Carbon is poison to tax reform

YOU don't need to be a professor to know that any policy sells better if it can be linked to tax cuts.

And packaging those tax cuts as "productivity enhancing tax reform" is even smarter: bribes are cheap and tawdry; compensation, Ross Garnaut tells us, is all too readily "unprincipled"; but tax reform is beyond reproach.

Good politics, however, may not be sound economics. And there are tax reforms and tax reforms. So the issue is whether the tax changes flagged in Garnaut's recent report on carbon pricing, and the ones the Prime Minister has hinted at since, would give us a better tax system and a stronger economy. The answer is they would not.

Start from the basics. What Garnaut is suggesting is that carbon pricing could yield a double dividend: environmental gains on the one hand, and tax changes that favour work and savings on the other.

What that suggestion ignores is the fact that introducing a carbon tax is neither a necessary nor sufficient condition for wider tax reform. As the Henry report concludes, there are many changes to our tax and welfare system that would reduce the distortions it creates at present. Those reforms can and should be made quite regardless of whether a carbon tax is introduced, as they would lead to a better, fairer, tax system.

As a result, ascribing the benefits of those changes to the carbon tax is not sensible. Rather, the relevant question is what additional gains a carbon tax would bring, above and beyond those tax reforms a sensible government would make in any event.

To answer that question, it is important to know that the first round impact of a carbon tax is to increase the distortions any practical tax system creates. That first round impact arises from what economists refer to as "tax interaction" effects. These occur when a new tax aggravates the harm existing taxes do to the incentives to work and save.

Consider, for example, a tax on driving on congested roads: a congestion charge, in other words. Such a charge may seem like a good idea. However, if most of those affected are travelling to or from work, that charge amounts to a tax on working.

But working is already taxed by income taxes, payroll taxes and the GST (as basically, we work so as to spend money consuming). And given the government's revenue needs, any viable tax system will have that distortion in it. Relative to that baseline, a further increase in marginal tax rates can have high economic costs. Those costs must be considered in assessing whether a congestion charge would yield net benefits.

Exactly the same issues arise, albeit more subtly, with a carbon tax or an emissions trading scheme.

Such a tax increases production costs, as firms incur outlays both in actually reducing emissions and in paying for however much carbon they continue to emit. Those cost increases raise prices and cut real wages, which discourages labour supply.

But because labour is already taxed, labour supply is already lower than it ideally would be. Moreover, the economic loss from discouraging labour supply rises rapidly as the effective tax rate on labour rises. As the carbon tax increases that effective tax rate, its costs are therefore greater than an assessment that viewed it in isolation would suggest.

Substantial revenues, however, are also likely to be raised by the carbon tax. Wouldn't the sheer mass of those revenues necessarily allow such large cuts to other distorting taxes as to yield net economic gains?

Unfortunately not. This is because the greater the revenues are, the greater are also the new tax's first-rounding distorting impacts. As a result, raising lots of dollars does not in itself allow the tax interaction effects to be more than offset.

In practice, however, even just offsetting those effects may not be achievable. And it certainly would not be achieved were the tax reforms Garnaut points to implemented.

Those reforms are based on the Henry Review's recommendations about income tax. But under the review's indicative income tax schedule, the marginal tax rate for individuals with taxable incomes between $25,000 and $37,000 would more than double, while that for individuals between $37,000 and $80,000 would increase from 30 per cent to 35 per cent.

Given that tax interaction effects of the carbon tax would be aggravating those increases, the result would be substantially higher economic losses.

And matters would be even worse were the proceeds from the carbon tax used to fund increases in the Low Income Tax Offset, as Julia Gillard has since proposed. That offset tapers as income rises, creating high effective marginal tax rates at the lower end of the income distribution.

The result of the Gillard approach would therefore be twofold: effective rates at the top would rise because of the carbon tax's interaction effects, reducing participation by the most highly skilled; while at the other end of the income scale, the increase in LITO would entrench disincentives our current tax-transfer system creates.

Gillard's claim that the overall effect would be to "boost productivity and participation" is therefore not credible.

Ultimately, Garnaut's linkage between wider tax reform and the carbon tax is a diversion from the real issue, which is whether unilaterally introducing a carbon tax in an economy whose comparative advantage lies in carbon-intensive production makes any sense.

And making tax reform hostage to a carbon tax can only compound the likely harm. It will jumble objectives, as buying support for the carbon tax comes to shape the tax changes that are adopted. And far from avoiding the unprincipled, rent-seeking transfers Garnaut rightly decries, it will expand the instruments that can be used to shuffle wealth around, multiplying the waste incurred along the way.

The final outcome is entirely predictable: a tax system that is even more complex, even less efficient and even less fair. Now justifying that really does require a professor.