For most of the latter-day tax-cutters, from Preston Manning to Hugh Segal with a couple of bank economists wedged in between, the first explanation seems to prevail. This has managed the amazing feat of making Paul Martin seem like the voice of fiscal rectitude. The notion that, because the Finance Minister has come in about $4-billion below his deficit targets, he now has $4-billion to splash about in tax cuts, can only be recorded as the final victory of the spin doctors.

The targets are benchmarks of precisely nothing. They were chosen not because they had any objective relevance to the economy, but because they were reckoned to be especially easy to beat. They are of no significance to anyone except Liberal campaign planners.

For that matter, neither is the deficit. It would be risible enough to be talking about returning a $4-billion "surplus" to the taxpayer even if we were anywhere near an actual surplus of revenues over outlays, rather than a "surplus" over Liberal propaganda. The problem is not the deficit any more, and hasn't been for some years. It's the debt, and more particularly, the cost of servicing debt.

So long as the federal debt remains near 75 per cent of GDP, and interest charges consequently consume in excess of one-third of revenues, annual budgets, whether deficit or surplus, are essentially meaningless. Until we have brought those ratios down a long way, every surplus dollar we can get our hands on will have be devoted to debt reduction. That certainly rules out the kind of unfunded tax cuts the Conservatives seem to be playing with. But it doesn't leave much more room even with offsetting spending cuts, as the Reform party proposes.

In any case, where did we get the idea that "putting more money in the hands of consumers" was the elixir of economic life? Even in the crudest Keynesian terms, a tax cut that is accompanied by equal or greater spending cuts would have no stimulative effect whatever: the one would "put money in," while the other "took money out." On the other hand, any stimulus from a tax cut that increased the deficit -- that is, with no offseting spending cuts -- would soon be unwound by the same combination of knock-on effects that defeats all such pump-priming efforts. Some of the extra demand from the tax cut would be dissipated in higher interest rates, some in purchases of imported goods, some in saving against anticipated future tax increases and so on.

The popular belief that the economy is "driven" by consumer spending, and that any time the economy is not growing at top speed it should be topped up with another dose of consumption, is perhaps the worst residue of the Keynesian experiment. Precisely the opposite is true: it is only income that is not consumed -- savings, in short -- that provides the capital to finance new investment. An economy that consumed everything it produced could not grow at all.

If the intent of cutting taxes, then, is to stimulate consumption, it will fail to increase economic growth in the short term, and reduce it in the long term.

The real case for cutting taxes is to improve incentives for production: to boost supply, that is, not demand. For this what matters is not to reduce tax revenues, but to reduce tax rates. A revenue-neutral tax cut, one that cut tax rates while broadening the tax base, would not be a short-term fix: nothing is. But it would increase the economy's long-run productive capacity.

Tax cuts, Keynesian-style, are conceived as a sort of gift to the taxpayers: they are simply handed back more income after tax, without any change in behaviour required. A supply-side tax cut, on the other hand, encourages people to act in ways that increase their income before tax: at lower marginal tax rates, they have more incentive to make the next investment, to work the extra hour, to save an additional dollar.

Moreover, the other half of the scissors -- a broader tax base, with fewer exemptions and deductions -- far from cancelling out the effects of lower rates, would actually amplify them. Decisions between alternative investments would be less influenced by tax considerations, and more by the real economic returns attached to each.

The only kind of tax cut that is remotely responsible in our present fiscal straits is a revenue neutral tax cut. Happily, that's also good policy for the longer term.