Source: Tax Justice Aotearoa
24 July 2026
The figures used by the Treasury to predict the rate of reduction in in-scope wealth in the event of a wealth tax being adopted, have been described as wrong by tax reform group, Tax Justice Aotearoa.
“There is no empirical evidence supporting Treasury’s apparent assumption that if you double the tax rate from 1% to 2%, you will get double the reduction in revenue from behavioural change. There is no evidence for a linear behavioural response from taxpayers to changes in tax rates,” says Tax Justice Aotearoa spokesperson Nick Miller.
A front page article in the NZ Herald yesterday relied on modelling Treasury undertook to assess the likely impact of the Labour Party’s proposed wealth tax in 2023 – a tax that was dropped following intervention from Labour leader Chris Hipkins. In 2023 Treasury estimated that a 1% would result in a 17.5% reduction in the in-scope wealth base, but noted the estimate was “subject to significant uncertainty”. The Herald article referred to the Green’s policy for a tax of 2.5% on wealth over $10 million and suggested that, using the 2023 methodology, 43.75% of the affected tax base would be moved off shore.
“The suggestion in the Herald article that a 2.5% wealth tax rate would result in a loss of 43.75% from the wealth base relies wholly on the assumption that there will be a linear behavioural response to increases in the wealth tax rate. This assumption is not mentioned in the 2023 Treasury advice to ministers cited in the Herald article,” says Miller.
“Behavioural changes are impacted by multiple factors, including the design of the tax, socio- economic pressures, geopolitical events, changes in tax in other jurisdictions, the efficiency of the revenue administration and any legislative responses to tax avoidance.”
Tax Justice Aotearoa has also criticised the Treasury’s assumption of a 17.5% reduction in the wealth base if a 1% tax was imposed on an exemption basis, as not appearing to be evidence based.
“The UK Wealth Commission Report from 2020 estimates a loss of 7-17% of the tax base on a 1% tax and this is, according to the report, based on evidence from countries with annual wealth taxes. The range alone indicates that estimating behavioural response is a highly inexact science”, says Miller.
“Furthermore the modelling in the Treasury paper appears to suggest a tapering off of the behavioural response. Table 2 in the paper shows an estimated yield of $1.5bn from a 1% tax applied to a $10m threshold and $2.7bn from a 2% tax applied to the same threshold. It is stated that the figures reflect an estimated 17.5% reduction in the wealth base for a 1% tax.”
“The report is silent about whether, as the Herald article suggests, the reduction in the base is linear – i.e. it would double from 17.5% to 35% if the tax rate doubled. The numbers in the report indicate an expected behavioural response resulting in a reduction in the base of 25% rather than 35% for a tax rate of 2%.*”
The Treasury does not appear to have modelled a linear response and has instead assumed that the behavioural response will taper off. The conclusions therefore drawn by the Herald article appear to be incorrect
*This point can be illustrated as follows:
A 1% tax charge on the $10m threshold yields $1.5bn after the 17.5% reduction. Therefore the yield before that reduction – the “gross” yield would be $1.8bn (1.5bn x 100/82.5). The “gross” yield from a 2% charge would be exactly double – around $3.6bn. Applying a 35% reduction (that is assuming a linear behavioural response) the expected yield would be around $2.35bn but the estimated yield in table 2 is instead $2.7bn. This indicates an overall reduction in the wealth base of about 25% rather than 35%.