Residential building has outgrown business investment. The crossover was 2017.
New Zealand has spent more of its annual output on houses than on all other fixed capital combined every year since 2017. In 1988, 27% of the country’s fixed capital investment was residential building. In 2024 the figure was 52%. Over the same period, business investment in non-residential buildings, plant and machinery has fallen from 11.6% of GDP to 7.3%. The house share of national investment has nearly doubled.
Those are not numbers anybody has tried to keep secret. They sit in Stats NZ’s national accounts, updated each quarter. But they are not routinely published as the lead item, and the compounded effect over 35 years is larger than most political discussion of the housing issue has allowed.
This article looks at what the Stats NZ and BIS numbers say about where New Zealand’s capital has actually gone since the late 1980s, and what that has produced.
Fixed capital investment has shifted to housing
Gross fixed capital formation is the accounting category for new buildings, plant and equipment. It is the physical stuff a country buys each year to build its future income. Stats NZ publishes it in nominal terms in the expenditure-side GDP tables, split by asset type.
| Year | Residential buildings (% GDP) | Non-residential + plant & machinery (% GDP) | Residential share of total fixed investment |
|---|---|---|---|
| 1988 | 4.2 | 11.6 | 27% |
| 1995 | 5.3 | 9.5 | 36% |
| 2000 | 5.5 | 8.2 | 40% |
| 2005 | 6.7 | 9.1 | 43% |
| 2010 | 4.5 | 7.2 | 39% |
| 2015 | 6.5 | 7.1 | 48% |
| 2017 | 7.6 | 7.1 | 52% |
| 2020 | 7.3 | 7.5 | 50% |
| 2024 | 7.8 | 7.3 | 52% |
Total fixed capital formation has moved within a narrow band — around 14 to 16% of GDP since 1988. The country has not increased the share of its output it sets aside for investment. What has changed is the composition. Houses grew; plant, machinery and non-residential buildings shrank as a share of the whole.
House purchase prices rose at nearly three times the rate of general inflation
The price signal matches the capital flow. Stats NZ’s Consumer Price Index includes a housing-purchase sub-group that tracks the cost of newly built dwellings.
| Year | Housing purchase (index 1990 = 100) | CPI all groups (index 1990 = 100) |
|---|---|---|
| 1990 | 100 | 100 |
| 1995 | 124 | 111 |
| 2000 | 138 | 119 |
| 2005 | 180 | 135 |
| 2010 | 216 | 155 |
| 2015 | 264 | 167 |
| 2020 | 326 | 180 |
| 2025 | 464 | 224 |
statsnz.cpi.cpiq.se904201; all-groups statsnz.cpi.cpiq.se9a.Over 35 years, house-purchase prices rose 364% while the general cost of living rose 124%. The gap is 240 percentage points. Housing purchase prices rose at close to three times the pace of general inflation. Rents, which Stats NZ has published separately since 1999, have tracked general inflation more closely, rising 71% against 93% for all groups over that window. The mismatch between purchase prices and rents is a well-known feature of the index methodology, and it does not undermine the broader point: the asset component of housing inflated dramatically faster than anything else households consume.
Business debt fell. Household debt tripled.
The Bank for International Settlements maintains a long-running series measuring credit to households and to non-financial corporations as shares of GDP, adjusted for structural breaks, back to 1991 for New Zealand.
| Year | NZ household debt (% GDP) | NZ non-financial corporate debt (% GDP) |
|---|---|---|
| 1991 | 32 | — |
| 2000 | 62 | 85 |
| 2005 | 83 | 90 |
| 2010 | 90 | 96 |
| 2015 | 90 | 85 |
| 2020 | 97 | 78 |
| 2025 (Q2) | 91 | 71 |
Household debt nearly tripled as a share of the economy between 1991 and 2020, and is now close to 90% of GDP. Non-financial corporate debt peaked in 2010 at 96% of GDP and has since fallen to 71%. Since 2015 New Zealand households have owed more, as a share of the economy, than New Zealand non-financial businesses. That is the opposite of the pattern in the 1990s and early 2000s.
The composition shift matters because the purpose of household and business debt is different. Household debt is overwhelmingly mortgages against existing or newly built dwellings. Business debt typically funds plant, equipment, inventory and the working capital of firms that employ people and produce exportable goods and services. When household credit rises and corporate credit falls, the banking system is doing more property lending and less productive lending. Under the current mix, that is what has happened in New Zealand for roughly a decade.
New Zealand is high, but not the highest among Anglophone peers
New Zealand is not alone in this pattern, and is not even the worst case among English-speaking developed economies.
| Country | Household debt % GDP (1991) | 2025 | Change |
|---|---|---|---|
| Australia | 47 | 114 | +67 pp |
| Canada | 58 | 100 | +42 pp |
| New Zealand | 32 | 91 | +59 pp |
| United Kingdom | 59 | 74 | +15 pp |
| United States | 63 | 68 | +5 pp |
| Germany | 51 | 49 | −2 pp |
Australia and Canada are both higher than New Zealand in absolute terms. Both Anglophone peers also had a larger household-debt build-up over the period. The United States peaked at around 100% of GDP in 2007 and then deleveraged sharply after the global financial crisis; it now sits below its 1990s level. Germany never engaged in the same buildup and has been flat.
The takeaway from the comparison is not that New Zealand is uniquely bad. It is that a group of English-speaking economies, including ours, ran a particular model — deregulated credit, tax-favoured owner-occupied housing, tight urban land supply, and floating exchange rates — that converted household income into mortgage debt faster than peer economies in Europe.
Where the borrowing actually comes from
New Zealand’s Net International Investment Position — the value of what New Zealanders own offshore minus what the rest of the world owns in New Zealand — is about −49% of GDP at end-2025. That is a large net foreign debt, but it is a substantial improvement on −85% in 1998.
| Year | Foreign assets owned by NZ (NZ$B) | Foreign liabilities of NZ (NZ$B) | Net IIP (% GDP) |
|---|---|---|---|
| 1989 | 13 | 57 | −63 |
| 1998 | 47 | 137 | −85 |
| 2010 | 155 | 307 | −78 |
| 2024 | 386 | 584 | −47 |
The improvement has two causes. Offshore financial assets held by New Zealanders — in KiwiSaver funds, pension funds, insurance reserves, ACC — have grown from NZ$13 billion in 1989 to NZ$386 billion in 2024. Much of that growth came after KiwiSaver began in 2007. Foreign liabilities grew too, but not as fast.
What the net-position number hides is the composition. A large share of the foreign-held liabilities is the lending book of New Zealand’s four Australian-owned retail banks, financed through their Australian parent companies and foreign wholesale funding. That lending is overwhelmingly mortgage lending. New Zealanders have partly financed their housing stock through a pipeline that runs from foreign wholesale markets, through four Australian-owned banks, to retail borrowers. The interest paid on those mortgages is a transfer from New Zealand borrowers to Australian bank shareholders; some is retained in the country as taxes and local wages, some is repatriated as dividends.
What the data settles
- Since 2017, New Zealand has invested more each year in residential buildings than in non-residential buildings, plant and machinery combined. That crossover has been sustained through 2024.
- Residential building’s share of total fixed capital investment has nearly doubled since 1988 (27% → 52%).
- Non-residential and plant investment has fallen from 11.6% of GDP to 7.3%. The country has not invested a larger overall share of GDP; it has redirected the composition.
- House-purchase prices have risen 364% since 1990, against 124% for the CPI all-groups index. The 240-percentage-point gap is a 35-year trend, not a recent shock.
- Household debt has tripled as a share of GDP since 1991 (32% to 91%). Corporate debt has fallen by a quarter since its 2010 peak. Households now owe more of GDP than businesses do.
What the data does not settle
The numbers show a clear pattern of capital reallocation; they do not prove the mechanism. Tax-favoured owner-occupied housing, restrictive urban planning, declining real interest rates, the end of the inflation era, financial deregulation from 1984, and the absence of a broad capital-gains tax are all plausible contributors, and the literature has not settled their relative weights for New Zealand specifically. The Tax Working Group of 2019 recommended a comprehensive capital-gains tax that would have addressed several of these factors; it was not adopted.
The BIS household-debt comparison also shows that the Anglophone economies (Australia, Canada, New Zealand, UK, US) moved together up to the mid-2000s and then diverged. The US and UK partly deleveraged after 2008; Australia, Canada and New Zealand did not. That divergence suggests country-specific policy choices — not global forces — determined the post-2010 path. Which choices specifically is a question this article’s data cannot resolve.
Finally, these are macro-level figures. They describe the direction of aggregate capital, not the welfare of individual households. A family that bought in 1998 and paid down its mortgage benefited from the capital-gain side of this trend. A family that rents or is paying down a post-2018 mortgage paid for it. Both are real, and both are inside the same aggregate.
Methods
Annual investment shares computed from Stats NZ National Accounts (SNA 2008) nominal expenditure-side GDP series, year ended March, sourced via the Gross Domestic Product December 2025 quarter release. House purchase and all-groups CPI use Stats NZ CPI quarterly series, annual averages. BIS household and non-financial corporate credit use the BIS WS_TC dataflow, quarterly, break-adjusted, retrieved via the BIS SDMX API. Net International Investment Position uses Stats NZ BPM6 annual series for total foreign assets and liabilities. All figures were retrieved on 19 April 2026.
Your turn
Does this capital-reallocation story match what you have seen — as a builder, a business operator, a property investor, or a first-home buyer? Add your view in the comments.