Kāinga Ora is a landlord whose business is managing the tenancies of some 72,000 public homes. Photo / Michael Craig
On the spectrum of business models, selling something expensive to people who can’t afford it lands somewhere between foolish and impossible (unless you’re a bank in which case such it can be rather lucrative).
Kāinga Ora is a landlord whose business is managing the tenancies of some 72,000 public homes, which is the vast majority of the more than 80,000 public housing places offered by the government.
Public housing is the modern iteration of what most people still tend to call “state housing”. Kāinga Ora provides state owned housing, and Community Housing Providers (CHPs) receive government funding to provide a similar service in a privately owned home.
Beginning in 2022, allegations have swirled that Kāinga Ora had become financially unviable.
Shortly after coming into office, the new Government commissioned a report by former Prime Minister Bill English to look into those allegations. That report was published in May and concluded the agency was “not financially viable under current settings”.
English had the receipts. He found operating deficits at Kāinga Ora were forecast to grow from $520million in 2022/23 to over $700m in 2026/27 and debt was forecast to blow out to $23 billion.
His conclusion did not go uncontested. In a response to English’s interim report, released under the Official Information Act, the old Kāinga Ora board said its most recent financials satisfied both the Treasury and the Ministry of Housing and Development. The board even had a crack at the English panel for having “relatively limited engagement” with the organisation.
The board was not the only organisation to take issue with the English report. Community groups, fearing market-led change in social housing, have also taken aim at English’s report, attacking both his diagnosis of the disease and the proscribed cure.
Kāinga Ora’s most direct ancestor is the Housing New Zealand Corporation, which was forged in its most recent form during Helen Clark’s fifth Labour Government. It largely represented a continuation of public housing provision, prior to the massive upheaval seen under the Fourth National Government in the 1990s.
HNZ and Kāinga Ora rent homes on a mostly commercial basis, but the income they receive from tenants is capped at 25 per cent of their income, with the Income-Related Rent Subsidy (IRRS) from the Government making up the rest.
In that sense, Kāinga Ora operates on a quasi-commercial model. It makes money by charging market rents to its tenants, but it is the Crown that picks up most of the tab (and has the greatest say in what defines “market rent”).
In the most recent annual report for the year ending June 2023, the agency received $469m in rent from tenants and $1.15b in IRRS. The more subsidy Kāinga Ora receives, the more properties it can afford to turn into public housing by building or buying housing and subsidising those tenancies with IRRS revenue.
Under the John Key-Bill English Government, HNZ was run on such a commercial basis that it even returned a dividend to the Crown.
Money still flowed predominantly from the Crown to HNZ in the form of rent subsidies, but considerable sums of money were also extracted in the form of a dividend each year - money that could have been reinvested in maintenance or growing the stock of public homes.
The Child Poverty Action Group (CPAG) commissioned its own “People’s Review of Kāinga Ora” in response to the English review. An analysis of the agency’s financials by co-convenor Alan Johnson for that report found that the government had withdrawn $576m in dividends while only contributing $131m in additional capital.
The English review found Kāinga Ora’s rental stock was “run-down and in need of significant investment”.
Johnson argued this was in no small part because previous governments had allowed the organisation to skimp on maintenance spending.
Johnson calculated that for the nine years between July 2009 and June 2018 maintenance spending as a proportion of building value averaged 3.1 per cent. The most recent five years under Labour saw maintenance spending average 3.7 per cent, although this was skewed by a binge on maintenance spending in 2022/23. Pressure to return a dividend may have been driving the low spend on maintenance.
The former Kāinga Ora board agreed, arguing that the poor state of its housing stock was partly a result of decisions taken by the last National Government, saying that “between 2007 and 2017 there was almost no renewal of homes. This led to a backlog of renewal needs of around 10,000 homes.”
In 2019, Kāinga Ora was formed through the merger of Housing New Zealand, Homes Land Community (formally Hobsonville Land Company) and KiwiBuild. The new agency was primarily a landlord, but it had bigger, broader ambitions too. It would be a developer, leading urban renewal projects imagined by the then Labour Government.
The organisation was briefly allowed to borrow on its own balance sheet rather than going through Treasury like other Government departments.
The official reason for this was that it gave Kāinga Ora the ability to manage its debt independently of the government, allowing it to grow naturally as its needs required, rather than at the whim of Treasury. The unofficial reason was that it flattered a core Crown debt figure that stalked the Labour Government’s nightmares by taking one of the most debt-laden parts of the government off the core Crown balance sheet.
This largely cosmetic debt-management exercise proved costly. In 2021, the average effective interest rate on Kāinga Ora’s Crown debt was 0.44 per cent, according to Johnson’s report. The same figure for its $5.5b of market debt was 1.95 per cent. It meant that that year alone Kāinga Ora was paying about $80m in interest costs it could have avoided if the Government had continued to force it to borrow through Treasury.
More alarmingly, the English review makes the case that Kāinga Ora used its independence irresponsibly. Backed by an asset base of tens of thousands of homes, steady income from the Crown, and an implicit guarantee that if it ever got into trouble, the Crown would bail it out, Kāinga Ora borrowed with abandon.
English found governance issues, and a costly passion for being an industry leader in things like green building and modular construction.
The organisation argued that a state-backed construction leader might have positive flow-on effects for the wider industry, but in at least one development, these add-ons came at an additional cost of $55,000 per unit. That’s something which arguably should have been trimmed in the interest of directing that money building more houses.
A document from Kāinga Ora cited by English said the positives of sustainable building were “considered to outweigh the financial performance metrics”.
English, horrified by the agency’s forecast deficits, begged to differ.
Kāinga Ora’s operating financial information for the 166-unit Elm Street development “did not include interest servicing costs which, if included, change the operating cashflow for the development to negative”.
Kāinga Ora, a homebuilder, was not thinking enough about one of the main costs to building houses: interest.
Johnson agreed the organisation’s governance was a problem
“I think governance isn’t necessarily as strong as it could have been,” he told the Herald.
“I think the model isn’t really one of independent governance. The minister has a hand in things, whether it’s a direct hand or simply one by virtue of persuasion and influence, but I don’t think the organisation is completely independent and I don’t think it ever has been to be honest,” he said.
The English review recommended paring back some of Kāinga Ora’s many new functions and reverting to being the state landlord.
Under Labour, Kāinga Ora became an urban development agency, fulfilling its ambitions of leading large-scale urban renewal projects that mixed all kinds of housing, public and private.
The English report argued the agency has lost its focus amidst the grand ambitions set by the former Government.
Johnson’s analysis actually agreed that paring back is needed.
He was deeply critical of Labour’s decision to turn Kāinga Ora into a state urban developer without much thought into what this would mean for its core business of providing housing and without giving it adequate income to support that new role.
He argued that many of Kāinga Ora’s financial woes come down to issues with its urban development arm, rather than its building of public homes. Two-thirds ($643m) of the $1b in losses recorded by Kāinga Ora in its first five years were losses in value associated with the sell-off of Kāinga Ora property, partly in the service of its urban development mandate.
“It is difficult to know, from the available evidence, if these write downs/loss of value was on account of over-valuation of Kāinga Ora’s assets or poor deals in these asset sales which gave advantage to purchasers. The available evidence points to the former,” he said.
Kāinga Ora itself seemed to agree, or at least it reckoned that it could no longer do urban development as a bold-on business cross-subsidised by its work as a landlord.
In a 2022 paper warning that the organisation’s finances were looking unsustainable, Kāinga Ora asked the then-Labour Government to consider properly funding its urban development work, which at that point was largely cross-subsidised by its primary job as a landlord.
One fix, the organisation suggested, was the “[c]onsideration of funding for delivery of urban development outcomes”.
It’s been a bad time for anyone in the businesses that involve a lot of borrowing. Interest rates have risen, squeezing margins and making some businesses unviable.
The English review found Kāinga Ora’s unsustainable deficits were driven by three things: ”Interest on the debt-financed capital investment programme and a significant uplift in other expenses including staffing and maintenance”.
The Johnson review agreed that interest and staffing costs were also a problem. His review found that in 2020, with Labour’s immense build programme well under way, Kāinga Ora’s interest costs were under 10 per cent of its overall revenue. By 2023, when interest rates had spiked, that figure rose to over 15 per cent, squeezing its finances.
This was no secret. Kāinga Ora briefings going back to 2022 warned that “surges in construction and maintenance costs as well as interest rate increases are challenging [the] model, with expenses outpacing revenue streams at an accelerating speed”. The same briefing warned that construction costs for some developments had increased by over 50 per cent.
Attributing blame here is difficult.
Facing a severe housing shortage and with low interest rates prior to the pandemic, most governments would have looked to borrow money to build more houses, and most governments would have been caught out by the surge in high interest rates (although the fact that interest rates rose as high as they did is partly, though not wholly, the fault of the last Government).
Interest also weighed upon Kāinga Ora as it struggled to actually get developments done. Here governance appears to have been a problem, with the organisation building too much all at once and failing to bring houses to market in time.
As the organisation scaled up to deliver more and more houses, the number of “works in progress” exploded.
“Works in progress” are houses that are begun but not yet completed. They are a problem for Kāinga Ora because a work in progress means it is paying interest on the cost of building a home, but not receiving any income from it. The longer a home is “in progress” the more costly the problem.
The value of works in progress was $880m in 2018, costing $19m in interest. This grew to $3698b in 2023, costing $128m in interest. It was forecast to stay between $1.8b and $2.9b in the years to 2028 costing a cumulative $597m in interest over that time.
High interest rates and slow construction mean the mere act of construction is costing Kāinga Ora a fortune before it is able to gain any revenue from the houses that it builds.
English said trimming the cost of “works in progress”, presumably by better phasing the works in progress was “one specific area where costs could be reduced”.
There are ways to reduce this cost, but they come with trade-offs. The Government could invest more in innovative modular construction, which cuts down construction time and cost. However, as we’ve already seen, the English review was quite critical of one pilot scheme Kāinga Ora ran to trial offsite construction techniques to reduce costs.
The Kāinga Ora board actually shot back at English for criticising these pilots, saying they were the solution to cost-escalation problems he found.
The board wrote to English saying Kāinga Ora was “now achieving 40-60 per cent construction cost reductions in pilot programmes. This has been achieved through detailed task scheduling that has doubled labour productivity”.
You can see evidence of this in the vast amount of (not always popular) modular homes going up around the country.
The Government could also reduce the amount spent on works in progress by building less, but that comes with the obvious consequence of building homes at a time when the housing wait list continues to grow.
As of March, there are 25,527 applicants on the register - up 6 per cent from a year ago.
Staffing was another part of that cost-escalation story, rising to 20 per cent of revenue by 2023 from just over 10 per cent in 2018. Put together, staff and interest costs in 2023 amounted to more than a third of the organisation’s overall revenue, dragging it away from the delivery of new houses.
There is less dissent on the other leg of the cost increase problem: staff. Both reports note the massive growth in the organisation.
English cited the tenancy management, including the Customer Contact centre, which increased from 527 FTEs (full time equivalent staff) in 2018 to 1026 in 2023.
Overall, the size of the organisation has doubled in the four years since it was formed with the 19/20 annual report recording 1611.87 FTS and the 22/23 annual report recording 3305 FTEs.
The Johnson review found 20 per cent of the agency’s revenue was being spent on staff in 2023, nearly double level of spending in 2018.
If the main problem with Kāinga Ora is interest rates and staff, the fix is fairly simple: reduce Kāinga Ora’s headcount and help the Reserve Bank get inflation under control, and maybe even reduce the number of houses Kāinga Ora is building to lower its borrowing costs.
But these were not quite fixes proposed by the English report, which argued for a radical change, not just to Kāinga Ora, but to the whole system of social housing.
He argued Kāinga Ora should have an altered entity form, turning it into a Crown Company, which would likely reinforce a stronger separation between the executive, the board, and the minister.
English also wanted to see financial responsibility for housing “consolidated” under the Minister for Housing and his ministry which would begin to “actively purchase” housing services from other organisations like Community Housing Providers (CHPs), new Community Housing Associations (CHAs), and Kāinga Ora. Under this model, Kāinga Ora would become one social housing landlord among many.
The focus for providing new housing would shift from debt-laden Kāinga Ora, to CHPs and CHAs.
CHPs are organisations, mainly not-for-profits, that offer social housing, much like Kāinga Ora.
They also receive IRRS money from the Government to provide social housing just like Kāinga Ora. Rent charged to tenants is capped at 25 per cent of income, with IRRS making up the difference.
As of March, there were more than 13,000 community housing places in New Zealand that receive IRRS, making them community provided public housing places (CHPs also provide other accommodation that does not receive IRRS). The sector has roughly doubled since Kāinga Ora was formed, adding 6700 places since 2019 (Kāinga Ora added 11,900 in the same time).
The English report argues that instead of Kāinga Ora being the default provider of public housing, the Government could “actively purchase housing from a range of providers, including CHPs and CHOs would be better placed to deliver tailored housing services in their communities”. The Government, CHPs and CHOs would be empowered to do this if there were better data available on what people needed and where.
English calls this “social investment”. It is often means the state being agnostic about whether the public or private sector delivers social “outcomes”.
In the case of housing, social investment would mean the Social Investment Agency looks at what is needed where, and the Ministry of Housing and Urban development “buys” those housing services from a CHP, CHA, or indeed Kāinga Ora.
The English review found that CHPs have their own debt problem, which is quite the opposite of Kāinga Ora’s. While English reckoned Kāinga Ora borrowed too much, he’s worried CHPs can’t borrow enough, or at low enough cost.
Because CHPs own the buildings they let out to public tenants, they too need to borrow money to build or buy them. The problem for CHPs, English found, is that they were classed as commercial borrowers by banks, meaning their cost of borrowing was higher than if they had been residential borrowers taking out a mortgage against a house.
“These loans do not account for the security provided by the asset or the government-guaranteed, long-term nature of IRRS revenues”, the review argued. It said this led to “higher credit margins and mismatched repayment structures”, which made “financing difficult for most CHPs and alternative tenures without substantial equity”.
Community Housing Aotearoa deputy chief executive Chris Glaudel told the Herald that the problem for CHPs went all the way back to the classification of CHPs as “investors” by the Reserve Bank roughly eight years ago, forcing banks to lend to CHPs on more onerous and expensive terms.
“The criteria required anyone with four or more properties as an investor. We had submitted that because CHPs really aren’t the same as a typically investor or speculator in residential real estate there should be a carve-out.
“For us, with a 25-year government contract guaranteeing market rents it simply isn’t the same risk profile [as normal property investment], yet there isn’t a separate asset class established to recognise that risk profile appropriately.
“There hasn’t been other forms of support from the Government to lower that cost even though it makes up one of the biggest expenses of new social housing.”
Glaudel told the Herald CHPs were accessing more expensive borrowing than an average owner-occupying borrower, despite probably being a safer bet for lenders because CHPs have guaranteed income - and not just any guaranteed income, income that is guaranteed by the Crown, something other businesses, and even owner-occupiers don’t have.
He said other countries recognised social housing as a special asset class. Doing so here might require the Beehive to have a word to the Reserve Bank to reclassify CHPs, which, if recent history is anything to go by, is unlikely to be welcomed by the Bank, which tends to look unfavourably at any Government goal coming between itself and its twin tasks of controlling inflation and maintaining a stable financial sector.
Glaudel said there were other options.
In Australia, the Government used the power of its own balance sheet to offer CHPs borrowing on far less costly terms than could be achieved on the private market. It would effectively allow CHPs to piggyback on the Government’s balance sheet, much the same way Kāinga Ora does. It would give CHPs access to cheap borrowing, but using the balance sheet to offer private organisations discounted borrowing rates is guaranteed to rankle Treasury .
It looks likely that the Government may attempt to work out some long-term contracts for CHPs to make this security of funding clearer, in the hope of lowering their borrowing costs. But this doesn’t necessarily solve the debt problem. It simply shifts it from Kāinga Ora to CHPs, which have less capacity to deal with it.
Glaudel acknowledged that CHPs, like all private and public providers of housing, used “a debt-based model”.
Outsourcing that borrowing to even smaller organisations in the private or community sector won’t be the solution - in fact it would almost certainly make the problem worse because no private organisation, however secure, is likely to be able to borrow on terms as favourable as the government itself. House for house, the government’s borrowing costs will always be cheaper.
Ten year Treasury yields, a key indicator of the government’s borrowing costs, are currently about 4.7 per cent, residential mortgage rates are above 6 per cent, and commercial lending rates are higher again, with some in double-digits.
Put another way, $10b of government borrowing for Kāinga Ora would cost about $480m in servicing costs.
Even if CHPs or another private provider were to access borrowing at rates akin to residential mortgage borrowers, their cost to borrow the same amount of money would be over $650m, a difference of about $170m a year. This means that even if lending conditions are made more favourable to CHPs, they will be unable to borrow as cheaply as Kāinga Ora, constraining their ability to replace it as a provider of housing supply.
Johnson is deeply sceptical the community sector has the answer.
“They don’t have access to cheaper finance [and] they don’t really have the balance sheets to run it,“ Johnson said.
“The reality is that, you need quite a bit of capital to stack up a project with the IRRS. I don’t think the CHPs, with with a few exceptions, have got that capital,” he said.
“Even with that capital, at the moment, they’re not going to build that number of houses that are required,” he said.
English turned the argument on its head, arguing that Kāinga Ora had misused its access to cheap capital.
“Historically, the Crown has kidded itself about the cost of owning state houses. It’s tended to borrow cheaply when actually it’s a higher risk asset - and you can tell that that’s the wrong strategy just by driving around looking at the houses: despite having access to cheap capital, its housing stock is poor,” English told the Herald.
“It’s a high risk asset running rental properties and running them for challenging tenants ... there’s a real cost and the Crown has historically not been willing to pay that cost,” he said.
One thing CHPs do have on their side is lower costs per unit of housing.
English noted that Kāinga Ora’s build costs seemed to be too high, implying CHPs could do better. MHUD agreed.
In September of last year it sent a report from a quantity surveyor back to ministers saying that Kāinga Ora’s build costs were higher than “a modest market home”, but they didn’t put this down to incompetence on Kāinga Ora’s incompetence.
Instead, they said this was due to “build features associated with Kāinga Ora’s requirements”. (Kāinga Ora requires houses to be fitted with certain accessibility features, given the large number of tenants with disabilities).The surveyor said the cheapest way to bring on new houses was through “acquisition”, meaning buying an existing home. That’s true, but it’s something the current Government fiercely criticised in opposition because it doesn’t add any additional houses to the national housing stock, and simply pushes up the price of market housing.
Back in 2022, National’s Nicola Willis slammed this approach saying “the Government’s buy-in approach does nothing to increase the supply of new homes, and instead pumps up the prices of existing houses”.
Kāinga Ora is in the business of being a landlord. It’s easy to think that renting properties to people who can’t afford them is a losing bet, but the way Kāinga Ora’s accounts work, this isn’t the case.
Kāinga Ora charges market rents for its properties and receives that income by sending the bill to tenants who pay no more than 25 per cent of their income, and the Ministry of Housing and Urban Development tops up the remainder with IRRS.
While this is certainly costly for the Crown, for Kāinga Ora, effectively sending two bills for one rental helps it stay as solvent as any other large landlord.
This creates perverse incentives for Kāinga Ora.
As market rents rise, so does its revenue. Like many landlords, Kāinga Ora can do quite well out of a housing crisis.
The same goes for CHPs, who are given the same formula for accessing IRRS payments: they charge market rent, for which the tenant pays no more than 25 per cent, and the Crown pays the rest.
This is a smart approach for creating a sector of solvent social landlords, but it means the Crown is left with a very expensive rent bill.
Recently, however, this hasn’t been the case.
Johnson’s research shows that the per unit IRRS received by the Kāinga Ora has actually been falling in real terms, despite market rents rising.
He reckons the IRRS is “an administrative number not much related to the foregone market rents it is supposed to be based on”.
This is good news for the Crown, which saves money, but bad news for Kāinga Ora which is trapped offering a commercial product but being unable to charge a commercial price for it .
The other question is whether Kāinga Ora’s operating model can ever work commercially. Many private landlords aren’t making cash profits each year. Those landlords aren’t really in the business of renting houses. Instead, they’re in the business of farming capital gains, which are collected when a rented property is sold.
Kāinga Ora doesn’t have that luxury.
The problem with that is that while Kāinga Ora might not be getting the full market price for its rentals or have the ability to flick them for a capital gain, it is very much charged commercial prices for building them.
This, over time, has created a deadly mismatch for the organisation, in which commercial costs are not balanced out by the means of earning a commercial return.
“The misalignment between the value of income-related rent subsidies and the inflationary pressures faced by Kāinga Ora as a business are a particular weakness in the organisation’s operating model,” Johnson’s report concluded.
Kāinga Ora itself wants to shift away from this structure. In a 2022 paper, warning that its financial position was becoming unstable, the organisation proposed cutting things like bringing existing housing stock up to modern heating and accessibility standards in order to save money. The organisation warned that just bringing these old homes up to standard at a time of high costs and high interest rates would add $9b to its borrowing position by 2081.
It pleaded with the Government to “move the revenue model away from market rent/IRRS model to one that moves more directly [in] line with the underlying cost structure”.
There are even stronger perverse incentives. Firms that build housing would typically respond to the economic cycle by putting off construction during a period of high interest rates and construction costs. You can see this occurring at the moment as the number of consents issued for new homes falls.
As a commercial landlord, Kāinga Ora should be following suit.
If its income were actually determined by market rents, it should listen to what the market is saying and cut back on its expensive build programme.
In social terms, maintaining the build programme despite the economic cycle is a good thing, but for an entity designed to stay solvent, it spells disaster. In 2022/23, for example costs grew by 23 per cent, while income from IRRS grew by just 2 per cent.
Financially, the organisation needs to pull back.
The problem is that the best, and perhaps only way of doing this, is to build fewer houses, and substitutes an economic problem for a social one.
Thomas Coughlan is Deputy Political Editor and covers politics from Parliament. He has worked for the Herald since 2021 and has worked in the press gallery since 2018.
"When it doesn’t perform, everybody notices. When it does perform, nobody notices.”