Normative Bases

Egalitarianism

The 50% CGT discount is counter to anyone with even modest Egalitarian sentiments. It looks to have failed to boost investment in productive Australian assets, and has had major negative effects in the process: feeding housing unaffordability, advantaging the wealthy, and draining public revenue. Reforming or removing it would help realign investment incentives with the broader public interest.

Boosting Investors at the Expense of Owner Occupiers (especially First-Home Buyers):

Before we get to the empirical evidence showing the effects of this policy on the Australian housing market, let's take a moment to consider the incentives that it creates. Once we come to terms with these, it will become clear that housing speculation was bound to increase since its introduction (even if there are other factors that also drive house price growth in Australia). The bottom line is that the CGT discount and negative gearing together create a tax-driven incentive to invest in property for capital growth — in general and particularly at higher leverage. Under current rules, investors can deduct unlimited rental losses from their wage income (negative gearing), then later pay tax on only half the capital gain when the property is sold. This combination considerably increases the overall profit expectations of investing in a negative-cashflow asset, favouring higher leverage. Housing is specifically an investment where ordinary people can access significant leverage, with high confidence of steady growth at low volatility, boosted by rental income, so the 50% CGT discount naturally amplifies existing cultural tendencies to invest in real estate. Finally, once investors (recognising these incentives) start piling into housing, the cycle becomes virtuous. Reducing the CGT discount would correspondingly reduce the reward for this strategy. Negative gearing on its own was the status quo before the change to the CGT discount, and it did not provide the same level of incentives on its own (though we will briefly discuss whether the tax system would be better off without it also).

There is solid empirical evidence to support this theoretical prediction: in particular, that the period after the introduction of this policy saw higher property speculation at higher leverage, which helped crowd out other buyers, especially first-home buyers, and may have contributed to a lasting increase in real estate growth expectations.^[1] In 2016, the Grattan Institute published a paper titled “Hot property: Negative gearing and capital gains tax reform” that presents a comprehensive look at these issues. Figure 7 in this report (below) shows that, because of the CGT discount, the effective tax rate declines with higher leverage, dropping below 20% for borrowing at 80% of investment value. Figure 9 (below) illustrates the striking fact that almost all the growth in property investment since 1994 has been because of loss-making landlords. Figure 10 (below) shows that reported rental losses specifically started trending upwards a short-time after the introduction of the tax-discount policy, suggesting that the CGT discount made real estate investors more comfortable with negatively gearing their properties.

In short, the evidence bears out the predicted first-order effects: promoting property speculation and particularly at higher leverage. But what of second-order effects - in particular, pushing up prices and crowding out non-investor buyers? The strength of these effects is hard to determine without a strong quantitative model, but there is circumstantial evidence to suggest that these also occurred. A chart from this Aussie Home Loans Report (see below) shows a dramatic fall in the ratio of first-home buyers to investors right around the expected time period. As this Australia Institute report describes, it is also around the time that property prices started to really diverge from household income. Again, in the absence of a really strong quantitative analysis, we can disclaim the position that all or most house price growth over this period was directly linked to the CGT policy change, but there is enough circumstantial evidence to support the theoretical prediction that a strong boost to leveraged speculation would push up prices.

Wealth Inequality and Distributional Effects

The benefits of the CGT discount flow overwhelmingly to higher-wealth individuals, contributing to inequality. Since those with more assets are more likely to realise large capital gains, they reap the lion’s share of this tax concession. Data from the Australia Institute show that 73.2% of the benefit of the CGT discount flows to the top decile of income earners. So there is no doubt that this policy change has promoted inequality.

The other issue is the inter-generational effects. Younger people have a higher ratio of wage income to capital income so pay a higher effective tax rate under this regime than older generations. Meanwhile the ever-increasing rise in property prices (partly linked to this policy) further biases the equation against them. Inevitably, these dynamics create a greater dependence on inherited wealth, thereby locking in future inequality and undermining fairness.

Fiscal Cost and Budget Considerations

The 50% CGT discount represents a large tax expenditure, meaning significant forgone revenue for the government. A PBO analysis requested by Adam Bandt shows the cost of CGT discounts just applied to residential properties peaked at $9.24 billion in 2021-22 (and was $5.22 billion in 2023-24) while the combined cost of CGT discounts + negative gearing deductions (which, recall, are incentivised by the CGT discount) amounted to $10.92 billion in 2023-24. This combined amount is projected to increase to $22.85 billion by 2034-35.

Lack of Effect on ASX Investment

One possible hope for a big capital gains discount is that it might boost investment in Australian equities, and thereby indirectly contribute to the growth of Australian companies. As it happens, a look at the performance of the All Ordinaries index (i.e. the ASX 500) in the years after the introduction of the change gives no reason to think this happened. In fact, the years 2000-2002 were very bad for the Australian stock market, as the All Ordinaries fell from 3,152.50 at the end of 1999 to 2,975.50 at the end of 2002 (from Wiki). This is not a knock-down argument against any positive effects of the policy change on productive Australian investment, but we are not aware of any research to support this notion either.

Policy Alternatives and Reform Options

This section on reform options draws heavily from “Reforming Australia’s 50 per cent capital gains tax discount incrementally”, a paper by Brett Freudenberg and John Minas available here.

Perhaps the most obvious, if politically improbable, approach to reform would be to return to the pre-1999 status quo whereby the only tax-discounting on capital gains income was based on inflation.^[3] In other words, capital gains are taxed the same as wage income, modulo inflation adjustment for assets held for more than one year. This would be consistent with a simple principle: all forms of income should be taxed at the same rate. One possible fear of completely unwinding this discount is that it may contribute to deflating the Australian housing market, which is integral to the nation's wealth and therefore the economy in general.^[4] But in theory, at least, it is the simplest and most logical option. Freudenberg and Minas argue that it is it “can be considered the tax policy ideal” from the point of view of efficiency, horizontal equity and vertical equity.

Whether for reasons of political plausibility or because of economic concerns, more incremental options are well worth considering. The Freudenberg and Minas paper is well worth reading in this regard, as they thoroughly evaluate a number of these options. For those who don't have the patience to read the paper, the reform they ultimately favour is “A reduction in the rate of the CGT discount together with introduction of an annual CGT discount limit amount”. In particular, they tentatively suggest a reduction in the CGT discount from 50 to 25% and a discount limit above $50,000. The result of such a policy change would be a multi-billion dollar annual budgetary saving, and a huge boost to the progressivity of Australia's tax regime.

^{^}
[Note from author] Although I have found many detailed reports and papers on this topic, I have been unable to find a good quantitative, causal analysis of the effect of this policy on the Australian housing market. The kind of approach I think would be ideal is to simulate the counterfactual scenario of no CGT change using a model that predicts investor activity in the residential property market, as proxied by the ratio of investment loans to total new residential loans. The reason for choosing this metric rather than just house prices themselves is that it more directly matches the theoretical prediction that this policy change should have affected leveraged property speculation specifically. My proposal for predicting this investor activity index is to construct a sub-model of investor profit expectations in the residential market that can be weighed against profit expectations in other available investment vehicles to measure the generic favourability of property investment. This sub-model of investor profit expectations would take into account such factors as the interest rate, rental yield, recent growth trends and the CGT rate. If the overall model could be made predictive, then the effect of the policy could be assessed by modelling the counterfactual scenario where the profit expectations were lower because of the maintenance of a higher CGT rate. The guess is that investment activity would be lower in this scenario, which would have given first-home buyers a better chance for longer (even if there are other drivers for the dramatic house price growth we have seen this century).
^{^}
[Note from author] Forgive the inability to include the image here - need to push a code update to allow this feature.
^{^}
Freudenberg and Minas suggest that an ideal policy may even drop the inflation adjustment: "The justification for inflation adjustments for capital gains is tenuous, given that other forms of capital income, such as rental and interest income, receive no such inflation adjustment." However, they concede that it would make the proposal much more politically palatable "especially if there was a return to a high inflationary environment".
^{^}
[Note from author] I have not been able to find any economic modelling to test the validity of this fear.

Policy Questions