When the Howard government first introduced the capital gains tax discount in 1999, the policy was expected to broadly pay for itself over time. But Treasury officials have now told a Senate inquiry the way the policy’s cost is measured today makes the outcome look very different.
As a reminder, the capital gains tax (CGT) discount allows individuals and trusts to pay tax on only 50% of a capital gain when an asset has been held for more than 12 months before being sold.
We’re set to hear from a Senate committee next week how it’s likely to change.
How it was meant to be ‘revenue neutral’
Treasury officials told a Senate inquiry last week that when the CGT discount was introduced, the policy modelling suggested it would be “broadly revenue neutral over time.”
In evidence to the committee, Treasury explained that the original modelling involved two stages.
First, officials calculated the static cost of replacing indexation with the 50% discount. On that basis alone, the change appeared to reduce tax revenue. But the modelling also included behavioural responses — changes in how investors would act once the tax settings changed.
Treasury officials told senators the modelling assumed investors would be more likely to sell assets and realise gains if the tax burden on those gains was reduced. It was designed to essentially eliminate the so-called “lock-in effect” which reduces the turnover of assets like homes, officials added.
That behavioural response of greater asset turnover was expected to increase the number of taxable capital gains events. More people selling, more people having to pay (albeit less) capital gains tax, therefore offsetting the initial loss in revenue.
Where it went wrong, and how much it costs taxpayers today
Today, the capital gains tax discount appears in the federal budget as a tax expenditure worth more than $20 billion a year.
But Treasury officials told the Senate inquiry that this number is calculated using a different benchmark from the one used when the policy was introduced.
The tax expenditure estimate compares the current system against a benchmark where capital gains are fully taxed on the nominal gain, with no discount.
Before 1999, however, Australia did not tax capital gains that way. Instead, gains were indexed for inflation, meaning taxpayers were taxed only on the real gain after inflation.
Treasury officials told the committee that this difference in benchmarks means the modern tax expenditure estimate cannot be directly compared with the original modelling for the reform.
In their testimony, Treasury officials stressed that the tax expenditure estimate reflects the difference between full nominal taxation and the current discounted system, rather than the difference between the old indexation system and the discount.
Because of that benchmark, the estimate captures the full value of the concession relative to a system that Australia has never actually operated.
Treasury told senators that understanding this distinction is important when interpreting the headline figure attached to the CGT discount in the budget.
The committee is expected to hand down its report into the capital gains tax discount in the coming days.
Maybe going back to indexation is the way to go. If you bought as investment for $100,000 and sold it 10 years later for $200,000, but inflation had doubled everything, you in effect made nothing. That was fair. But the worst scenario when things changed was what if you sold at a loss? Suppose your investment had halved to $50,000, under the indexation you could claim an effective loss of $150,000. UNder the new scheme, you had to halve the loss to $25,000. Not only did you not get the indexation, but could only claim half your loss. That is a huge difference.