Since then a couple of sets of numbers have come through that shed some light on the points I was trying to make.
First up is the lag issue - the length of time taken between population growth occurring and houses being built. It is often noted that housing markets cannot react quickly to population shocks. It takes time to find sites, draw up plans and get a builder on board; more so as cities intensify and there is a shift from stand alone houses to terraces and apartments.
The Interest.co.nz website (see note 1) ran an interesting set of numbers recently, comparing building permits for new dwellings and code of compliance certificates (CCCs) issued by the Auckland Council.
CCCs are issued when a building is completed. Comparing the numbers of building permits issued with the number of code of compliance issued suggests that there is a significant time delay between a new building consent being issued and the building receiving a CCC when it is completed.
The data suggests that the lag is about 2 years, based on the numbers in the table below, which are for the Auckland Region.
The data suggests that in 2021, up to 14,000 dwellings may be built in the region.
Note, the 2 year lag is between the building permit being issued and the building being finished. For the building permit to be issued, a site needs to have been found, if needed resource consents obtained and building plans prepared, checked and consented. This adds further time. So the number of building consents issued in any year is likely to reflect the planning and design work undertaken during the previous year or two. But by the time the building consent is issued, these planning related issues have been addressed. The two year lag between building consent being issued and completion must be because of factors like labour shortages and possibly the growing number of multi-unit developments which take longer to build (where all units are completed at the same time).
During the period between population growth occurring and a new dwelling being finished - perhaps 3 years at least - then there can be extra demands on the housing stock. These are short term pressures. What is important is the housing market response over the medium term. Many commentators seem to latch onto the short term pressure and use this as support for the view that planning restrictions are creating a structural housing shortfall. But what the data, and most texts on housing markets note, is that there is always a delay between demand and supply. That delay is not necessarily the result of planning.
Next bit of information relates to whether the (delayed) supply response is sufficient to meet demands. The question has always been, what is a reasonable estimate of demand?
Here the other interesting bit of information is the revision to the net migration numbers by Stats NZ. Stats NZ have been busy revising how they count inwards and outwards migration. As a result of these revisions the number of long term migrants (people staying for more than 12 months over a 16 month period) is lower than previously estimated. Inward migration has been the biggest source of population growth over the past 5 years or so, so a revision is important.
The table below lists the old and new net international migration estimates, for NZ as a whole.
The difference is large in the 2017 and 2018 periods.
Apart from the obvious point that revisions of this scale drive home the point that it is very hard for the building sector to estimate demand if the numbers keep on moving; the other question is over the make up of these numbers. The numbers are for all types of migrants. A migrant is an overseas resident who arrives in New Zealand and cumulatively spends 12 out of the next 16 months in New Zealand. Migrants may be NZ'ers returning, people who come to work, visit for a long period of time or for study, or to shift permanently. For example, an international student who spends the term-time in New Zealand and holiday time overseas, over successive years, will be counted as a migrant arrival at the time of their initial border crossing if they satisfy the 12/16 month criterion.
Net international migration took off in 2014 and peaked last year. Below is the ‘old’ measure of net migration, and while the numbers have been revised downward, the pattern remains.
Migration was slowing down or negative in the period 2009 to 2012. Based on the above data on building consents and completions, the number of house completions in 2015 most likely reflected the conditions prevailing in 2012 (ie a much more bleak picture).
As for the causes of the lift after 2014, in June 2016 the NZ Herald ran a story that international student numbers from new migrant source countries, like India and the Philippines, are contributing to net migration numbers hitting record highs, according to analysis into arrival and departure card data.
To further complicate things, the revised net migration numbers are national figures, not for the Auckland Region. To work out what they may mean for the Auckland Region, we need to make a few adjustments.
But before we do that, we need to understand that at the regional level migration is the difference between the number of people who have moved to, and departed from, a given area. Sub national net migration includes both international migration and internal migration gains and losses.
There is no reliable evidence as to the number of NZ residents leaving Auckland for other regions (ie internal migrants) for the period from 2014. Between 2001 and 2006, Auckland ‘lost’ about 18,000 people to other regions. For the period 2006 to 2013, the census records a net loss of 4,650 people. So this number bounces around a lot. From 2014 to 2019, internal migration out of Auckland could have been very high. Until the census data comes out, we dont really know.
Then there is the number of international migrants who stay in Auckland.
Stats NZ, for their yearly population estimates, make a stab at working out how many international migrants head to Auckland, versus how many head to other regions. Their estimate covers both internal and international migrants for the regions, but at the national level, the migration total must only be for international migrants (for internal migration, a loss from one region is made up by a gain in another region).
The following figures are Stats NZ estimates of migration gain for Auckland and NZ, and Auckland’s share.
The Auckland share of total migration gain for the country has dropped. This may be because of more local residents leaving Auckland, or fewer international migrants heading to Auckland. We don’t know.
If we take the above Auckland region shares and apply them to the revised national migration figures, then we get the following:
The revised migration gain can then be added to the estimate of natural increase to get total population growth for the Auckland region.
Does 6,000 fewer people make that much of a difference? At 3 people per house, this is 2,000 fewer houses. Whether the dwelling demand is 3 people per house is debatable. Many migrants are younger people - the median age of Auckland is dropping. At 2013, the average number of people per occupied dwelling in the Auckland Region was 3, but a younger age profile suggests a higher number of people per dwelling.
If we stick with 3 people per dwelling, then the revised population estimates result in the following dwelling demand.
We can then compare the dwelling demand with dwelling supply, but with a 2 year lag built in between building consents being issued and houses being built. The table below has the dwelling 'supply' lagged by 2 years. For example dwelling demand from population growth in 2014 is matched to building consents issued in 2012.
So there is a demand and supply imbalance, but that imbalance is because of the lag between demand becoming apparent and supply cranking up.
If we ask “did the rising demand result in a signal that building consents needed to be ramped up”, then we can say yes, but it is not a fast process. If we add a year into the supply chain between population growth and a house being completed, so the whole process takes three years, then we can see the housing market responding in the graph below.
Taking 2010 as a starting point, the figure below shows three things:
1. Demand as estimated by population growth (adjusted down based on the revised migration data)
2. Building consents for new dwellings, based on a years gap between demand and the consent being issued. So the 2011 consents are probably based on population demand in the previous year, being, 2010.
3. Completions are delayed by 2 years from the year the building consent was issued, so completions are based on estimated demand three years previous.
What is apparent is the large step up in population growth between 2013 and 2014. Consents started to rise in response, but completions were low because they were based on what was consented two years previous. By 2018, with population growth dropping back a bit, completions have almost caught up.
There will still be a shortage of houses built up over the period 2010 to 2018, and the above suggests that the shortfall will take a while to work its way out of the system, unless population growth takes off again, or falls back quickly.
This analysis reinforces the need for there to be sufficient zone capacity to meet housing demand, but provides some caution as to the presumed role of a lack of capacity due to zoning, and the slow pace of adding more capacity, in fueling land and house price rises. In particular, the AUP (OP) was made operative in late 2016. By 2016, the housing market was responding to the increase in population growth from a few years back, based on the capacity available in 2013. The AUP (OP) has helpfully added capacity (as any plan review should and would have), but this is capacity for future growth. If a lack of zoned capacity was not the cause of house price growth (or only a small cause), then what is the main driver?
Notes: https://www.interest.co.nz/property/100382/number-new-homes-being-completed-auckland-could-increase-about-third-over-next-two
