Broadening the tax base is critical in the longer term, to avoid large increases in personal income taxes and GST.
That is, in so many words, the view of the tax working group examining how to overhaul the tax system so that it is more efficient, fairer and, crucially, able to continue to deliver the revenue needed to fund what we have come to expect from the state.
Any substantial increase in the tax base is bound to be unpopular, and present difficult transitional and boundary issues.
An invertebrate Government would have no shortage of reasons to shrink from such changes.
But it set up the working group for good reasons. The status quo is far too reliant on taxing personal and corporate income in a world where labour and capital are mobile and New Zealand is not well placed to compete for either.
The group's most recent meeting focused on ways of bringing property into the tax net, particularly a capital gains tax, a land tax, and taxing a deemed return on rental properties.
It received conflicting advice from officials on the merits of a capital gains tax.
The Treasury likes it. It reckons it would bring in at least $1.5 billion a year even if owner-occupied housing were excluded, which would nearly cover the cost of bringing the top personal tax rate down to 30 per cent.
It would also make the tax system more progressive, hitting hardest those best placed to bear it.
Inland Revenue officials on the other hand were more focused on the practical difficulties.
For example, applied to the family home it would have lock-in effects and present a barrier to labour mobility. Exempting the family home would have created distortions in investment and land use decisions. Neither option is efficient.
On balance, IRD argued, in the real world the disadvantages of a capital gains tax would outweigh the advantages.
A land tax by contrast is efficient in that the tax base is immobile (it cannot take the next plane to Perth or London) and the supply of it is fixed.
As the supply of land is inelastic (unaffected by prices), market prices depend on what buyers are prepared to pay rather than on landowners' expenses.
Even though the tax would be paid at regular intervals, probably annually, it really represents a one-off tax on the current owners of land, officials say.
The introduction of land tax would trigger a drop in the value of land, by the net present value of the future tax liability. It would not affect those who acquire land after its introduction, whose increased outgoings would be exactly compensated for by a lower purchase price.
The potential tax base is huge - getting on for half a trillion dollars, of which about a fifth is farm land. That allows low rates to deliver a worthwhile increase in revenue and therefore scope for tax relief elsewhere in a revenue-neutral package.
Estimates of how much land values would fall depend on the rate of the tax and other assumptions used in the modelling.
Work by economists Andrew Coleman and Arthur Grimes at the Wellington think tank Motu, which assumed a 1 per cent tax rate, concluded values would drop by a sixth. The IRD, using some different assumptions, including the ability of businesses to deduct land tax payments from their taxable income, but keeping a 1 per cent rate, concludes the fall could be more than a quarter.
Clearly this raises concerns about highly-geared households, even though land is typically only half the value of residential property, diluting the impact, and suggests that the rate would need to be set well below 1 per cent.
The banks, we may assume, would be unimpressed to see the value of the security underpinning so much of their loan books reduced at a stroke.
But it is worth remembering the sheer size of the increase in house prices during the boom. In round numbers the value of the housing stock doubled from $300 billion in 2003 to $600 billion in 2007 and has retreated only modestly, about 8 per cent, since then.
Residential property represents about two-thirds of the value of privately held land and is split roughly two to one between owner-occupied and investment properties.
Would a tax-induced price shock roughly comparable in scale to what has occurred over the past couple of years, or which reversed one year's house price inflation during the boom, be the end of the world?
A land tax would be disproportionately hard on superannuitants, however. In general they would have to devote a higher share of their income to paying a land tax than other property-owners on similar incomes.
The obvious solution would be to increase New Zealand Superannuation payments. That would be a better response than whittling away the integrity of the tax through carve-outs or differential rates - precisely the defects which led to the abandonment of the previous land tax.
The working group agreed it is critical that if a land tax were introduced, the same rate must be applied across all land types.
Another base-broadening option it considered is a tax targeted at rental properties, applying a version of the 2001 McLeod review's risk-free return method (RFRM).
The working group's minutes described the rental property sector as "the glaring hole in the tax system".
In aggregate the $200 billion or so invested in rental properties yield not one red cent to the public purse. It is a net tax shelter. What is wrong with that picture?
The working group said that if a capital gains tax is not the preferred option, RFRM on rental properties is an option worth considering.
Under this system rents would not be taxed and expenses relating to the investment would not be deductible.
Instead the tax base is the amount that would have been earned if the investor's equity in the property had been invested in a risk-free Government bond, excluding that part of the return on the bond which is just compensation for inflation.
So if the bond yield is 6 per cent and inflation 2 per cent, then the investor's equity is deemed to return 4 per cent and that is taxed at the investor's marginal income tax rate.
The difficulties are that it could raise rents and would therefore be regressive, in contrast to a capital gains tax which is progressive or a land tax which is more or less proportional.
Under the RFRM people are taxed even if they make losses.
And applying it only to investment properties only increases the incentive to invest in owner-occupied housing - already high because home owners escape tax on both their imputed rentals and capital gains.