Skip to main content

Wealth report – a fruit salad of information

Tax Alert - May 2023

Tax has been front of mind for the general population since the release of Inland Revenue’s High-Wealth Research Project (the Report) and the companion piece prepared by The Treasury looking at Tax and Transfer Progressivity in New Zealand. Given the size of the work produced and the speed with which the media cycle works, the attention inevitably goes towards the numbers conveniently included within the 2 page press release or highlighted by the Minister of Revenue Hon David Parker in a speech, and those become the commonly accepted ‘facts’ – “the ‘rich’ 311 family units surveyed only pay 8.9% tax”, “only 7% of their income is personal income”, “their average wealth is $276 million and have total wealth of $85 billion”, “the ‘average New Zealander’ pays 20.2% tax”.

Reading the headlines, and then reading the report, can leave an ‘average New Zealander’ desperately trying to recall high school statistics lessons and thinking of that famous quote ‘lies, damned lies, and statistics’. So let me put this on the table, I am not a statistician, but I am someone who likes cold-hard facts and being able to analyse data, and to use another statistics idiom, ‘compare apples to apples’.

With that in mind, this article is a presentation of some of the information from the Report. But first, a statistics refresher:

The ‘mean’ is the average when the sum of a collection of numbers is divided by the count of numbers in the collection.

The ‘median’ is the middle result when all pieces of data are placed in order from smallest to largest.

Which is better, well it depends on context, but comparing a median result to a mean result is a bit like comparing apples to oranges, or to use another fruit analogy, cherry picking. The ability to misuse the mean when the median may be more appropriate is helpfully illustrated by the joke “Bill Gates walks into a bar, and everyone inside becomes a millionaire, on average” – here the use of the mean artificially inflates the wealth of the population in the bar.

Before getting into the detail, it’s worth noting upfront that frustratingly the Report does not clearly present data to allow readers to form opinions. Some data is only provided for certain years, some data has absolute numbers, some only has percentages, some have a mean but not the median, some has the median but not the mean, and some of the data has been presented only in bar graph format without totals (Inland Revenue have subsequently supplied me with some of this data) – it’s a fruit salad of approaches. Ultimately the presentation of the Report means that it is necessary to read the narrative to get the fullest possible picture rather than just looking at a series of numbers in a table and jumping to conclusions - possibly a good thing, provided you're not time poor.

In the interests of privacy, the Report does not provide information on the wealth of any individual/family group, but it is very clear from the data that there is a significant range of net worth’s, and one can only assume there must be some significant outliers – to play on the joke above, Bill Gates walks into a bar of 311 millionaires, and everyone inside becomes a billionaire, on average.

Those who are interested in juicy details about New Zealand’s wealthiest families will be disappointed, there is no information about the highest net worth, or how many billionaires are included in the population.

Net worth distribution

Net worth








Less than $50 million














$50 million - $250 million














Above $250 million














Median total net worth

Median total net worth







Mean net worth

Mean total net worth







So, from the above you can ascertain that total wealth (mean x 311) was $64 billion, $73.9 billion and $85.8 billion in 2015, 2018 and 2021 respectively.

You can also conclude:

  • that wealth is increasing, with the population moving up the distribution bands between 2015 and 2021;
  • that 75% of the sample population has less than $250 million wealth in 2021;
  • that using the median total net worth in 2021 of $106 million presents a more accurate presentation of ‘average’ net worth than the mean of $276 million.

It should also be noted, for the record, that the methodology used in the Report uses a mix of actual data and assumptions, so none of the numbers above are precise numbers, they are approximations. The growth in wealth in this population is, unsurprisingly, heavily made up of unrealised gains – some of which will be attributable to inflation (which is unadjusted, hence the gains include inflationary gains).

While not factored into any calculations, Inland Revenue asked the survey population to provide 50 years of information about inheritances. Over 50 years, an estimated total of $411 million was passed down within sixty-six family units; with the mean inheritance being $6.2 million and the median $1.3 million. The inference you can take from this data is other than some potential outliers with significant inheritances, the surveyed population is largely made up of ‘self-made’ millionaires.

The Report focuses on determining what is total economic income in order to quantify the effective tax rate (ETR) of the survey population. The Report notes that annual economic income for the project population varied between $1 billion in 2017 to $14.6 billion in 2021. The Report does not provide a clear breakdown of economic income, rather the data is provided through a bar graph at figure 12.1; the "All income values for project population" breakdown below has been supplied by Inland Revenue.

All income values for project population














Base income














Trustee income














Property income




























Business entities














All-income plus imputed rent














The Report notes that in 2018 the median family economic income is $8 million, and the mean family economic income is $22 million. Again, this indicates that economic income is skewed by outliers in the dataset.

The Report notes that income in 2021 is “relatively high” due to a combination of higher payments of dividends and salaries and buoyant asset prices. This is an understatement, as when you sum the ‘all-income’ amounts across the survey period, 2021 makes up 47% of all income. It seems that it’s all peaches and cream in 2021, but the dataset ends before the economy started to upset the apple cart in 2022.

When reading the table above, it needs to be noted that the process of building up to economic income is complicated and is focused on increases in asset values. In most cases, asset values are estimated based on assumed annual growth rates rather than verified information/actual data (the only exception is significant holdings of listed company shares), as such the Minister’s comment that “[f]or the first time, we have hard data confirming fundamental unfairness in our tax system” may over-estimate the hardness of that data.

The Minister noted in his speech that “[t]he financial affairs of the very wealthy are often complex and can involve partnerships plus hundreds of companies and multiple trusts.” While the Minister is talking in generalisations, the Report notes:

  • Total entities (which includes companies and trading trusts) included within ‘business income’: 2,695 (this results in a mean of 8.66 per responding family)
  • Total trusts: 1,279 (this results in a mean of 4.14 per responding family)
  • Total partnerships: 88 (this results in a mean of 0.28 per responding family)

Of the 2,695 business entities, it is acknowledged that the number could be higher, this figure only includes companies with gross assets and/or taxable income of greater than $1 million, or trading trusts with over $100,000 of income. However, these numbers don’t support the complex web of affairs portrayed by the Minister.

The Report notes that about 55% of businesses are owned via a trust.

The Report acknowledges that business income can be volatile, noting that business entity income ranged from -$899m in 2019 to $9,013 million in 2021.

Business entity income value














All business entities














Percentage of income over survey period














Included within business income are non-taxed distributions (e.g., returns of capital) to shareholders and estimated capital gains (calculated by estimating the equity value of unlisted entities using either a multiple of EBITDA, a revenue multiple or a multiple of asset values; with a 25% illiquidity discount applied). Data was collected on business sales, but the report writers decided not to include any data about actual realised capital gains; therefore, all amounts in the table are speculative.

The Report provides data about the average amount of property owned on a collective basis; being 5,107 residential properties and 1,879 non-residential properties; no data is provided about the mean or median number of properties owned by family unit. The value of the mean investment in property increased from $24.6m in 2015 to $43.3m in 2021. The calculation of asset values varied depending on whether the property was residential or non-residential. Residential properties have generally been valued using third-party automated property valuation services. Non-residential properties have been valued by taking the 2015 rateable value and applying an annual growth rate:

  • Forestry: 13.2%
  • Farms: 0.7% - 19.4%
  • Commercial: 5.4% - 21.1%
  • Industrial: 7.1% - 15.9%

Considering the growth rates being applied, the Report notes that 59% of calculated capital gains comes from non-residential property. Property gains have been calculated using the automated valuations rather than actual sales data due to concerns with mixing actual sales prices with automated valuation information from 2015. As such no data on actual realised gains is provided for property. Data in the Report, extrapolated from Table 9.3, indicates that approximately 32% of capital gains for property are realised, with the remainder being unrealised gains. 

Portfolio financial assets are investments in equity and debt instruments. The Report applies a variety of methods to attribute capital gains to different investment types, all of which apply a rate of return rather than actual data.

45 members of the survey population were identified as having significant holdings in listed companies. Because data in relation to share prices is more readily available, the value of shareholdings was tracked for a 17-year period from 1 April 2004 to 31 March 2021. Over this period, total capital gains were $6 billion, with only $1.7m being realised capital gains; noting that capital losses arose in 2008-2009 and 2015-2016. A significant portion of gains occurred in 2021 alone.

The Report presents a summary of the ETR for the survey population, which starts with base income (essentially what is taxable under the Income Tax Act) and then contrasts this to all-income (essentially all forms of economic income):

Type of ETR


Weighted mean1




Base income 












     -  All-income plus imputed rental ETR






     -  All-income plus imputed rental netting transfers






      -  All-income plus imputed rental and GST ETR






1 Note, the effective tax rate calculations in the Report use an 'income weighted-mean' rather than a 'simple mean'.

The bottom three rows of the table adjust tax on economic income to factor in the benefit of living in your own home compared to the cost of having to rent it, superannuation benefits (note, the median population age was 68) and GST. The statistic chosen to be focused on is the median ‘all-income plus imputed rental ETR’ of 8.9%, which is interesting as the Minister himself said “I prefer estimates of income that exclude imputed rents and capital gains on the owner-occupied home, at all income and wealth levels.”

What does this translate into actual amounts of tax paid? Again, the Report does not clearly present this information, with the following breakdown supplied to me by Inland Revenue. The years 2016-2020 were broadly similar, with 2021 having a large increase in tax due to the increase in the top personal tax rate which took effect from the 2022 tax year.

Tax paid by source


































































































2 This is lower because company tax has been attributed to individuals trusts due to dividends paid, not because companies paid less tax.

There is no year-on-year data provided about the median value of tax paid per year by family group.

In relation to GST, it is worth noting that the estimates of GST paid specifically exclude any GST paid on purchases of motor vehicles or housing stock because the amounts are large and purchased infrequently. The Report also notes that some high-wealth families reported spending in excess of their taxable income, but neglects to join the dots with the median survey respondent age being 68 (despite the Report highlighting that retirees spend their savings in retirement).

What statistics are missing

The Report was designed with a specific purpose in mind, being to understand the wealth of a population of families. While it aims to be objective, the objectivity of the report may have been better served by clearer data and less “sound bite” selections of data. This would have reduced the risk of the Report being seen as directed towards a particular outcome and avoided any question about the political neutrality of the Inland Revenue.

While not called out, the Report indirectly highlights some of the key issues that a capital gains tax design would need to address (if one was to be progressed as some commentators are now calling for): critically the subjectivity of valuations, the volatility in economic cycles and the treatment of unrealised gains and losses.

As a snapshot, while it is undeniable that the survey population has a lot of wealth (which, in reality, may actually bear little resemblance to the data in the report), that is just one side of the story. The volatility of the business income also serves to highlight the personal risk-taking of these individuals, with history showing us that fortunes can be lost a lot easier than they are made.

The research also didn’t extend as far as to understand the impact of these hard-working individuals on New Zealand; for example, the value that those businesses bring to the economy and society through building infrastructure or creating the goods and services we need and want, the number of employees working in their businesses (and the tax paid by those employees), the philanthropic activities and charitable donations made – often in a modest way.

So, what is next? Now it’s down to politics.

May 2023 - Tax Alerts

Did you find this useful?

Thanks for your feedback

If you would like to help improve further, please complete a 3-minute survey