Behind the contentious KiwiBuild numbers
It’s one of the most important numbers in politics, more contested than GDP: the number of KiwiBuild homes that will be built between now and 2028.
The promise is well-known: 100,000 over 10 years, but the reality could be quite different. So far, just 33 have been completed, with 77 more on the way and a pipeline for many thousands more slated to begin construction in the large urban development projects announced in Auckland and Wellington.
National MP Andrew Bayly told the House the Reserve Bank’s November Monetary Policy Statement (MPS) had assumed KiwiBuild would only deliver 7000 to 14,000 additional homes by July 2022. If KiwiBuild runs exactly to plan, it should deliver 28,000 homes over that period.
This was initially reported as the MPS claiming KiwiBuild would add 7000 to 14,000 houses.
But observers balked at this. The Reserve Bank is far too neutral to give something so candid as a number of KiwiBuild houses in its MPS. Even at the post-MPS press conference, this reporter was only able to get vague details on the Bank’s estimates from repeated questions.
But information released to Newsroom under the Official Information Act shows how the estimate was arrived at.
In September, staff from the Reserve Bank contacted Treasury for the latest estimates on the number of KiwiBuild homes it expects will be built. A Treasury staffer said it did not forecast the number of houses, but it did forecast the level of residential investment.
The figures date to the Budget Economic and Fiscal Update, or BEFU. This means they are in fact the same numbers that were reported on in May, just cut a different way.
In May, Treasury forecast KiwiBuild would add $2.5 billion “additional nominal residential investment” over the next five years - revised downwards from $5 billion at Christmas.
The number of houses, quoted in the House, relates to Reserve Bank estimates based on this Treasury forecast. This was not included in the MPS, as had been reported, but extracted from the Bank at Parliament’s Finance and Expenditure Committee.
After delivering the MPS, the Reserve Bank’s Governor, currently Adrian Orr, appears before Parliament’s committee to answer questions about the Bank’s forecast and decisions.
This is where the figure emerged.
A representative from the Reserve Bank told Newsroom FEC members asked the Bank about how many new houses would be built and it cautioned them that its forecasts were “based on the value of new builds, rather than the number of houses”.
This Thursday, the committee published its report, based on that hearing, which outlined how the 7000-14,000 number was arrived at.
The Bank said the number of new builds can be estimated based on different construction cost assumptions. The figure of 14,000 (which RBNZ says is actually 14,200 houses) is based on a KiwiBuild home being delivered at 50 percent of the average value of a building consent.
The 7100 figure estimates KiwiBuild houses will be built at the average consent value.
Using the average consent value raises some important questions for estimates of how many KiwiBuild homes will be delivered.
The average consent value jumps around a fair amount. Figures from Statistics NZ show that the average consent value in May and June 2018 was roughly $350,000, but the average value in January and February hovered around $400,000.
Using this estimate could bode well for the Government. KiwiBuild is meant to massively reduce construction costs by encouraging companies to explore prefabrication and mass development.
New Zealand’s residential construction industry is also geared unhealthily towards building expensive homes. Just 5 percent of new builds are priced at the lower quartile of property prices, meaning the value of new consents is skewed away from the likely cost of a new KiwiBuild home.
That means you can fairly discount the 7100 figure as an extreme lowball — for all the accusations levelled against KiwiBuild, the homes are unlikely to conform to the luxury standards of most new builds.
But there’s also some good news for the opposition.
New consents are also spread relatively evenly between Auckland and the rest of the country. In the year to June, 38 percent of consents were issued in Auckland.
This could skew the average consent price downwards, as the cost of building in Auckland is far higher than the rest of the country.
But KiwiBuild is aimed at addressing the housing crisis where it is most acute — and that means its programme is heavily skewed in favour of Auckland, where 50 percent of the planned 100,000 homes will be built.
This means the average cost of a KiwiBuild could edge closer to the average value of a consent. Even though most consented homes are relatively luxurious, a luxury home in Southland can cost less than a a rotting shack in Grey Lynn.
Housing Minister Phil Twyford has disputed the estimate since it was first after the Budget.
Both the optimistic and pessimistic forecasts have their risks.
De-risking residential construction could come with massive gains in productivity, increasing existing capacity, or there could be a financial shock in the next four years. A shock would cause overstretched construction firms to collapse, leading to a huge slowdown in construction; with the forecast being slashed by 50 percent between the election and Budget, it’s difficult to tell.
Twyford’s next moments of truth will come at the Half-Year Economic and Fiscal Update next week, and then the next Budget, when Treasury releases its latest forecasts.