The tax unit for an asset-holding tax should be the asset!

Gavin Putland

Recommendation 10 of the Draft Report of the IPART Review of State Taxation suggests “changing the tax unit for land tax from joint ownership to the individual” as a means of reducing complexity caused by aggregation of site values.

That raises the question: As land tax is a tax on site values, wouldn’t it be simplest if the tax unit were the site? Our first submission to this Review proposed a mechanism which would indeed make the tax unit the site, without creating winners and losers in the transition to the new system. With a little elaboration, the transitional arrangement can account for the effects of aggregation before doing away with it.

Origin of the problem

Because the acquisition of major assets requires income in excess of necessary consumption, the distribution of major assets, including land, is more unequal than the distribution of income. And because land tax (in NSW and other Australian jurisdictions) exempts owner-occupied residential land, the ownership of taxable land is more concentrated than the ownership of land in general. For these reasons, land tax would be strongly progressive even if applied at a flat rate with no threshold. The purpose of a threshold is not so much to make the tax more progressive as to limit the number of taxpayers, especially among swinging voters. But a threshold, by itself, creates the opportunity for landowners to minimize tax by holding a large number of low-valued sites, this strategy being especially viable for urban speculators who “buy at the fringe and wait”. To close the loophole opened by the threshold, site values are aggregated before the threshold is applied.

In whose hands, then, should site values be aggregated for the purpose of taxation? That is the question addressed by Recommendation 10. But the mere fact that values must be aggregated in someone’s hands explains why the tax unit is not simply the site.

Solution – with no losers

The first proposal in our first submission said in part:

Replace all recurrent property taxes levied by the State, including land tax and any “levies” or “duties” that amount to recurrent taxes on property, with an incremental land tax (ILT)that is, a recurrent site-value tax with a flat rate, and with an inflation-adjusted threshold that varies from site to site so that the ILT initially payable on each site in the first year of the new system is equal (in real terms) to the total recurrent property tax payable to the State in respect of the same site and its improvements in the last year of the old system…

This proposal assumes that the tax unit is the site, which is obviously not the case under present aggregation provisions. Therefore the transition to the new system requires a formula for apportioning the taxable aggregate site value among the individual sites. That our first submission neither offered such a formula nor acknowledged the need for it may explain why the Tribunal members were not persuaded by the “ILT” proposal. But the omission is easily remedied.

First note that the formula would not be contentious. The distribution of the annual tax bill among the properties owned by one tax unit does not affect the total annual tax payable by that unit unless and until one of those properties is transferred with the tax liability attached. And while the annual tax liability attached to that property affects its market value (i.e. its transfer price), the total market value of the properties owned by that tax unit is insensitive to the distribution of the total tax liability among the properties. So the formula for apportioning the tax liability does not need to be designed with great political sensitivity. It only needs to be simple and unambiguous.

For example, the total annual tax liability for the several properties owned by one tax unit could be distributed among the properties in proportion to their site values. The tax thus calculated for each site would then determine the threshold for that site. In the case of fractional ownership, the share of a property or site owned by the tax unit could be treated as a property or site under the foregoing procedure in order to determine the tax payable on each share. A simple addition would then yield the tax payable on each (whole) site, and thence the threshold for that site.

Redistributive effects

As the transition to the ILT would not create winners and losers, it would not of itself have any redistributive effect and, therefore, would not of itself frustrate the intent of the existing aggregation provisions. After the transition, however, there could be a redistributive effect as taxpayers arranged their affairs in response to the new rules. In particular, speculators could buy large numbers of low-valued sites in order to take advantage of multiple thresholds. But the reforms proposed in our first submission would counteract that effect in two ways.

First, the ILT would replace not only land tax, but all recurrent taxes on property, including those not based on site values. This would tend to make the recurrent tax on a property more sensitive to the site value, which is the appreciating component of the overall value. Thus the tax burden would become more concentrated on speculative gains, reducing the speculative motive. Second, the other tax proposed in our submission, namely the “site windfall tax” to replace conveyancing stamp duties and lump-sum development levies, would also target speculative gains.

In our view, for the purposes of equity, the deliberate concentration of the tax burden on speculative gains is a more than satisfactory substitute for aggregation. And of course to eliminate aggregation altogether – by making the tax unit the site – is a greater simplification than any other adjustment of the tax unit for aggregation purposes.