Negative gearing helps poorer people most

Should we get rid of tax provisions that provide a greater benefit to the rich than the poor? That’s the argument we hear from those who want negative gearing abolished: that the benefit is much larger at high incomes.

If you believe this then you should line up: there are plenty of other tax concessions that we should get rid of.

In particular, let’s get rid of all the GST exemptions, as they all provide greater value to high income earners.

For example, the GST exemption for food provides a benefit of $365 to the poorest 10% of households, but $631 per year to the richest 10% of households. Therefore logic says we must get rid of it.

It is much starker for health: a benefit of only $171 per year to the poorest households, and a much larger $488 to the richest. Definitely should be abolished.

The same goes for education, water and childcare. Cancel, abolish, remove!

In total, the benefit of all the GST concessions is $679 per year for the lowest 10% of households, while the benefit to the richest households is $1,643 per year. Get rid of them all!

The figures are shown in the graph below.

The trend is clear: the rich obtain a larger benefit from these GST exemptions.

So why do people oppose the removal of these exemptions — often the same people supporting the abolition of negative gearing?

Because the relative benefit, or the benefit as a percentage of income, for these GST exemptions is larger at low incomes than high incomes. This means that these GST exemptions are progressive, and removing them would be regressive (although the impact of such a change could be offset by other tax changes).

This is the standard approach to looking at the income distribution of tax provisions. All commentators uses this approach (except when they are going after negative gearing).

The correct proportionate analysis of the GST exemptions is shown in the graph below.

So how does this relate to negative gearing? It means we should measure the tax benefit of negative gearing as a proportion of income, rather than in dollar terms.

And when we do that, guess what? The largest relative benefit is at low income levels rather than high income levels. In fact the benefit is particularly substantial at low income levels. This is shown in the graph below.

So abolishing negative gearing would be regressive — or at the very least not particularly progressive, if the lowest decile is left out. Why omit this decile? It has been argued that the figures for this decile are problematic because low income earners couldn’t borrow to invest, so it is more likely that they are avoiding tax.

However, a more likely option is that this is made up of non-working spouses who negatively gear while supported by their working partner. These households are actually paying more tax than if the working spouse was negatively gearing, because the working partner would have a higher tax rate. Other investors may be making one-off losses and earning more in other years. No clear tax avoidance with these options.

As a result, people who appear to be low income negative gearers may be in higher income households, or have higher income over the longer term. However, we haven’t seen the negative gearing abolitionists discuss this issue or present data on it.

The other argument sometimes made is that we should measure the benefit of negative gearing as a proportion of taxable income before negative gearing. However, this would differ from all other distributional analysis: for example, the GST analysis discussed above doesn’t compare the concessions with household income before the value of the concessions. Consistency argues for the same treatment of negative gearing. Commentators arguing for inconsistent analysis of negative gearing are just assuming the conclusion they want to prove.

There is further inconsistency in the arguments of some commentators who argue we shouldn’t look at the absolute number of nurses who negatively gear, but instead we should look at the proportion of nurses who negatively gear. And, unselfconsciously, they then measure the distribution of the benefits of negative gearing in absolute dollars, rather than as a proportion of income. Go figure.

There is a broader issue: how have we come to the point that the assessment of tax changes is largely based on who benefits or loses from the policy? While this should be a relevant issue, it should not be the deciding factor.

The relative benefits of negative gearing are greatest at low income levels; but this shouldn’t determine what is done. Much more important are the impact on the budget, the housing market and economy more broadly. This is where the debate should be held, not in facile and misleading discussions about the distribution of benefits and losses.

Michael Potter is a Research Fellow in the Economics Program at the Centre for Independent Studies