Over the past seven years, New Zealand has experienced one of the fastest rates of house price inflation in the developed world.
According to the Bank for International Settlements (the central banks' bank) residential prices in New Zealand ended last year 71 per cent higher than their level in 2010, a compound annual growth rate of 8 per cent. That compares with a cumulative rise of 43 per cent in Australia and an average 27 per cent across all advanced economies.
Among advanced economies, only Hong Kong and Iceland saw faster rises in residential prices. Indeed, after adjusting for consumer price inflation only Hong Kong outstripped New Zealand's 54 per cent real rise.
As a general rule, when we are an outlier in some economic measure it is not a good sign.
Clearly, multiple factors are at play in the latest surge of house price inflation we have suffered or enjoyed (depending on your circumstances).
But one of them, undoubtedly, is tax. We have a tax system which discourages saving and encourages borrowing to buy housing, whether to live in or to rent out.
The tax system is being reviewed, but its terms of reference, for fairly transparent political reasons, put owner-occupied housing off limits.
That leaves landlords in the gun.
Officials' estimates of the relative marginal effective tax rates on the real returns from various types of investment show rental property is handsomely tax-advantaged and owner-occupied housing even more so.
We have a tax system which discourages saving and encourages borrowing to buy housing, whether to live in or to rent out.
The New Zealand Property Investors Federation disputes this. These arguments get very arcane. But they tend to ignore the crucial influence of leverage. An investment largely bought with borrowed money in a rising market will see the buyer's equity grow much more rapidly than the asset value. If the capital gain is untaxed, so much the better.
Conversely, leverage will amplify the investor's losses in a falling market.
It is the interaction between the tax system and the banking system which tips the scales in favour of housing over other investments.
Good luck, for example, going to a bank and saying: "I reckon Fletcher Building is oversold. I want to buy 100,000 of their shares and I would like the bank to put up 80 per cent of the money, please. I'll happily give you a mortgage over the shares."
The response is liable to be: "Never mind that. What about a mortgage over your house? You can do what you like with the money."
A simple way of looking at the issue would be to compare the tax paid by landlords with the value of the rental housing stock.
The Property Investors Federation, citing Inland Revenue data, says landlords paid $1.3-$1.5 billion a year in tax over the five years to 2016. Assuming, conservatively, that they own a quarter of the housing stock by value, and taking the Reserve Bank's estimates of the value of the national housing stock over that period, that would represent a tax yield of about 1 per cent a year.
Not exactly onerous, especially in times of rampant house price inflation.
Westpac's chief economist Dominick Stephens has used a model he developed some years ago to calculate the impact various tax changes would have on house prices and rents.
The model estimates the value of an average New Zealand house to a rational investor, based on the net present value of the rent that the property could generate, less the expenses. It is sensitive to inflation, interest rates and the tax system, and gives excellent signals as to what might happen to house prices when one of these factors changes, Stephens says.
For example, from 2014 to 2017 mortgage rates fell sharply and the model correctly predicted that this was a period of strong house price growth.
The reason for looking at what it is rational for investors to pay is that they tend to be the marginal buyers who set the prices would-be owner-occupiers have to outbid.
Westpac's model predicts that a capital gains tax at a rate of 10 per cent, which would exempt the family home, would reduce house prices by 10.9 per cent, all else being equal, but push up rents 5.5 per cent. It does not predict how long it would take for those effects to work through.
A capital gains tax is not the only option. A property tax of 0.5 per cent on the value of the property (house and land), again exempting the family home, would have a similar impact. By reducing the price an investor would be willing to pay, it would increase home ownership rates, reduce house prices by 10.5 per cent and put rents up 5.2 per cent.
A land tax of 1 per cent on the unimproved value of the land on which a dwelling sits, again exempting the family home as per the terms of reference, would reduce house prices, all else being equal, by 9.5 per cent, the model predicts, but push up rents by 4.8 per cent.
Properties for which land makes up a greater proportion of the value, such as houses with large sections, would experience a greater percentage decline in price, while apartments would experience a smaller percentage decline, Stephens says. "Auckland prices would probably fall further than prices elsewhere in New Zealand. This is because land makes up a greater proportion of the value of Auckland properties than in other regions."
The model assumes that one-third of the adjustment to a tax change would come through higher rents and two-thirds through lower house prices.
That is the assumption Stephens has least confidence in, he says, in that it is based on observations when the model was developed in 2007. But if rents rose less, house prices would fall more.