As the Tax Working Group starts work on its final report, the riding instructions it has received from the Government are demanding.
Finance Minister Grant Robertson wants a package of measures that would (among other things) be revenue-neutral and would reduce inequality.
This presents Sir Michael Cullen and his colleagues with a number of problems.
The first is: revenue neutral over what period? If the main expansion of the tax base in the offing is to tax a much wider range of capital gains than is currently the case, and it is not to be odiously retrospective, it will take years for the revenue yield to build up.
The second is that while it is true that the tax and transfer systems between them do less than in most developed countries to narrow the gap between inequality of market incomes and inequality of disposable incomes, it is transfer payments, not the tax system, that do the heavy lifting.
• Brian Fallow: Tax review redefines "income"
• Brian Fallow: Company tax rate - leave it where it is, Cullen review told
• Brian Fallow: Who said tax was meant to be fair?
Officials' advice on possible changes to income tax through lowering rates or raising thresholds argue that "if the aim of the reform is to target the lowest income households it would be sensible to be recycling any revenue raised through extending the taxation of capital income into increasing transfers".
It is difficult and expensive to make income tax more progressive if you rule out a higher rate at the top end — and the review's terms of reference forbid that.
Of the seven possible changes officials consider, the two which have the biggest impact on inequality (at least on a static basis) would be to halve the tax paid on the first $14,000 of income, from 10.5 per cent now to 5.25 per cent, or to introduce a tax-free band for the first $7000.
Both would be worth $730 a year to people in the second lowest income decile and twice that to those in the top decile. But it would still only nibble at inequality.
Another option, reducing the second rate — which applies to the $14,000-$48,000 income band — from 17.5 to 14.75 per cent, would be worth about $10 a week to a full-time worker on the average wage, but of course do nothing for part-time workers or beneficiaries earning less than $14,000.
One way of improving the progressivity of the income tax system would be to deal with the issue of bracket creep or fiscal drag.
The takeaway from all this is that when nearly half of all households are net recipients of money from the state (receiving more in transfer payments than they pay in income tax) it has to be welfare reform — the remit of a different advisory group — that is crucial for reducing income inequality. It would have to deal in particular with the abatement regimes which impose punishingly high effective marginal tax rates on wages earned.
That said, one way of improving the progressivity of the income tax system would be to deal with the issue of bracket creep or fiscal drag, which pushes more and more taxpayers into higher tax brackets.
Since the thresholds and rates were last adjusted in 2010, the effect of this stealth tax increase has been to reduce the proportion of individual taxpayers in the lowest tax bracket from 27 to 22 per cent and those in the second lowest bracket from 50 to 46 per cent, while increasing the proportion in the second highest bracket from 13 to 16 per cent and those in the top bracket from 10 to 17 per cent.
If thresholds had been adjusted annually for inflation they would be 8.8 per cent higher by now, raising the $14,000 threshold to $15,300, the $48,000 threshold to $52,200 and the top threshold from $70,000 to $72,200. If those changes were introduced for the next tax year, the fiscal cost would be around $1.7 billion, officials estimate.
Another factor limiting the working group's options is that the virtual absence of capital gains tax is not the only respect in which New Zealand's tax base is conspicuously narrower than international peers'.
We also don't do social security taxes.
In particular, New Zealand superannuation is a charge on the entire tax base and not funded by a payroll tax that widens the tax wedge on wages — the gap between what it costs to employ someone and their take-home pay — which is low by developed country standards.
Economist Andrew Coleman calls it extraordinary that the tax review has virtually ignored this feature of the status quo, because it is the single largest difference between New Zealand's tax system and that of most other developed countries: "In 2016 New Zealand raised approximately 1.1 per cent of GDP in social security taxes (the ACC levy). The OECD average was 9 per cent."
The virtual absence of payroll taxes could be seen as a positive feature of the tax system, contributing to a high employment rate by international standards. Certainly, the lamentations we are hearing from business lobby groups about increases in the minimum wage suggest widening the tax wedge on wages would not go down well.
But Coleman points out a key feature of social security taxes is that they apply to labour incomes, not capital incomes.
The absence of them makes labour appear more heavily taxed by international standards.
OECD data on the tax rates for someone on the average wage ranks New Zealand the 10th highest among 34 developed countries if social security taxes are ignored, but second lowest when social security contributions are counted.
Because so little revenue is raised by social security taxes in New Zealand, the amount of revenue raised from income taxes is very large by OECD standards.
This means that New Zealand has relatively high effective rates on capital income — or at least on capital income that comes from assets other than property — and that contributes, Coleman argues, to the capital-shallowness of New Zealand business.
He questions the assumption that capital income and labour income should be taxed at the same rates.
That is anomalous in the OECD, unsupported by theoretical considerations and may be contributing to New Zealand's low capital intensity and poor economic productivity, he says.