Are we too negative about negative gearing?

As someone who doesn’t own a house, I don’t think about housing very much. But I have been seeing a lot of recent articles saying something needs to be done about negative gearing. When looking for analysis though, the most comprehensive piece was this 2016 note by Grattan.

However, I’m both a secret tax nerd and always need to spend some time thinking about things. I need to explicitly work through the “why” of things before I know whether I agree or disagree - namely “why do we have negative gearing in Australia”, “why may it be seen as appropriate or inappropriate”, and “what evidence do we have”. Ok the last isn’t a why, but is pretty important too.

For fun, I thought I would walk through my thinking below, keen for thoughts from others - and happy to be persuaded.

The Grattan piece is definitely the best piece I’ve found about negative gearing in Australia - I enjoyed this piece a lot and didn’t really find much else willing to talk trade-offs - so I will be using the Grattan piece to aid the discussion. However, I might not reach the same conclusions. For those willing to look outside Australia, the NZ IRD advice on disallowing the deductibility of interest of rental properties is pretty relevant.

tl;dr Matt’s conclusions there are trade-offs from limiting negative gearing - but any claim that it will “improve housing affordability” through reduced rents is political posturing rather than reality. This is not a distortionary subsidy in of itself. Instead the key trade-off of limiting negative gearing is between the tax planning potential of negative gearing (benefit) vs the of risk of higher rental costs/lower housing supply.

If the concern is truly housing affordability, maybe push for something actually relevant - like densification and the removal of dumb heritage listing for things like car parks and oil tanks.

Note: I’m a newbie in terms of Aussie tax - so any feel free to share any comprehensive or important pieces post Henry tax review as I just won’t know about them!

---

Why can housing investors negatively gear!

Let’s cover the knee-jerk feeling on this first - it definitely feels like bullshiz that someone renting out a property can claim a bunch of expenses against their tax bill, while renters can’t claim rent and home owners can’t claim the cost of their house.

This initial feeling around fairness is what pushes people against any type of negative gearing - in New Zealand this thinking, and its corresponding support from the public economics community, lead to it going even further with a residential property only form of ringfencing, earlier removal of depreciation expenses, and interest expenses no longer being allowed to be claimed! Note: New Zealand also doesn’t have a long-term capital gains tax.

But the policy makers of the past were (probably) not silly or corrupt - so what is the reason why we i) allow expensing of rental properties but not other housing expenses ii) allow individuals to claim these losses against other income as a form of “total taxable income” rather than ringfencing these losses?

Why can we only expense on investment properties?

Expensing is part of an income tax system - where the goal is to raise government revenue from the genuine income someone earns. For an investment property - as in any type of business - the income earned is the revenue (i.e. rents received) net of expenses. As a result, we deduct such expenses from this estimate of income.

Now what about owner occupied property! Well the income earned from this property (imputed rent) is not taxed at all - as a result, if the income is untaxed it is a bit weird to allow a deduction. Hence no deductions - this would change if we started to tax imputed rent.

What about rent paid! Well rent paid isn’t income - it is payment for a service. The income is taxed for the person you are paying rent too, and it is at that level that deductions are allowed.

Why comprehensive income?

Cool, so we’ve motivated why we allow deductions - but why can people pool together all of their sources of income and expenses in one. Shouldn’t the deduction only apply to rental income?

That is the essence of the debate here - ringfencing of income and expenses by income source.

In Australia we have a comprehensive income tax - meaning that the “goal” is to treat a dollar as a dollar irrespective of the source. Now this is often violated (capital gains discount, super subsidies), but the principle is that by allowing individuals to treat income and expenses consistently across all of their economic decisions we minimise distortions about what to do - and also generate horizontal equity by treating all taxpayer who earn the same income the same way.

In this world, the decision to ringfence will either do nothing or increase tax rates on investors relative to comprehensive income treatment. So what are the channels where higher effective tax paid due to ringfencing may influence outcomes:

• By reducing the incentive to hold a rental property, the short term price of owner occupier properties will fall.

• By increasing the relative tax rate on property people will rebalance investment portfolios to other assets.

• By reducing the incentive to supply rental property rents will rise.

• By reducing the return for adding housing supply housing affordability will worsen.

• By limiting the ability to “shift income” through time, it will reduce tax planning activity.

We’ll work through these ideas a little bit more below - but given how scant the evidence currently is, it is very hard to be too definitive on the magnitude of things!

The suggested alternative

Lets look at the specific suggestion out there - not allowing rental property deductions to be claimed against wage and salary earnings, but allowing deductions to be “carried forward” if investment income is below these expenses. This is a much narrower change than the ones that have occurred in New Zealand.

I will leave the changes to the capital gains discount and tax favoured status of superannuation savings - as they are separate topics that are worthy of further discussion in of themselves.

The 2016 Grattan report that motivates this change claims both:

1. There were be significant fiscal savings from the policy change.

2. There would be no to limited implications for rents.

Although I agree the rental argument can be overplayed (although in parts I feel the report does underplay it), that is because the net present value of the fiscal benefits are significantly smaller than suggested in the report.

The proposal is intended to only change the timing of when someone can claim expenses - namely they have to claim the benefit of the deduction at a later date, when they earn positive income from the asset.

In this way, it increases government revenue now and reduces it in the future. In so far as an individual could put the money they’ve paid in tax in the bank to earn some risk free income, their tax liability is higher in “present value” terms and instead the government benefits from that. But it seems a stretch to say the net medium-term value of this is 80% of the short-term gross increase tax revenue (which is what the report estimates).

Such a large medium term fiscal gain would need to be due to some individuals running a lifetime loss on their property - and so never using these deductions.

However, then the argument that this wouldn’t increase rents stretches credibility - it is hard to argue that these property investors who are operating at a loss are earning “economic rents” that we can tax without them either increasing rents or selling to individuals who will treat this more as a business.

And if they instead sell to owner-occupiers, that adjustment would initially lower prices for those who are able to transition from renting to owning - but in terms of an affordability crisis this is not the most vulnerable group … who instead will be the individuals who remain renting and face higher rent.

But we need to equalise ETRs!

Ahhh effective tax rates. Definitely agree that equalising them is a good first principle - but to do this we need to calculate them properly.

I’ve always had a bit of an issue with ETR graphs comparing equity and debt finance that only looks for the single investors perspective - as that debt is someone elses savings which are already facing income tax. Let me explain.

Say that you have can invest for $100 and you make a $10 return per year. Nice. If you use some equity to make that investment you pay tax on the whole $10. But if you keep hold of your $100 and borrow $100 to fund the investment at a 5% interest rate your “profit” is only $10-$5, so you are only taxed on that. Hey presto your tax payment is lower as your taxable income is lower!

However, we also need to recognise there is someone lending the $100 to earn that $5, and that person is paying income tax on that. We can imagine “looking through” the person lending, and treating them as someone purchasing a share of the asset and receiving this income stream as their return on it - which is then taxed. Recognising this the full $10 is taxed when we allow expensing, and if we didn’t allow the deduction we would be taxing the $5 component twice!

Or another way to frame this is that the required rate of return for lending will be higher due to income tax, and that is really part of the ETR associated with debt financing investment - which these examples miss!

Allowing legitimate deductions isn’t a “subsidy” for investor for any business - including a rental business. Furthermore, disallowing deductions when the income is taxed does not “level the playing field” against individuals whose income stream from purchasing a property is not taxed at all!

If the issue is that rental income is not being taxed properly (say if the capital gain discount is too high), then sorting that out is the solution - not limiting expensing.

But this is passive income!

The passive-active distinction isn’t really economics - it comes from legalistic definitions of income as income definitions were debated in legislation and the courts through the years. So for this debate I don’t find this too useful.

But even if we did buy it - the most convincing argument I’ve heard is that being able to “borrow” passive funds to “earn from” passive funds allows tax planning and abuse without the creation of economic value - I am not sure the idea of supplying housing services is passive.

Matching a renter with the house, maintaining the property, taking on the risks associated with the property - these are all standard business activities the involve the provision of a service with clear economic value.

But these tend to be high income individuals!

Ok, if you think that the tax scale should be more progressive make it more progressive - what does this have to do with denying a legitimate business expense.

But this brings us to the best argument to my mind here is that it could be used as a vehicle to shift taxable income through time and lower someones lifetime tax liability! This is easily the main convincing argument for me, and all the yarns about major fiscal benefits without rents changing are a side story.

So if someone has the numbers on that definitely hit me up!

But if this is the case we should be thinking similarly about trusts. As this work from ANU’s TTPI highlights the trust operated by your local GP, or the mining magnate, is where a lot of this tax planning behaviour is taking place!

Now looking at residential property investors, the competing arguments about limiting deductions in this case are as follows:

1. By allowing tax planning, two individuals in the same income position are being taxed differently - violating horizontal equity. Limiting deductions (while allowing carry forwards) solves this.

2. By disallowing this type of planning an “accidental subsidy” on low income rentals will be removed - which may lead to higher rents or restrictions to housing supply.

If planning isn’t occurring, both these arguments evaporate - and this is just a weird policy that will force landlords to spend more time with accountants (compliance costs).

If planning is consequential, the suggested change by Grattan has merit - but we would need to be mindful that this either doesn’t change or worsens affordability for vulnerable individuals by increasing rents - and complementary more “pro-affordability” measures should be considered. Like zoning for density, or providing income support to vulnerable individuals.