The payroll tax that dare not speak its name

Milton Friedman's celebrated put-down of American "Social Security" [Barron's, Aug.24, 1998] is mostly applicable to Australia's Superannuation Guarantee (SG),

http://www.hoover.org/publications/hoover-digest/article/6459.

In Friedman's redacted words, the SG is "...a combination of two separate programs, neither of which would have a chance of being enacted on its own. One program is... a flat-rate tax on wages..., a regressive tax imposed at one of the worst places in the economic structure, where it introduces a gap between what the worker receives and what it costs the employer. ...The second component is a... program in which the largest subsidies go to the people who have earned the highest incomes. Yet when you tie the two together you have a sacred cow. I have always said it is one of the greatest marketing miracles..."

Defenders of the SG will triumphantly retort that it isn't a tax, because it isn't payable to any government; it's a compulsory transfer of income from employers to super funds on behalf of workers.

But the key word is "compulsory." That a compulsory transfer doesn't pass through the government's hands doesn't make it any less offensive to liberty, property or economic efficiency, but merely takes it "off-budget," enabling the government to pretend that its economic footprint is smaller than it really is.

In all other respects, Australia's federally mandated, employer-funded 9% superannuation contribution is equivalent to a federally-funded 9% contribution paid for by a 9% federal payroll tax. At least the Americans are honest enough to call their tax a tax.

In as much as the SG is equivalent to a tax-and-transfer program, it defeats its own purpose, namely to reduce public expenditure on social security. Indeed, the effective payroll tax already raises slightly more revenue than the GST, and about twice as much as is needed to pay for the current age pension!

Worse, the effective payroll tax applies to Australian labour embodied in goods and services exported from Australia or competing with imports, but not to foreign labour embodied in goods and services imported into Australia or competing with Australian exports.

Thus it is a tax on Australian jobs, Australian products and Australian competitiveness, hence a tax on Australians' capacity to save for their retirements!

Moreover, because the tax feeds into prices, it is partly paid by the retirees who were meant to benefit from it; that is, it leads to churning.

The mining industry, which is not labour-intensive relative to the value of its output, can probably afford this de-facto tax on its payrolls. The same can't be said for other trade-exposed industries, especially when the high dollar (due in part to the rude health of mining) is damaging their competitiveness - a problem mentioned five times in Wayne Swan's budget speech.

The GST, of course, also feeds into prices. But because it taxes goods and services in the country of consumption rather than the country of production, it doesn't discriminate against Australian products as the SG does.

Hence lifting the SG rate from 9% to 12% in order to increase saving - as the superannuation industry demands, as the Government intends, and as the voters seem to want - is patently worse than lifting the GST rate from 10% to (say) 13.5% for the same purpose. Imagine the political reaction to the latter!

If we must have a tax to finance superannuation contributions, what sort of tax should it be? Friedman again: "In my opinion the least bad tax is the property tax on the unimproved value of land, the Henry George argument of many, many years ago" [interview in the Times Herald (Norristown, PA), Dec.1, 1978],

http://cooperativeindividualism.org/friedman-milton_interview-1978.html.

That's because the land cannot be created or destroyed or moved by the taxpayer, while its value depends not on any activity of the taxpayer, but only on the potential for such activity. Collecting the "tax" doesn't suppress that potential, but rather compels the owner of the land to realize its potential, in order to get income to pay the tax; whereas other taxes reduce economic efficiency, this one enforces economic efficiency.

Indeed, as long as a "tax" on the value of land is kept at a level that doesn't cause the resale price to fall, it is not so much as tax as the withdrawal of an unearned benefit; its effect is to prevent the full privatization of publicly generated increases in the value of the land.

In that case, the "tax" is really the rent on that part of the land value which has not yet been privatized. It is therefore no more distorting than if the government had a stake in the land, just as a private entity - such as a super fund - might have a stake in it.

It follows that if the Commonwealth must intervene to boost retirement incomes, there is no more efficient way to do it than to expand the age pension, and to pay for the expansion by imposing a charge on land values - presumably with an exemption for residential land owned and occupied by persons of pensionable age, in order to avoid churning. All other savings - including superannuation - would be purely voluntary.

When that happens, Wile E. Coyote will catch the Road Runner, Tweety Pie will go down the hatch and James Bond will get lasered in half from the crotch upward.

With more than $1.2 trillion worth of assets under management, and with its very own payroll tax with which to buy up more and more of the asset classes that have traditionally conferred political power, the superannuation industry will soon become, if it has not already become, the most powerful vested interest that has ever existed in this country.

Any attempt to cut off the tax stream will be portrayed by the super funds as an attack on their hardworking members, notwithstanding that the members pay the tax.

Rational but poorly funded arguments against the system will not be able to compete with well-funded, hysterical soundbites in its favour. Compulsory super is here to stay. Indeed, as the clamour for a 12% rate shows, it is here to expand.

So the best we can hope for is that Friedman's "least bad tax" will be used, not to fund the age pension, but to replace the effective payroll tax that funds compulsory super contributions.

Hence the best chance of reducing the distortionary taxes that presently pay for the age pension is to displace that pension by expanding superannuation funded by the "least bad tax."

Here's how the system might work:

The present de-facto payroll tax is abolished (the saving being passed on in higher wages and lower prices). In its place, every person of working age, whether working or not, receives two compulsory contributions to a private super fund, those contributions being:

Part A: a levy payable by that person, of so many percent per annum on the value of residential land owned and occupied by that person; and

Part B: a per-capita share of the revenue from a similar levy on the value of commercial land, rural land, and non-owner-occupied residential land.

Part A is off-budget and is strictly for the benefit of the payer, neutralizing the political sensitivity of a levy on the "family home." As it is not payable by retirees, it largely avoids the "asset-rich, income-poor" issue that often arises with municipal rates.

Part B, which could be on- or off-budget, is meant to become a non-means-tested replacement for the age pension. Consistent with this purpose, the payout from Part B is taken as a lifetime annuity and its preservation age is greater than for Part A. But, as a political sweetener, and as a dividend from the economic expansion that follows the abolition of the de-facto payroll tax, the preservation age for Part B is less than the present pension age, and won't need to be increased as the pension age will be.

The preservation age for Part B is taken as the upper limit of "working age": Part A is not payable by, nor Part B for the benefit of, persons past that age.

As a test of political feasibility, let us estimate the rate per annum at which the two levies would raise enough revenue to replace the SG.

In 2009/10, Australians earned about $549 billion in wages and salaries [according to ABS 5204.0 Tab.6]. The SG should have raised 9% of that, i.e. $49.4 billion.

The alternative revenue base would be land values as assessed in mid 2009. Because residential land owned and occupied by persons of pensionable age is exempt, taking the base as the total value of residential, commercial and rural land would overestimate the base, hence underestimate the rate.

To the contrary, one could argue that a lower rate, causing a lower level of saving, is acceptable because the abolition of the de-facto payroll tax would reduce prices for retirees.

In any case, the total value of residential, rural and commercial land in mid 2009 was $2821 billion [ABS 5204.0 Tab.61], and if this is taken as the base, the required rate is 1.75% per annum.

For a non-retired home owner with a land value of $300,000, the new superannuation charge would be $5250 per annum. In the case of land purchased before the reform and still mortgaged, the lender would pay the charge for its part of the equity in the land. For the rest, the owner would be compensated by a fall in prices (comparable to abolition of the GST) and better job opportunities.

One statement omitted from the above "redacted" quote concerns the American payroll tax, of which Friedman said: "You cannot conceive of that tax being enacted by itself as part of a financing program."

In Australia we not only conceive of it, but actually do it: payroll tax, called by its proper name and used for general revenue, is the mainstay of State budgets.

So, for interest's sake, let us estimate the rate at which the "least bad tax" would replace the payroll tax that does speak its name.

In 2009/10, Australian State and Territory governments collected a total of $16.8 billion in payroll tax [ABS 5506.0]. On the same land-value base as above, the required rate is 0.6% per annum. This is a weighted average for the whole of Australia; individual States and Territories could deviate from the average.

In this case, because State payroll taxes are paid by retirees (through prices of goods and services) but not earmarked for their benefit, there is no argument for excluding residential land owned and occupied by retirees from the alternative revenue base.

But, as with council rates, there is a strong argument for allowing "asset-rich, income-poor" owners, presumably including many retirees, to defer some of the payment until the land is next sold (which, if the land is bequeathed, would mean the first sale after the bequest).

So the substitute for payroll tax could be implemented as a State surcharge on council rates. The additional 0.6% per annum would apply to the land value only, not the combined value of the land and building(s).

If State payroll taxes were replaced by annual charges on land values, the result would be more jobs, better pay, lower prices, and improved international competitiveness.

If the federal payroll tax masquerading as the Superannuation Guarantee, which raises almost three times as much revenue as State payroll taxes, were similarly replaced, the benefits would be proportionally greater.