Michael Cullen has been adamant about the principles that will guide tax cut policy under the Labour government.   But his criteria are not nearly as binding as they might initially sound – a point that was forcefully rammed home by the prime minister’s assurance that the criteria would indeed be satisfied and tax cuts will be forthcoming.

To discuss the first two criteria in tandem, Cullen has suggested that the government will neither borrow nor cut services to pay for tax cuts.   But in economic terms, the cost of something is always the opportunity cost of the best foregone alternative.   By definition, the cost of tax cuts is foregone services or foregone debt repayment (unless the finance minister had undergone a conversion to supply-side economics, and now believes that tax cuts increase revenues).

So what do the first two criteria really tell us? Presumably, no extra borrowing is just a restatement of the government’s long-standing goal of holding the debt-to-GDP ratio constant.   As I argued in an earlier column (15/12/07), this means running cash deficits of around $2bn per year.   Starting from the current cash surplus, this criterion allows the government an extra $5bn of spending or tax cuts in coming years (over and above the amount allowed by the normal growth in tax revenues).

"No cuts to services" is an equally obscure criterion.   Government spending expands along with the economy every year, so services are hardly ever cut in a nominal sense.   And budget projections already allow for spending increases to maintain expenditure in real (after inflation) terms.   So even if the government used all of its $2bn annual budget allowance for new spending to fund tax cuts, services still wouldn’t be "cut" in real terms.

So under the "strict" guidelines of the first two criteria, the government is allowed a tax cut programme of no more than $5bn total over the next three years.   By point of comparison, the income tax regime proposed by the National party at the 2005 election would cost $2.6bn.

It is on the third criterion – that tax cuts will not exacerbate inflationary pressures – that Cullen begins to find himself in serious logical strife.   There are only two ways that tax cuts won’t increase the fiscal stimulus and add to inflationary pressures.   The first is if tax cuts are offset by reducing spending (which Cullen’s second criterion ruled out).   The second is if tax cuts are funded from upside surprises in tax revenues (stemming, for example, from the dairy boom), in which case the "tax cuts" aren’t tax cuts at all in net terms.

So what does the third criteria mean in reality?   The government’s latest fiscal projections already feature a substantial fiscal stimulus, and the Reserve Bank’s monetary policy decisions already incorporates the impact of that expected stimulus.   Cullen’s third criterion is presumably just a promise not to make a bad situation worse, and does not imply any further restriction on the government’s options than the first two criteria did.

Cullen’s fourth criterion – that the tax cuts won’t lead to greater inequality – comes closest to relevancy, but even it is misleading.   Since income tax payments are highly skewed to begin with (14% of taxpayers pay 53% of income tax), it is very difficult for tax cuts not to "favour" the rich.   Seemingly fair proposals, like a proportionate reduction in everyone’s income tax, would fail a strict interpretation of this criterion.

The tax-cut option this criterion implies is a tax-free income threshold (making income below $6000 tax-free would cost around $2bn p.a.).   But that option lacks appeal no matter your goals.   The main benefits of cutting taxes come through reducing deadweight losses to society, and increasing incentives to work.   The former goal argues for reducing the top marginal rates (where the deadweight costs are greatest).   However increased work incentives would be best achieved by reducing marginal rates for middle income households most likely to be caught in the "working for families" welfare trap of high abatement rates.   A tax-free threshold is unlikely to achieve either function.

Alternatively, if the policy goal is to reduce inequality, tax cuts are a terribly inefficient instrument.   People in poverty tend to be paying very little in the way of taxes to begin with, and would receive little of the overall benefit from tax cuts.   Transfers, not tax progressivity, account for the majority of income inequality reduction in developed nations.

As guiding criteria for policy, Cullen’s tax cut principles are meaningless. As political criteria, they will no doubt be invaluable.   Labour’s tax cuts will almost certainly violate Cullen’s criteria by entailing less spending and more borrowing, more inflationary pressures, and increased inequality.   But National’s tax cuts will presumably do the same, only more so.   Cullen’s criteria simply allow the Labour government to defend an arbitrary line in the sand.