Photo by Nick Fewings on Unsplash

There are a large number of trusts, particularly discretionary trusts, in Australia. This was highlighted in one of the Australian Labour Party’s (ALP) tax reform proposals in its 2019 federal election campaign (‘A Fairer Tax System: Discretionary Trusts Reform’). It is also evident in statistics released annually by the Australian Tax Office (ATO) – in 2015-16, returns were filed for 845,925 trusts, which is only 95,241 less than the number of returns filed for companies (941,166).

Practitioners have been both aware of the growing number and popularity of the trust since the late 1970s, and familiar with trusts being used for traditional reasons (intergenerational wealth transfer and succession planning) or for carrying on business; and being used alone or as part of more sophisticated structures. However, academic literature has only focused on the increased use of trusts since 2005, and then the focus was on public trusts with only very limited analysis (largely technical and applied analysis of select aspects of the tax treatment) of private trusts. The existing literature often attributed some of the growth and popularity of trusts to income tax. However, it never clearly identified which part of the income tax treatment or settings made using a trust so favourable.

My recent Sydney Law Review article fills this gap.

Entity taxation and flow-through taxation

There are two main paradigms that can be applied to tax income derived through a vehicle (vehicle here means a company, trust or partnership). The first is called entity taxation and the second is flow-through taxation.

The key difference between the two is that under entity taxation, the tax system recognises the vehicle as being a separate entity to its owners and applies tax at both the vehicle and owner levels. In contrast, under flow-through taxation, the vehicle is not treated as a separate entity. Rather, it is treated as what is referred to in tax as a “fiscally transparent” or a pass-through (pass-thru in US literature) vehicle. This means that tax is not applied at the vehicle level, it is only applied at the owner level.

This difference is apparent in many design aspects of the two paradigms: the treatment for losses (which are quarantined at the vehicle level under entity taxation or attributed to owners under a pure version of flow-through taxation); the way in which income is distributed (under entity taxation) or attributed (under flow through taxation) to owners; and whether tax preferences (anything that reduces the tax rate that applies to that income) are used up at the vehicle level (under entity taxation) or are carried along with the income to owners (under flow through taxation).

The use of trusts continued to grow despite the introduction of dividend imputation

Why do we care about these two methods? Because, anecdotally, the growing use and popularity of the trust has often been explained as resulting from the fact that the income tax treatment of trusts is closer on the design spectrum to flow-through taxation than the income tax treatment for companies. This is coupled with the acceptance of a legal precedent from the early 1900s that allowed a trustee to carry on business through a trust, changes to the income tax treatment for companies during the 1970s that made using companies more unfavourable, and the use of trusts in avoidance practices and marketed structures.

The extension of that view is that using a company should have become comparatively more attractive after Australia introduced the imputation system for companies in 1987 because the imputation design is more of a flow-through design than the system that preceded it. The problem is that this explanation runs contrary to and does not explain the continued growth in the use of trusts since 1987.

The article draws on an historical data table that was collated by the author in collaboration with the ATO, and then published as Snapshot Table 6 in the ATO’s Tax Statistics 2015-16. That table showed the number of returns filed for trusts, companies and partnerships for each income year between 1958-59 and 2015-16. This is much more extensive than any previous analysis. For example, the Coalition Government’s 2015 Re:Think Tax Discussion Paper only set out the number of companies, trusts and partnerships from 1990-91 onwards.

Analysis of the more extended data shows the following key points (numbers are based on returns filed):

  • Generally, the growth in the number of trusts has at least kept pace with, and has at times outpaced, the growth in the number of companies;
  • The number of partnerships peaked in 1993-94 and has been declining since. The number of trusts overtook the number of partnerships in 2002-03. In 2015-16, there were more than double the number of trusts (845,925) than partnerships (321,360).

The article also analysed further material from ATO Tax Statistics, including Trusts Table 1 from ATO Tax Statistics 2015-16, which showed the following points in relation to the period between 1996-97 and 2015-16:

  • At each point, the vast majority of trusts were private trusts. For example, private trusts comprised over 99% of the total trust population in 1996-97 and a little over 98.71% in 2015-16;
  • Consistently across this period, the majority of trusts were discretionary trusts, and the number of discretionary trusts has grown income-year-on-income-year with a few exceptions;
  • The total business income derived through trusts in 2015-16 ($368,468,000,000) was 16.78 times higher than the 1989-90 figure ($21,984,566,019). Most of this income is derived through micro, small and medium sized trusts (classifications used by the ATO).

Why use a trust?

The article argues that the continued growth in trusts post-imputation is due to a combination of the following factors:

  • As the primary rules for taxing trusts in Division 6 of the Income Tax Assessment Act 1936 (Div. 6) is largely patterned on a flow through model, income retains its character and any attached preferences as it passes to beneficiaries. This is not the case (even functionally) for income derived through companies;
  • Discretionary trusts allow for income splitting, which can reduce the overall tax liability of the trust. It is noted however that such trusts are subject to general and specific anti-avoidance rules.
  • Discounted/concessional treatment is provided to trusts under the capital gains tax rules. This can halve the rate that applies to certain capital gains.
  • The capital gains tax event that brings to account amounts that have not been previously taxed to the trustee and would not otherwise be taxed in the beneficiary’s hands (CGT event E4) does not apply to discretionary trusts.
  • The High Court’s decision in Harmer v Federal Commissioner of Taxation (1991) has been accepted as the precedent for a very broad definition for the key allocation mechanism in Div. 6 (present entitlement).
  • Although both Coalition and the ALP have, at different times, proposed to reform Div. 6, those attempts have been unsuccessful. This is in part due to politics (historically, the lobbying power of the National Farmers’ Federation and small business) and also the technical complexity involved.

The article argues that trusts continue to be the primary alternative vehicle, rather than the partnership, for two main reasons:

  • Trusts are more flexible. In particular, it is possible to give different people rights to the income and capital, whereas this is not possible using a partnership. Flexibility is greatest with a discretionary trust because the trustee’s power of appointment gives the trustee choice of appointee and amount, and there can be variation of appointee across time.
  • The application of the capital gains tax rules to partnerships were initially very uncertain.

The article finishes with a discussion of future directions and challenges for the income tax treatment of private trusts. The article sketches out a skeleton of features of a possible design for taxing private trusts that addresses the most problematic aspects of the design in Div. 6. The proposal could be easily adapted for a different alternative vehicle, if introduced.


This post is a high level summary of the article – “Why We Use Private Trusts in Australia: The Income Tax Dimension Explained” (2019) 41(2) Sydney Law Review 217.

This article has 4 comments

  1. Alex – very interested in this in the context of inter-generational wealth transfer. Very interested to know to whether you have done any work around Ken Asprey’s or Ken Henry’s proposals for death/gift duties or a wealth tax?
    Joe Roach

  2. Joe Thanks for your comment I am also interested. Peter Groenwegen the doyen of Australian Public Finance economists, believed that a wealth tax should have been
    introduced with the GST.

  3. Hi Joe and Wayne, many thanks for your comments, it’s really interesting to read your thoughts and nice to engage in discussion about this really good point.

    My apologies for the delay in responding, I’m currently on maternity leave.

    For background, my PhD thesis and work to date has focused on business trusts, with one exception noted below. I do think about intergenerational wealth transfer and we discuss it in the Masters level Taxation of Trusts and Tax Policy courses I teach at UNSW.

    Joe – I agree that the discussions and proposals in the Asprey and Henry Reports are really valuable in this area. The full article in Sydney Law Review that this post summarised raises a comment around taxation of capital that Professor Parsons made in the Asprey Report – see pages 250-1 in the full article.

    In another article, “The Rule Against Perpetuities – How We Got Here and Future Directions” (2017) 46 Australian Tax Review 71, I did some analysis of wealth tax and CGT. The article looks at wealth tax, wealth transfer tax and the avoidance rules (generation skipping tax, which confusingly for Australians is also called the GST) in the US at pages 92-5. If you email me ([email protected]), I’ll send you a copy.

    Anecdotally, after living over in the US for nearly 4 years now, I think about wealth tax a lot. We live in Seattle (Bill Gates’ hometown), a few suburbs over from both Gates’ and Bezos’ homes / compounds. And yet, the number of homeless people here is incredibly high – people live in “tent cities” under highway bypasses and in covered carparks. It seems like it is the result of several factors.

    Michael Graetz’s book “Death By A Thousand Cuts” is great at explaining the background to why the wealth tax thresholds in the US are so high and why there are so many exemptions.

  4. Hi Alex, It is good to see this part of your work in print. It demonstrates the importance of particular income tax history in understanding the role of trusts today and over the past four or five decades. This is particularly valuable in an international and comparative context because the uses of trusts and their fiscal, economic and social significance in each country reflect the history and conditions of that country.
    Joe’s point is a difficult one for an Australian researcher because we haven’t had inheritance taxes since the 1970s; we even privilege transfers on death by giving an effective CGT rollover. Any public discussion of the topic now is political poison. I expect that is why Henry did no more than wave his hand at the issue and say blandly that we ought to think about it. The old view, which is probably still sound, was that you can’t have a comprehensive tax base without taxing inheritances or estates. On the consumption side, this goes back to Lord Kaldor in the 1950s.
    Although any inheritance tax is leaky – Henry cited research suggesting that about 1/3 of a medium sized US estate typically escaped inheritance tax – it has an impact beyond the comprehensiveness and progressivity of the tax system. It is no accident that the US has many large charitable foundations established by very wealthy people. The knowledge that wealth passed privately to the next generation will be diminished by tax encourages philanthropy.
    There are other issues around the use of trusts in the US, notably asset protection, which can happen without any tax-avoidance angle. There is a lot of US literature about this, particularly on grantor trusts.

Leave a comment

Your email address will not be published. Required fields are marked *