Last week, the New Zealand Government announced a major building densification law change that will see up to 105,500 extra homes built over the next 5-8 years across New Zealand:
The Government, with the support of National, is proposing to urgently change the Resource Management Act (RMA) to enable more houses to be built in cities.
It plans to pass a bill to make councils in Auckland, Hamilton, Tauranga, Wellington and Christchurch implement intensification requirements under the National Policy Statement – Urban Development at least a year earlier than the current requirement.
Councils will need to have their intensification policies and rules in place by August 2023….
These will enable landowners to build up to three homes of up to three storeys on most sites (up to 50% maximum coverage of the site) without the need for a resource consent. Previously, district plans would typically only allow for one home of up to two storeys…
The Government estimates the proposed changes will result in 48,200 to 105,500 additional dwellings over the next five to eight years, above what was expected from councils implementing the National Policy Statement’s intensification policies…
“Today National and Labour are coming together to to say an emphatic ‘yes’ to housing in our backyards,” [National’s Housing spokesperson Nicola Willis said].
The move has led to calls for Australian policymakers to emulate New Zealand.
However, recent research by Mark Limb and Cameron Murray, which assesses the outcomes of planning for density in established suburbs over a twenty-year period, suggests that zoning for density does not lower dwelling prices.
Below is the Abstract:
Does planning for higher density increase housing development and decrease housing prices? We study the outcomes of planning for density in established suburbs over a twenty-year period using a large site-level dataset on dwelling stock, planning regulations, and prices, in 19 planned densification areas (activity centres) comprising 25,775 sites in Brisbane, Australia. Planning rules in these areas were repeatedly relaxed to allow for higher density; a policy change that should have observable price effects. To study the effect of zoning, we create a variable for each site called zoned capacity, which is the estimated number of additional dwellings able to be built under the planning code. Only 2% of the zoned capacity was taken up in any five-year study interval. Zoned capacity doubled over the whole twenty-year study period (going from 0.9x total dwellings to 1.4x), however despite these changes, 78% of sites with zoned capacity in the first period remained undeveloped. Higher rates of new housing supply are robustly related to higher prices despite demand arguably seeing a similar increase across locations. Our zoned capacity variable has no relationship to price across numerous regression models and is robust to various data selection choices. It could be that planning is not a binding constraint on new housing in Brisbane—yet price growth over our study period is comparable to other Australian cities. This evidence suggests that private housing markets will not rapidly supply new housing and cause significant price reductions, even if the planning system allows it.
And below is the Discussion and Conclusion:
Using a large site-level database across twenty years and 19 activity centres that were targeted in the planning system for residential densification, we have tested which of two causal stories about housing supply is consistent with the empirical record. The expected and observed patterns are summarised in Table 7. Overall, rising prices appeared related to rising new housing construction, even after controlling for zoned capacity, and these results are robust to various data selection choices. This contradicts the predictions of the planning story, which argues that areas with more zoned capacity will see more dwellings construction and lower prices after a demand shock.
Instead, the data shows patterns consistent with the dynamic economic story whereby viable sites are developed at a rate that depends on price growth.
Like (Monkkonen et al. , 2020) we find that the interaction of our regulatory metric, lagged zoned capacity, to be positively related to new dwellings (though our result is not robust to all data selection choices). This is, of course, by design. Planning regulations are intended to locate new dwelling supply where it is planned rather than where it is not. Because of the lack of relationship between zoned capacity and price, even when interacted with a lag price, we also cannot conclude that regulation has most effect on the dwelling market in locations where prices are higher. Instead, we see this data as supporting the idea that higher prices are necessary for densification and for zoned capacity to be taken up at viable sites. This is consistent with the regression models of (Monkkonen et al. , 2020), whose negative relationship between existing density and multifamily permits also suggests that existing development reduces the economic viability of redevelopment, which is an alternative way of saying that higher prices are required to make redevelopment of existing uses viable.
Our approach is limited in the following ways. First, our planning story assumption that there is common demand shock across all locations in our data at each time period could be wrong. This would mean that the planning story predictions about the relationship between price and quantity would be the opposite (i.e. consistent with the dynamic economic story). Yet, there should still be a large observable negative relationship between zoned capacity and price, which does not appear in most model variations. In the one instance it does, the relationship is extremely small, implying a price elasticity with respect to zoned capacity of 0.003. Put another way, a doubling of zoned capacity (a 100% increase) from this already high average amount might reduce prices five years later by 0.3%. Even if this occurred in every five-year period, that is only a 3% price effect after compounding over twenty years. Second, our price information may be too limited, as we do not have unique prices at each site, nor hedonic price indexes. But if supply variation does not affect these local averages over five year windows, then it is hard to argue that is has a meaningful price effect. While this analysis cannot exclude the possibility that there are small undetectable negative price effects from policy changes that increase zoned capacity, it can exclude the idea that if these price effects exist that they are large or important for explaining the overall pattern of housing prices.
The regression results coupled with the overall descriptive data supports the idea that zoned capacity alone is not a major determinant of the rate of new dwelling supply, nor dwelling price. Only 22% of the sites with zoned capacity in 1996 had any redevelopment in the subsequent twenty year period. Despite all the development that did occur, zoned capacity in every activity centre continued to grow over the 20-year study period, and in total zoned capacity was twice as high at the end of the study period than the start. Meanwhile, the increase in dwellings in these activity centres was in line with the city average, with many below it.
It may be the case that the rate of new housing supply in Brisbane is unconstrained by planning, either in these activity centre location or overall. If this is true, however, then any claim of large price effects from supply constraints runs into difficulty. Over the 2001-2016 period, for example, Brisbane dwelling prices grew more than those in Sydney and Melbourne ABS (2018). Yet when we study those periods in our data it seems that zoned capacity had no relationship to price. It is more likely that the planning story of housing supply assumed by the various policy interest groups in the YIMBY movement is not an accurate description of housing supply.