When the Government crafts a package of tax reform its options will be limited by two big constraints.

One is fiscal: the self-imposed need for the package to be broadly revenue-neutral overall, which inevitably means it is an exercise in robbing Peter to pay Paul

The other is the need for it to be seen as reasonably fair and equitable.

That is essential, Finance Minister Bill English says, if tax changes are to stick, and only changes that stick would be worth the political price of robbing Peter.

The tax working group's report talks of both vertical and horizontal equity.

Vertical equity is about how the tax burden is distributed among different levels of income. The longstanding principle is those on higher incomes should pay a higher proportion of their incomes in tax.

Horizontal equity is about whether those with similar incomes, or in similar financial circumstances, pay a similar amount of tax.

Vertical equity considerations are likely to count against, for example, funding income tax cuts at the top end by raising GST, which would have a disproportionate impact on people on low incomes.

Horizontal equity considerations suggest cuts to the top personal rates and company rate are more likely to be funded from the property sector, in whose favour the current system, the working group concludes, is biased.

That is on the assumption that landlords are likely to be, or have been, in the upper tax brackets.

The tax working group makes a theoretical case for the Government to get more of its revenue from GST, than the 21 per cent it does now. Consumption taxes are seen as less damaging to economic growth than taxes on personal and corporate incomes.

New Zealand's version is still regarded as a model of its kind, particularly for its comprehensiveness, and at the margin it might tip more disposable income into saving than spending.

The problem is that GST represents a bigger share of income and expenditure for those on lower incomes than those on higher ones.

Prime Minister John Key has said if GST rose, people on lower incomes would need compensating.

This is not just a matter of raising benefit and superannuation payments. Compensating those in the workforce is much more difficult and expensive.

One of the scenarios modelled for the tax working group has GST rising from 12.5 to 15 per cent and offsetting that by cuts to the two lower income tax brackets (10.5 per cent instead of 12.5 per cent up to $14,000 and 19 per cent instead of 21 per cent of up $48,000).

Those income tax cuts would cost around $2 billion a year, while the GST increase would bring in an extra $2.15 billion, or $1.9 billion allowing for automatic inflation-indexed increases in benefit and superannuation payments. On average, for people in the lower income deciles it would be a wash.

The catch is the "on average". This does not to take into account the complex interactions between the tax systems and social assistance programmes such as Working for Families.

"In effect Working for Families means that households [with children] in the bottom half of the income distribution effectively pay no income tax or receive tax credits," the working group said. That makes it difficult, to say the least, to fully compensate people for higher GST by cutting income tax rates.

Opinions differ as to how progressive the income tax scales should be.

Is it fair for 44 per cent of the income tax take to come from 10 per cent of taxpayers?

And is it sustainable in an age of labour mobility, given the income gap with Australia and most other developed countries?

The working group says it is not sustainable, especially with inflation over time pushing more and more people into the top tax bracket, and calls for rate cuts at the upper end of the scale. Much of its report is about exploring how that could be funded.

Labour's finance spokesman, David Cunliffe, is sceptical that the Government would achieve a full offset for an increase in GST, so that any change in that direction would be unacceptably regressive.

But on the horizontal equity front, Cunliffe sees a strong alignment of views between the major parties, at least in principle.

"However the Government knows walking the talk there means losing a few friends in the property sector."

The tax working group strongly advocates aligning tax rates to avoid the distortions arising from the fact that different rates of tax apply depending on whether assets are held directly or through companies, trusts, or portfolio investment entities (PIEs).

"The tax system also biases investment in favour of rental property investment."

Although aggregate rental income has grown strongly over the past 15 years or so, the deductions available to property investors have meant that tax losses in the sector have grown even more.

Some economists argue the usefulness of rental properties as a tax shelter has been a major driver of high house-price inflation (property values roughly doubled between 2002 and 2007), pushing home ownership out of reach of more and more people.

Advocates of more vigorous capital markets see the tax advantages of rental property as channelling too much of New Zealanders' savings into an relatively sterile asset class from the standpoint of economic growth.

And from the Government's point of view the net tax yield from a sector with some $200 billion invested in it has been negative in recent years.

But the working group report makes it clear there are also drawbacks to the various options for taxing property.

A capital gains tax was opposed by most members. Practical concerns include the risk of locking people into assets they would otherwise sell in order to defer paying the tax.

A land tax, by contrast, was supported by most of the working group's members as a means of funding tax cuts elsewhere.

The potential tax base is very large, hundreds of billions of dollars, allowing a low rate to be applied, such as 0.5 per cent, and still bring in more than $2 billion a year.

It would immediately cut the value of existing land holdings by reducing how much, all else being equal, future buyers would be prepared to pay for it. Views differ on how big that effect would be.

After that one-off impact, a land tax would not reduce the relative attractiveness of rental property as an investment.

It would be hard on those who are asset-rich but cash poor, such as the retired and Maori incorporations, and there would be pressure to exempt farm land.

It would also compound distortions already embedded in local body rates.

An earlier attempt, by the 2001 McLeod tax review, to trade off a wealth tax that includes the family home for deep cuts in income tax rates, met ferocious opposition.

A land tax would have a narrower base and a lower rate but would probably get a similar reaction, especially from middle-income home-owners.

A third option, which could be targeted at rental property, would be to apply a version of McLeod's risk-free return method (RFRM).

Instead of taxing a landlord's rental income and allowing a deduction for expenses including interest and depreciation, what would be taxed is a deemed return on the equity in the property.

It would be deemed to have earned the same income that sum would have earned if it had invested in a risk-free asset such as Government bonds and this what would be taxed.

The working group estimates this could bring in around $850 million a year, and more if holiday homes rented out for some of the year were included. But for so novel a tax such estimates need to be treated with caution.

Its critics point to the fact that property investors would have to pay the tax even in years when property values fell, and that if they are already cashflow-negative, the tax would compound that.

As with a capital gains tax on rental property, how it scored for vertical equity would depend on how far landlords could pass on increased liabilities to tenants.

The fourth option for lifting the tax take from property would be to deny a depreciation deduction for buildings. The working group questions having such a deduction for buildings which in fact increase in value.

Removing the depreciation on buildings would save revenue of up to $1.3 billion, it says, and it is the kind of tax change which could be implemented quickly.

It would take the Government a long way towards the estimated $1.6 billion cost of aligning the top personal, company, trust and PIE rates at 30 per cent.

But $1.3 billion is an rough upper boundary for the potential yield from scrapping building depreciation.

Deloittes, for instance, reckons that when due allowance is made for buildings such as industrial premises which do in fact depreciate, the yield might be only half that.

So while the tax review has given the Government plenty of options to consider, it hasn't given it easy ones.

PROS AND CONS
Revenue-raising options:
RAISING GST

Pro: GST is a robust and efficient tax, and shifting tax from incomes to spending might improve saving.

Con: It is very hard to prevent a rise in GST hitting those on lower incomes harder.

CAPITAL GAINS TAX

Pro: Potentially very lucrative, allowing more tax relief elsewhere.

Con: Lots of practical difficulties and the IRD, which would have to administer it, hates the idea.

LAND TAX

Pro: Broad base, low rate and could bring in billions.

Con: Liable to be undermined by exemptions as in the past. Hard on the retired, Maori trusts and farmers.

RISK FREE RATE OF RETURN FORMULA APPLIED TO RENTAL PROPERTIES

Pro: Targeted at a sector that seems undertaxed now.

Con: Because it is based on equity, it could perversely encourage more gearing in the rental property sector. Could flow through to tenants.

SCRAP BUILDING DEPRECIATION

Pro: Could be done quickly.

Con: It is not easy to distinguish buildings which do depreciate from those which don't.