Sheet 01
Load case — the problem
Half the state stands on one cracking column
New Zealand leans on personal income tax harder than almost any country on Earth: 41.7% of all tax revenue, against an OECD average of 23.7% — the third-highest share of 38 member countries. Every payslip in the country is a structural member.
That was sound engineering for the economy it was designed in: a nation of stable, salaried, lifetime jobs. It is not the economy being built around us. Automation is shifting output from wages to capital, and AI is doing it faster. Platform work is decoupling income from employment, and high earners can now choose their tax residence from a departure lounge.
"Almost 40 percent of global employment is exposed to AI… In advanced economies, about 60 percent of jobs may be impacted."IMF — AI Will Transform the Global Economy, January 2024
A tax system is infrastructure. When an engineer finds more than half the load on a member that's losing capacity, they don't wait for it to fail — they redesign the load path. That is all this site proposes.
Sources: OECD Revenue Statistics 2025 — New Zealand (2023 data) · IMF — AI Will Transform the Global Economy, Jan 2024
Exhibit A
Where the load sits today — FY2025
Treasury's audited accounts, year ended 30 June 2025.
Internal circulation — the state paying itself (≈$12.1b; Sheet 02)
Data table & source
| Tax | FY2025 revenue | Share |
|---|---|---|
| Personal income tax (PAYE, other persons, FBT, net) | $61,849m | 51% |
| GST (net) | $29,550m | 24% |
| Company tax (incl. NRWT, net) | $17,496m | 14% |
| RWT on interest & dividends | $4,507m | 4% |
| Excises, duties, levies | $7,656m | 6% |
| Total taxation revenue | $121,058m | 100% |
Source: Financial Statements of the Government of New Zealand, year ended 30 June 2025, Note 4.
Sheet 02
Load analysis — dead load
A fifth of the work column is the state paying itself
Before sizing a structure, an engineer separates live load from dead load — the weight of the structure itself. Do that to the $66.4 billion work column and something remarkable appears: about $12 billion of it never enters the Crown from outside at all. The government pays a nurse, a teacher, a constable — then takes PAYE straight back out of the cheque it just wrote. It pays NZ Super gross, then withholds tax from its own pension. It does the same with Jobseeker and every main benefit. Revenue and expense both inflate by the same amount, and cancel.
The pieces, each from a source already on this site: the Crown's audited wage bill is $35.0 billion (FSGNZ Note 9 — salaries and wages, excluding employer KiwiSaver, which isn't taxed through PAYE); at the ~23% average source-deduction rate implied by IRD's PAYE take, that's ≈$8.0b of the Crown taxing its own payroll. Tax withheld from gross NZ Super — ~13.5%, computed from the same Work and Income gross/net tables the calculator below uses — is ≈$3.1b. Tax on the main benefits adds ≈$1.0b. Together: ≈$12.1b of internal circulation — the deck's own weight, carried by the deck.
This isn't a conspiracy, and the official record shouldn't be restated: gross treatment keeps a public nurse's payslip comparable with a private one's, and every OECD statistic on this site is gross, so the 41.7% comparison stays honest only if the headline stays. But flows are not net transfers. The genuine external load that earned income carries is nearer $54 billion than $66 billion — and any argument about replacing it should know the difference.
Here is the rare thing about this honesty adjustment: both halves point the same way. The work column carries less real load than 55% suggests — and replacing it costs less than the headline implies, because a clean transition simply rebases public salaries, Super and benefits to today's net amounts. Nobody's take-home changes by a cent; the churn just stops being pumped in a circle. That takes ≈$12b off the deck, and the auto-balanced land value tax falls from about 4.15% to about 3.4%. The load test below has the toggle — dissolve it yourself.
Sources: FSGNZ 2025, Note 9 (salaries & wages $35,039m) and Note 4 (PAYE source deductions $52,460m) · Work and Income gross/net rate tables (tax share of NZ Super ≈ 13–14%) · main benefits $9.6b from Note 8, taxed at the 10.5% bottom rate. All three are deliberately simple estimates, printed so you can argue with them.
Exhibit DL
The work column, X-rayed
$66.4b of tax on earned income, FY2025 — split into what arrives from outside the Crown and what is internal circulation.
External load — $54.3b
Dead load — the state paying itself, ≈$12.1b
Income tax on people & savings, FY2025
$54.3b external
$12.1b
PAYE on the Crown's payroll
≈$8.0b
$35.0b of audited public wages × ~23% average source deduction.
Tax withheld from NZ Super
≈$3.1b
~13.5% of $23.2b gross — a smaller pension wearing a tax costume.
Tax on main benefits
≈$1.0b
Jobseeker, supported living & sole parent are paid gross, then taxed.
Sheet 03
General arrangement — the architecture
Three columns that can't run away
The redesign rests on one legal distinction: you exist without an Act of Parliament; a company doesn't. Natural persons are never taxed on what they earn, save, or build. The state is funded instead by what the community creates, what people choose to spend, and what exists only by legislative grant.
Column I · New
Land value
$1.5 trillion base · immobile
An annual tax on the unimproved value of land — the number already printed on every rating notice. Buildings, renovations and improvements are never touched. Site value is created by the community around the site; the community collects it back.
Replaces personal income tax & RWT · valued 3-yearly by law since 1896
Column II · In service
Consumption
GST 15% · $29.6b in FY2025
The cleanest consumption tax in the OECD stays exactly as it is: broad, single-rate, almost no exemptions. It taxes what you choose to spend — never what you earn, save, or invest — and it's collected in small pieces along the whole production chain.
671,324 GST-registered businesses already file it (IRD, 2025)
Column III · In service, renamed
Privilege
Company tax 28% · $17.5b in FY2025
Company tax stays — rebuilt on an honest foundation. It is the price of limited liability, separate legal personality, and access to capital markets: privileges that exist only by statute and are enforced entirely at public expense. Prefer not to pay it? Trade with full personal liability.
756,821 companies on the register (Companies Office, June 2026)
Decommissioned
Income tax
$66.4b to relocate
No tax on wages. No tax on savings interest or dividends. No tax on capital gains. PAYE disappears from every payslip; the IR3 from every autumn. The century-old distinction between "income" and "capital gain" — the engine of every avoidance scheme — stops mattering, because neither is taxed.
The load moves to Columns I–III — tested below
The full design — including why executive pay loses its deductibility while ordinary wages keep it — is on the plan page.
Sheet 04
Alignment check — the incentives
A government that only gets a raise when you do
Here is the part of the redesign nobody talks about, and it may matter more than the rates. Every tax base teaches the government a behaviour. Today's system pays the Crown when nominal wages rise — so inflation itself is a revenue policy, and bracket creep quietly raises taxes without a single vote in Parliament.
Under the modern architecture, there are exactly three ways for the government to collect more revenue — and every one of them requires the country to actually do better first:
If the Crown wants more GST
It must grow the economy
GST rises only when New Zealanders earn enough to spend more. Every policy that helps businesses sell, hire, and produce pays the government back through the till — and every recession docks its pay. The Treasury's own forecasts show it: GST tracks the economy, dollar for dollar.
If the Crown wants more land value tax
It must build the country
Land value is created by what surrounds the land: transport, schools, hospitals, jobs, safe streets. Under LVT, a rail line that lifts land values along its route pays for itself through the very values it created. Infrastructure stops being a cost centre and becomes the revenue engine.
If the Crown wants more privilege revenue
It must make NZ a good place to run a company
The privilege charge is a slice of profits. It grows when enterprises are profitable — when courts are fast, rules are stable, and markets are open. A government hungry for corporate revenue must keep the privilege worth paying for.
And what it can no longer do
Profit from inflation and bracket creep
With no income tax there are no brackets to creep. The state's revenue is chained to real prosperity — spending, land value, profit — not to the nominal wage drift it can engineer itself. If it wants more than growth provides, it must raise a visible rate and defend it at an election.
And it hands every voter the bill
There's a reason the biggest tax in New Zealand is also the least talked about. PAYE is engineered to be invisible: the money leaves your pay before you ever hold it, and for the 3.57 million New Zealanders on automatic assessment there isn't even a form — IRD's proudest metric is that 85% of them never have to think about income tax at all. What you never see, you never question. Nobody has ever marched down Lambton Quay over PAYE; every council in the country hears about the rates bill.
Economists call this salience, and it is not a side effect — it is the design. Withholding was invented so the state could take more with less friction, and the man who helped build it said so himself. Milton Friedman spent 1941–43 at the US Treasury constructing the machinery of collecting income tax at source, and counted it among the great regrets of his life:
"It never occurred to me at the time that I was helping to develop machinery that would make possible a government that I would come to criticize severely as too large, too intrusive, too destructive of freedom. Yet, that was precisely what I was doing."Milton Friedman, Two Lucky People (1998) — on inventing income-tax withholding
New Zealand switched the same machine on in 1958, and it has run silently ever since. The same Friedman, remember, called the tax on unimproved land value "the least bad tax" — and under this architecture his two verdicts finally point the same way. The withholding machine is dismantled, and the price of government arrives the way a rates bill does: a number, on paper, addressed to you. A tax you can read is a government you can audit. When every household sees exactly what the state costs, every household starts asking the questions that invisible money never provokes — is the department delivering? was that programme worth it? should the rate rise, or the spending fall? That conversation is what the spending page exists to arm — and our companion site taxdollars.nz takes it further, breaking the whole $141.7 billion down interactively to dollars per person. It is precisely the conversation withholding switched off.
Sources: Milton & Rose Friedman, Two Lucky People: Memoirs (University of Chicago Press, 1998) · FEE, Wartime Origins of Modern Income-Tax Withholding · IRD Annual Report 2024–25 (3.57m automatic assessments; 85% required no further action).
It retrains the country's savings, too
The incentives don't stop at the Beehive. Today the tax system quietly steers every household's savings into land and away from the world. Buy a rental and the location gain is untaxed. Buy more than $50,000 of overseas shares and you enter the FIF regime — taxed every year on a deemed 5% return, whether the shares rose, paid nothing, or fell. It is complex enough that ordinary savers and their advisers simply avoid it, and honest enough about nothing: in a down year, the only winner is the government. Even the current government has conceded the pain — its 2026 proposal merely lifts the threshold, which was set in the year 2000 and has never moved.
Under the architecture the steering wheel turns the other way. The FIF regime vanishes with the rest of personal income taxation — a stake in an overseas startup, an index fund, a global market is simply untaxed, while land carries an annual tax. The safest-feeling investment in New Zealand stops being a section in Auckland and starts being whatever is actually productive.
Where the incentives point today
Savings flow into land
Location gains: untaxed. Offshore diversification: the FIF regime's deemed 5% — tax due even in a losing year, plus the paperwork. The rational Kiwi portfolio has been a mortgage on dirt for thirty years, and house prices show it.
Where they point under the architecture
Savings flow into enterprise
No FIF, no tax on dividends or gains — global markets and overseas startups are open to every saver, at last on equal footing. Holding land now carries the annual tax; building and backing things carries none. Capital goes looking for productivity.
FIF regime: IRD — foreign investment funds (de minimis NZ$50,000 cost, fair dividend rate 5%) · IRD tax policy information sheet, May 2026 (proposal to lift the threshold to $100,000 — unchanged since 2000).
Exhibit B
GST is the economy's dial, live
Net GST revenue and the economy that drives it, indexed to 2000 = 100. When the country grows, so does the Crown's GST — automatically, at the same slope.
Net GST revenue
Nominal GDP
Sources: Treasury tax outturn data (net GST as collected by IRD & Customs, June years) · Treasury fiscal time series (nominal GDP, June years). The step at FY2011 is the GST rate change from 12.5% to 15% on 1 October 2010 — a visible, voted-on rate rise, which is exactly how the architecture says revenue beyond growth should be raised.
Sheet 05
Forward load case — the robots
When the robots clock in, the state still gets paid
Here is the stress test every tax system now faces: a robot pays no PAYE. Under today's architecture, every task that moves from a person to a machine is a leak in the revenue base — and the flow has started. The World Economic Forum's employer survey expects roughly a third of all work tasks to be done by technology alone by 2030, up from 22% today. The IMF puts 60% of advanced-economy jobs in AI's blast radius. For New Zealand specifically, PwC estimated a quarter of jobs at high automation risk by the mid-2030s.
The machines are no longer hypothetical. There were 4.66 million industrial robots working in the world's factories in 2024, growing 9% a year. Humanoids shipped in the tens of thousands last year; Bank of America's current forecast has that reaching 1.2 million a year by 2030, and Morgan Stanley sees over a billion in use by 2050. Every forecast could be wrong by years — the direction isn't.
Now watch what happens to this architecture. Nothing.
Industrial robots in service
4,664,000
IFR, 2024 — +9%/yr; installations pass 700,000/yr by 2028.
Humanoids forecast, 2030
1.2m/yr
Bank of America, Mar 2026 — from ~20,000 shipped in 2025.
A home humanoid, today
US$499/mo
1X NEO — subscription first, US$20k to own. US deliveries from 2026.
Tasks done by tech alone, 2030
~34%
Up from 22% today — WEF Future of Jobs Report 2025.
Robots on the payroll
Their output lands in company profits
A robot pays no PAYE — but it doesn't invoice, either. Whatever it produces flows straight to the bottom line of the company that owns it, where the privilege charge takes its 28% before a dollar reaches shareholders. No "robot tax", no definition of what counts as a robot: the profits are the meter.
Robots in the home
Every subscription carries GST
Households won't mostly buy robots — they'll rent them from companies, the way 1X sells its NEO at US$499 a month and New Zealand's own Halter runs cow-collar robotics at $9.90 per cow per month. The IFR reports robot-as-a-service fleets growing 31% in a year. Every monthly robot bill is GST-carrying revenue for a privilege-charged company — a brand-new consumption category the architecture taxes automatically.
Robots on the land
The dark factory still pays land value tax
A warehouse can run without workers; it cannot run without a location. Robotic factories, automated ports and data centres sit on exactly the land the LVT taxes — and the better the automation, the more that well-located land is worth. The column that owes nothing to wages doesn't notice who does the work.
Run the same five years against the current system and the contrast is stark: every automated checkout, milking shed and warehouse pick quietly shrinks the PAYE base that carries half the state — while the profits it creates are the only extra thing the Crown collects on. New Zealand automates because our labour is scarce — the kiwifruit harvest alone needs over 20,000 seasonal workers, which is why Tauranga's Robotics Plus (now Yamaha-owned) builds orchard robots and 77% of rotary milking sheds already run partial automation. A tax system should cheer that on, not depend on its failure. This one does: if the robots come slowly, untaxed wages carry on; if they come fast, profits, subscriptions and land carry the state. Either way, the deck stays supported.
Sources: IFR World Robotics 2025 (4,663,698 operational; 542,076 installed 2024; 700,000/yr by 2028) · Bank of America, Physical AI (Mar 2026) (~20,000 humanoids 2025 → 1.2m/yr 2030) · Morgan Stanley (Apr 2025) ($5t market, 1b+ humanoids by 2050) · 1X NEO order page (US$499/mo or US$20,000) · IFR service robots 2025 (RaaS fleets +31%) · Halter pricing · WEF Future of Jobs 2025 · PwC NZ (2018) · Yamaha–Robotics Plus (2025). Humanoid volume figures are bank forecasts, dated and attributed — today's actual shipments are ~20,000/yr; the argument doesn't depend on which forecast wins.
Sheet 06
Structural test — the load test
Can three columns carry $121 billion? You decide what they carry.
This is the whole argument, live — in two connected parts. First, set the load: the deck starts at every dollar the Crown collected in FY2025, and you can lighten it by privatising whole functions of the state. Then set the GST and privilege rates; the land value tax solves itself, rising or falling until the columns exactly carry whatever state you chose. And because your rating notice is a slice of the same arithmetic, part two turns the national rate into your household's bill — live, as you move the sliders.
The load test · part one
Set the load. Set the rates. Watch the deck.
All figures are FY2025 actuals held static: $1.5 trillion of non-government land, GST raising $1.97b per point of rate, company tax $625m per point. Real-world responses (land prices falling as LVT rises — partly the point) are discussed under the diagram.
The load — what the columns must carry
Privatising removes a function's FY2025 cost from what taxes must raise — the cost doesn't vanish; households buy it directly instead (honesty note below). Ending Super stops the payment itself, to roughly 900,000 superannuitants. Dissolving the churn removes the dead load from Sheet 02: take-home pay unchanged, the circular $12.1b just stops being pumped.
Column I — land value tax
4.2%
Per year, on unimproved land value only. Untick to set the rate yourself and see the surplus or shortfall.
Column II — GST
15%
Today's rate is 15%. Each point raises about $1.97b.
Column III — privilege charge (today's company tax)
28%
Today's company tax rate is 28%. Each point raises about $625m.
Land value tax$0
GST$0
GST, second round — untaxed pay spent$0
Privilege charge (company tax)$0
Excises & levies$0
IRD running costs saved$0
Total raised$0
Static arithmetic: FY2025 revenue ($121.06b target, Treasury Note 4), $1.5t taxable land base (Stats NZ national balance sheet, Dec 2025, all non-government land before design exclusions such as Māori freehold land), straight-line rate scaling for GST and company tax. Second-round GST: the $66.4b of abolished income tax & RWT stays in household hands; we deduct the LVT falling on household-owned land (46% of the base — Stats NZ), assume 75% of the remainder is spent, and count GST at rate÷(100+rate) of that gross spending. It moves with your sliders: raise the LVT and the second round shrinks, honestly. Smaller IRD: with no personal income tax, no RWT, no FIF regime and no IR3s, we assume half of IRD's $756m operating appropriation (IR annual report 2024–25) is saved — counted here as revenue-equivalent. The load toggles: "privatise" removes the FY2025 core Crown cost from what taxes must raise, holding the deficit unchanged — health $30.3b (Treasury functional classification); schools & ECE $17.0b (the $20.9b education function less Vote Tertiary, which stays publicly funded, fees-and-loans system unchanged). The cost does not vanish — it lands on households as direct bills, and the calculator in part two charges them at prices set so the loop closes exactly: health $5,693 × 5.32 million residents = $30.3b, schooling $14,400 × ~1.18 million children = $17.0b. Every dollar off the deck reappears on a household's bill somewhere. Those prices assume private delivery at the public system's unit cost — no insurer margins or loss of Pharmac-scale purchasing added, no competition dividend claimed either. That neutrality is deliberate: no developed country has fully handed schools or healthcare to the market, so there is no clean evidence for either hope — the most market-driven rich-country health system (the United States) is also the world's most expensive per person, while voucher-style school competition (Chile, Sweden) has produced decades of genuinely mixed results. The health figure also includes ≈$1.3b of one-off Holidays Act back-pay, so the steady-state price is slightly lower. Per enrolled learner the schooling figure is ≈$16,200; we blend across all 0–17s because the calculator doesn't ask children's ages. The burden falls most heavily on older households (health) and families with children (schooling). The architecture is deliberately agnostic about the size of the state; the toggles show the arithmetic, not a recommendation. End superannuation: removes NZ Super's audited $23.2b (Note 8) from the load, along with the ≈$3.1b of tax the Crown was withholding from it — from both the deck and the second-round pool, and counted once if the churn toggle is also on. Be clear about what the toggle is: not an efficiency but the end of the state's single biggest transfer, roughly 900,000 pensions averaging about $26,000 gross a year. It exists to show the arithmetic of the largest line on the books, not to recommend it. Dissolve the churn: ≈$12.1b of the income tax being replaced is internal circulation — PAYE on the Crown's own audited $35.0b wage bill (≈$8.0b at the ~23% average source-deduction rate), tax withheld from gross NZ Super (≈$3.1b, from the same Work and Income tables the calculator uses) and from main benefits (≈$1.0b) — see Sheet 02. The toggle rebases those payments to today's net amounts: every take-home is unchanged, the deck loses $12.1b, and the same $12.1b leaves the second-round GST pool, since rebased workers and superannuitants see no windfall to spend. The churn also adjusts to whatever else you've toggled: end Super and its ≈$3.1b of withheld tax is gone; privatise health or education and their workers leave the Crown payroll (Health NZ's audited salaries of $11.56b; the $7.7b school payroll, 102,000 staff), taking ≈$2.6b and ≈$1.8b of payroll churn with them — each dollar counted once, wherever it lands. Auto-balance solves the land value tax rate so the columns exactly carry whatever load you set (capped at 8%). None of these estimates is Treasury-modelled; all are deliberately simple and printed here so you can argue with them. A serious LVT also lowers land prices as it capitalises — which shrinks the base at higher rates — so the real path is a staged, pre-announced transition with rates recalibrated as the base responds. Full honesty clauses on the data page.
Part two — your share of the state you just set
Your official land value is printed on your council rating notice (or free at qv.co.nz). This calculator uses the LVT and GST rates from the load test above, live — privatise a function or move a slider up there, and your bill changes down here.
The load test · part two
What would you pay?
Start from an example household — or enter your own numbers. Renting? Your land value is 0: you own no land, so your land value tax is zero — and your pay is untaxed.
Try an example
Static comparison at 2025–26 income tax rates — couples are taxed as two even earners, so a single-earner couple pays somewhat more today than shown — with LVT and GST at the live rates from part one above (with nothing privatised, the LVT auto-balances at about 4.2%), held at today's land values; "GST now" is fixed at today's 15%. If you privatise health or education in part one, your household's direct bills are added on flat national averages: health per person (all ages — real costs skew strongly to over-65s) and schooling per child aged 0–17. It ignores ACC levies, company and trust structures, and the land-price fall a real transition would produce (which lowers the LVT). A real scheme includes deferral as of right for the asset-rich, income-poor — see questions.
Example households: minimum wage $23.50/hr × 40 hrs (Minimum Wage Order 2025); median gross household income $109,556, YE June 2025 (Stats NZ); average household spending ≈ $88,000 in 2025 dollars (Stats NZ HES 2023, CPI-adjusted); $250,000 is comfortably a top-decile household income; NZ Superannuation from 1 April 2026 is $1,294.74 gross a fortnight for a single person living alone ($33,663/yr) and $1,968.56 gross combined for a couple who both qualify ($51,183/yr) (Work and Income rates table). Each example adapts to the household selector: the single variants use one full-time minimum wage, roughly the median full-time individual wage (≈$72,000), a $180,000 top-band earner, and the single-living-alone Super rate; couples use two minimum wages, the median household income, a $250,000 professional couple, and the couple Super rate. Direct-bill averages when functions are privatised: health $5,693 per person (core Crown health $30,311m ÷ 5,324,700 residents, Stats NZ); schooling ≈ $14,400 per child aged 0–17 (education function minus Vote Tertiary ≈ $16,976m ÷ ~1.18m children — the child count is our estimate from Stats NZ age shares). Flat averages both: Treasury's 2025 Long-term Fiscal Statement notes those aged 85+ cost more than five times as much in health spending as someone aged 60–64. Average residential land values from the latest rating revaluations: Wellington $621,000 (WCC, 2024), Christchurch $475,384 and Dunedin $340,650 (QV, 2025), Auckland ≈ $750,000 (computed from the official $1.29m average capital value — Auckland publishes no residential land-value average). High-earner land values are leafy-suburb single-house estimates (Remuera / Oriental Bay / Fendalton / Māori Hill tier), cross-checked against QV property records.
Sheet 07
As-built survey — the machinery
Every part is already in service
This architecture needs no new bureaucracy, because New Zealand — almost uniquely — has been maintaining every component for generations:
- The land register: 2,791,560 parcels in the LINZ cadastre, every boundary surveyed and published as open data; ownership guaranteed by the state since 1870.
- The valuations: every rateable property receives an official land value, separate from its buildings, at least every three years — audited by the Valuer-General, as required by the Rating Valuations Act 1998.
- The GST machine: 671,324 registered businesses already file it, chain-collected, with the OECD's cleanest base.
- The companies register: 756,821 companies already file and pay the privilege charge under its old name.
You cannot hide a hectare of Epsom in the Cayman Islands, and a company that wants limited liability must stand on a public register to get it. The two hard problems of tax administration — finding the base and valuing it — were solved here before anyone alive was born.
Sources: LINZ NZ Primary Parcels · Rating Valuations Act 1998, ss 2, 7, 9 · IRD GST filer statistics, 2025 · NZ Companies Office statistics, June 2026
Parcels surveyed
2,791,560
LINZ cadastre — the tax base, mapped to the boundary peg.
Land revalued
every 3 yrs
By law, with statutory objection rights.
GST filers
671,324
Businesses already collecting Column II.
Registered companies
756,821
Every one already pays for its privileges.
Sheet 08
Issue register — go deeper
Check the drawings yourself
Every figure on this site traces to an audited account, a statute, or an official dataset. Start where you're most sceptical.
Data room
The government's own numbers
Twenty-five years of net GST and company tax, Treasury's forecasts to 2030, the OECD comparison, and the $1.5 trillion land base — every chart with its table and source.
Drawing set
The architecture in full
The natural-person / statutory-entity distinction, the three columns, the executive-pay deductibility fix, and why the design sets the how while Parliament keeps the how much.
Cross-examination
The hard questions
Grandma's villa. Māori land. Falling land prices. "GST hits the poor hardest." "Companies will leave." The strongest objections we know of, answered without spin.
Companion site · external
taxdollars.nz — where it all goes
The other side of the ledger, interactive: every dollar of the Government's $141.7 billion spend in 2024/25, from NZ Super to defence — what each line costs per person, and where your own tax goes.
Provenance
The single tax — where Column I comes from
Henry George's 135-year-old idea, New Zealand's own land tax of 1878–1990, a century of land-value council rating — and what the modern architecture keeps, and deliberately changes.