i.e. uncoordinated demands on infrastructure. MDRS means we'll be scrambling to cope with growth patterns, and will need to pay for (lumpy) infrastructure upgrades that provide lower value-for-money than more co-ordinated growth.
Ironically that problem of missing middle supply is worsened by the "all or nothing" nature of the NPS-UD. If you have land zoned for 6 stories, and the market doesn't want much of that, you're more likely to landbank until demand finally arrives, than to develop to less than the planning envelope (e.g. 3 stories). If you have land zoned for only 3 stories, in line with where the market's at, you have less reason to delay.
(Sheridan Titman's 1985 American Economic Review paper first argued this point... there's a brief explanation here:
Had the NPS-UD standards been a bit more graduated - e.g. 6 stories in some areas stepping down to 3 nearby, rather than the binary in/out which creates a lot of heat over walking catchments, centres boundaries, etc - we would get just as much housing but in typologies people actually want. That would reduce greenfields demand.
One more comment: a great example of zoning inhibiting choice is here in Wellington.
Two developers developed prime Te Aro land to 2/3 stories, because that's where the market's at. Around them are many high rises.
The developers were also concerned that Council would impose MINIMUM density requirements. So they wanted to corner the medium-density Te Aro market before that choice was outlawed.
Council then went ahead and, seemingly in vengeance, prohibited further medium-density development. Now you can't build to less than 6 stories in Te Aro.
This does two things:
1. Gives these developers a duopoly on medium-density housing in that area. The typology the market wanted is now more expensive in price because of lack of competition (a decision directly in opposition to the competitive markets objective of the NPS-UD, I should add).
2. Makes other Te Aro landowners less likely to develop their land. If the highest-and-best-use (most profitable development) is 3 stories, as the Paddington and Aro suggests, then prohibiting this will lead to more land held in low-density use until such time in the (possibly far-distant) future when 6 stories is possible.
It's one of the clearest examples I've seen of planners restricting choice in ways that directly work against their stated goals of seeing the area developed.
Here's a write-up (the proposed district plan rule of minimum 6 stories hasn't been much reported but it's in the plan):
I do think there will be some substitution effects, but equally this could go in the opposite direction towards lower density greenfield growth.
Regarding your Wellington example, I understand the reason two and three storeys was developed on that site was due to infrastructure limitations, not the market. Happy to chat about that more in person sometime.
Thanks. Great overview. Unfortunately, I do not share your enthusiasm for National's proposed policy to mandate mixed use along the PT corridors and residential areas around commercial centres. Whilst I am a big fan of mixed use in the right places, carte blanche enablement as suggested is very likely to undermine the network of commercial centres that the NPSUD itself is centred on, by dispersing the provision of commercial floorspace over very large areas/corridors. There is only so much commercial floorspace that can be sustained by demand and many commercial centres are already finding things challenging.
Kia ora Adele, thank you and thanks for engaging on this important topic.
I'm not sure I see the same risk as you do. Where we have zoned for Mixed Use development under the Unitary Plan we have seen a good deal of medium to high density apartment buildings but most have added fairly little additional commercial space so not undermined existing centres. Enabling Mixed Use should not include any requirement for additional commercial space, i.e. I don't think it needs to include any frontage controls. It is however, really important that we do allow for commercial spaces where the market sees demand, as increased density over time with developments up to 6+ storeys will allow for growth of commercial centres.
I guess we'll have to agree to disagree. Experience down here (Christchurch) would suggest it more likely that the corridors would be taken up for single storey commercial development or 2/3 storey townhouses rather than the mixed use 6 storey utopia sought by this policy. There is no precedence for that here (even in large commercial centres themselves) and unlikely to be commercial feasible in the foreseeable future.
Different set of circumstances down here. Even the mixed-use zone in the central city has controls on the type and scale of commercial activity that can establish in order to support central city (core) recovery after the earthquakes. The basis for that has been thoroughly tested. Another one size fits all approach might not be desirable in this case too. I do think a more nuanced approach might work though, focused on enabling just smaller scale convenience type of activities along the corridors, or at nodes. I believe that is the intent anyway. The devil will be in the (lack of) detail...
Really useful deep-dive into the details and a well-rounded assessment. Thanks Malcolm.
Good write up brother, think the policy is a bit of a utopian view that we can density well. On the theory side it is great to really encourage higher density in key spots, but it doesn’t take into what market can/wants to deliver
That’s right, constraining medium density in suburbs won’t necessarily translate into greater demand for higher density in centres. We need both to meet demand and variety of choices.
Of course, this all depends on the outcome of the election which is far from certain, and the mechanics of MMP may well have an impact. But as you so eloquently express, a mixture of good, bad and plain odd stuff in National’s Housing Policy.
Also missing is effective design standards, recognising the psychological need for human scale and developing/reinforcing communities, really serious central government support for physical, social and environmental infrastructure. Time for much more form-based zoning? Time to re-visit 'A Pattern Language' by Alexander and others?
Isn't National's proposal BETTER for densification than the current system?
You say: "[a 30 year capacity rule] does not mean we will actually get the capacity where we need it, which is in the most central locations close to jobs, education and services."
Where do you think "we need it"? Do we need more housing concentrated in central locations? Or do we need more housing dispersed around the city?
The NPS-UD drafters thought we need it concentrated near metropolitan centres and high-quality public transport. That's why they imposed upzoning that improves the feasibility of development in these locations.
The MDRS drafters thought we needed development to be dispersed everywhere throughout the city, so that urban form would be less dense and infrastructure per new dwelling would cost more. That is why they reduced the incentive to build near centres and public transport routes by upzoning to improve the feasibility of developing land a long way from these places. (Perhaps more likely is that they didn't think about it at all).
The MDRS cancelled out the effect of the NPS-UD, in other words. The MDRS is not good for density or emissions. If you like the NPS-UD on density and emissions grounds you should not like the MDRS.
Since the MDRS makes it profitable to pepper-pot 2-3 dwelling and 2-3 story developments miles away from public transport, it will encourage development to occur in those places, which will preclude development that would otherwise have occurred closer to metropolitan centres and PT. That is not good for urban density. It will lock in a car dependent urban form.
Some people don't understand that allowing medium density on sites a long way from where the population is most dense (MDRS) does not help promote density over the city. If you want a higher-density city, you need to put more people where the people already are (NPS-UD).
I have started wondering whether people are confusing site density with urban density. These are completely different concepts. Higher site density in the wrong places means lower urban density, a la the MDRS.
National's infrastructure contributions proposal will also significantly reduce the feasibility of greenfields development vis-a-vis infill development. That will promote a more dense urban form.
So in conclusion, I really don't understand why density advocates are treating the (possible) death of the MDRS as anything but a good thing. Apart from tribal/political affiliations being triggered, what is the actual problem? It's clearly an improvement on the status quo in terms of density and emissions.
(To state the obvious: the MDRS on top of the NPS-UD will not lead to MORE housing being built overall or any improvement in housing affordability. It will just move development around and change how it looks. Developers don't build out a certain percentage of zoned capacity each year. If that was the case, increasing housing supply would be easy: Councils could just increase zoned capacity and watch development grow proportionally. A cursory glance at the numbers shows that this is simply not what happens. A deeper look at the economic theory relating to new housing supply confirms that)
Hi Tim, apologies but somehow I missed the notification for this comment at the time.
I support the MDRS as it enables 3 storeys across the entire urban area. Not just the 3 house x 3 storey rules, but enabling terraced housing and walk-up apartments of three storeys in all neighbourhoods. This is critical for enabling housing choice, as I have wrote about previously. https://betterthingsarepossible.substack.com/p/we-should-allow-three-storeys-everywhere
I don’t consider this to be completing competing with densification around the rapid transit stations or major centres. In those areas, we should be seeking real density of 6+ storeys. Probably at least 10 storeys in central areas of high access and high amenity. People seeking a terraced house or walk-up apartment in a suburb, will not necessarily choose an apartment in higher density developments if we constrain supply in suburban areas. I often hear people talking about how they wouldn’t mind living in apartment but don't want to be in high density developments surrounded by busy and heavily trafficked streets. They would however, be happy to live in a low to medium density suburban apartment. If that is not available then they will look towards a stand-alone house instead, including in greenfield areas. I think Ockham's development in Auckland showcase the demand for suburban apartment living. Close to good public transport but not always on busy main roads or in town centres. This is still a better outcome than greenfield growth.
I short, I think its overly simple to think demand for one typology or area will easily switch in either direction. We should enable housing choice and then use price levers where appropriate to encourage and discourage growth. E.g. the infrastructure contributions proposal for greenfield areas.
No worries Malcolm. Thanks for your reply.
Do you not think the outcomes from NPS-UD+MDRS vis-a-vis NPS-UD alone will involve substitution effects?
In my understanding, the difference in outcomes between those two policy packages is pure substitution.
NPS-UD alone will lead to a higher share of development around centres and transit than NPS-UD+MDRS. The latter will have a higher share of dispersed development than the former.
Am I to understand you don't see that happening? I don't understand why.
As for the actual typologies, I agree, anecdotally many people are comfortable with medium-density (3 storey) stuff and would not contemplate living in the high-density (6 storey) stuff.
Offering them low or high density without the medium option inhibits choice, I agree.
The MDRS achieves that but at the cost of inefficient land use patterns, i.e. uncoordinated demands on infrastructure. MDRS means we'll be scrambling to cope with growth patterns, and will need to pay for (lumpy) infrastructure upgrades that provide lower value-for-money than more co-ordinated growth.
Ironically that problem of missing middle supply is worsened by the "all or nothing" nature of the NPS-UD. If you have land zoned for 6 stories, and the market doesn't want much of that, you're more likely to landbank until demand finally arrives, than to develop to less than the planning envelope (e.g. 3 stories). If you have land zoned for only 3 stories, in line with where the market's at, you have less reason to delay.
(Sheridan Titman's 1985 American Economic Review paper first argued this point... there's a brief explanation here:
https://www.fresheconomicthinking.com/p/why-i-am-anti-anti-zoning)
Had the NPS-UD standards been a bit more graduated - e.g. 6 stories in some areas stepping down to 3 nearby, rather than the binary in/out which creates a lot of heat over walking catchments, centres boundaries, etc - we would get just as much housing but in typologies people actually want. That would reduce greenfields demand.
One more comment: a great example of zoning inhibiting choice is here in Wellington.
Two developers developed prime Te Aro land to 2/3 stories, because that's where the market's at. Around them are many high rises.
The developers were also concerned that Council would impose MINIMUM density requirements. So they wanted to corner the medium-density Te Aro market before that choice was outlawed.
Council then went ahead and, seemingly in vengeance, prohibited further medium-density development. Now you can't build to less than 6 stories in Te Aro.
This does two things:
1. Gives these developers a duopoly on medium-density housing in that area. The typology the market wanted is now more expensive in price because of lack of competition (a decision directly in opposition to the competitive markets objective of the NPS-UD, I should add).
2. Makes other Te Aro landowners less likely to develop their land. If the highest-and-best-use (most profitable development) is 3 stories, as the Paddington and Aro suggests, then prohibiting this will lead to more land held in low-density use until such time in the (possibly far-distant) future when 6 stories is possible.
It's one of the clearest examples I've seen of planners restricting choice in ways that directly work against their stated goals of seeing the area developed.
Here's a write-up (the proposed district plan rule of minimum 6 stories hasn't been much reported but it's in the plan):
https://www.stuff.co.nz/life-style/homed/housing-affordability/300267853/wellingtons-developing-drama-city-taking-the-low-road-when-high-rise-is-answer-to-housing-crisis
Hi Tim
I do think there will be some substitution effects, but equally this could go in the opposite direction towards lower density greenfield growth.
Regarding your Wellington example, I understand the reason two and three storeys was developed on that site was due to infrastructure limitations, not the market. Happy to chat about that more in person sometime.
Thanks. Great overview. Unfortunately, I do not share your enthusiasm for National's proposed policy to mandate mixed use along the PT corridors and residential areas around commercial centres. Whilst I am a big fan of mixed use in the right places, carte blanche enablement as suggested is very likely to undermine the network of commercial centres that the NPSUD itself is centred on, by dispersing the provision of commercial floorspace over very large areas/corridors. There is only so much commercial floorspace that can be sustained by demand and many commercial centres are already finding things challenging.
Kia ora Adele, thank you and thanks for engaging on this important topic.
I'm not sure I see the same risk as you do. Where we have zoned for Mixed Use development under the Unitary Plan we have seen a good deal of medium to high density apartment buildings but most have added fairly little additional commercial space so not undermined existing centres. Enabling Mixed Use should not include any requirement for additional commercial space, i.e. I don't think it needs to include any frontage controls. It is however, really important that we do allow for commercial spaces where the market sees demand, as increased density over time with developments up to 6+ storeys will allow for growth of commercial centres.
I guess we'll have to agree to disagree. Experience down here (Christchurch) would suggest it more likely that the corridors would be taken up for single storey commercial development or 2/3 storey townhouses rather than the mixed use 6 storey utopia sought by this policy. There is no precedence for that here (even in large commercial centres themselves) and unlikely to be commercial feasible in the foreseeable future.
Different set of circumstances down here. Even the mixed-use zone in the central city has controls on the type and scale of commercial activity that can establish in order to support central city (core) recovery after the earthquakes. The basis for that has been thoroughly tested. Another one size fits all approach might not be desirable in this case too. I do think a more nuanced approach might work though, focused on enabling just smaller scale convenience type of activities along the corridors, or at nodes. I believe that is the intent anyway. The devil will be in the (lack of) detail...
Totally agree with the latter points, really needs to focus on smaller scale.