Brian Fallow: Imposing a payroll tax would be sheer Muldoonacy

Some propositions should be ruled out on first principles.

One is the idea the Labour Government could impose a new tax on labour in order to lessen tax on the returns to capital.

Yet Finance Minister Michael Cullen remains coy on whether a payroll tax is under consideration as part of the "big and bold" review of business taxation. Nothing can be ruled in or out yet, his office says.

But he confirmed in Parliament this month, in response to National's John Key, that the Treasury had, at his request two years ago, calculated that, on a given set of assumptions, it would take a payroll tax of 5.4 per cent to replace the revenue lost by cutting the company tax rate to 20 per cent.

If Cullen really is toying with this idea, he should stop and drive it from his mind with blows and curses.

Or his colleagues should do it for him. It is Muldoonacy.

One reason the 3.6 per cent unemployment rate is enviably low by the standards of other developed countries is that the tax wedge on wages - the difference between what it costs to employ someone and his or her take-home pay - is relatively narrow. Widening it from the low 20 per cents to the high 20s by an additional tax on wages and salaries would undermine that advantage.

While the tax might be paid by employers (all but the smallest if the Australian states' model is followed) the economic cost will largely fall on workers. Either their after-tax pay will be less than it might otherwise have been or there will be fewer of them.

Business Roundtable executive director Roger Kerr said that where labour was internationally mobile firms might be forced to increase gross wages and maintain after-tax pay to compensate for a payroll tax. All else being equal, profits would be squeezed.

"In this case, fewer such people would be employed over time as firms would make adjustments to maintain a normal return on capital."

Kerr said firms which exported or had to compete with imports would be unable to pass on higher costs to customers. They would also reduce their demand for labour.

Firms in the non-tradeables sector which escape such competitive disciplines might be tempted to pass on the increased cost, which would be inflationary and push interest rates higher than they otherwise would be.

A payroll tax would be a retreat from the sound tax policy principle of neutrality.

It would increase the incentive, wherever the option arises, to recognise income as business profit rather than wages or salary. That, of course, would be especially so if the company tax rate were slashed.

Policing that boundary, one suspects, would become a nightmare.

In light of these fairly obvious drawbacks, why might Cullen be tempted by the idea?

It may be that he sees it as a step towards increasing the capital-to-labour ratio in the economy - the amount of plant, machinery and software firms invest in a worker.

In some ways, a payroll tax is equivalent to a tax on consumption like GST, in that most people spend most of what they earn in wages. Indeed, the Reserve Bank estimates that the household savings rate is negative with people spending about $1.14 for every $1 of income.

But if you wanted to change the tax mix to encourage a shift to less consumption and more investment, why not just increase GST and cut income tax?

There has undoubtedly been a major structural shift in the economy in the past 15 years to a permanently tighter labour market. Unemployment has trended down from more than 11 per cent in the early 1990s to below 4 per cent now, while the participation rate - the proportion of working-age people in the workforce - has risen to internationally high levels.

Even the economic slowdown now clearly under way is only expected to push it back up towards 5 per cent, where it was earlier in the decade.

The problem is that most economic growth in the past 10 years or so is explained by an increase in labour use rather than labour productivity - more hands to the pump rather than investing in a more efficient pump.

With the population ageing, and on average three people leaving the country for every four who immigrate, that formula for economic growth won't work any longer.

The low-hanging fruit has long been plucked from the labour supply tree and it would require a costly cherrypicker to get at what is left.

The Australians are richer than we are not because they are smarter, harder working or more enterprising, but because they have about 50 per cent more physical capital employed per worker, boosting productivity and incomes.

So they can afford to inflate employers' labour costs with a compulsory superannuation contribution of 9 per cent and state payroll taxes of about 6 per cent.

But it would be putting a heavy cart before the horse to impose a payroll tax here in hopes that it would encourage firms to substitute capital for labour.

They have enough incentive to do that already, given the experiences of labour shortages in the past few years.

Substituting one tax for another won't do any good. Lowering the overall tax burden might.

Business investment has increased markedly in the past few years. That growth is tapering off now as the economy slows and a falling exchange rate makes imported gear more costly.

It has occurred at a time when labour costs have been rising - but for the right reason, as the normal lagged response to a boom and a particularly tight labour market. Higher wage costs which did nothing for wage-earners would be a different thing.

Reserve Bank Governor Alan Bollard last week sounded a warning about "unrealistic" wage increases, yet the bank's own research tells it that wages tend to follow inflation, and indeed struggle to keep up with it, rather than drive it as they might have in the past.

Bollard's concern is evidently more about expectations, that higher-than-normal wage settlements will erode price setters' confidence that they still operate in a low-inflation environment.

In any case, real wages will have to rise if the brain drain is to be reduced.

In that context, a payroll tax would be an own goal indeed.