Neutralizing stamp duty and development levies

It is appropriate that the surge in Melbourne home prices has rekindled debate on conveyancing stamp duty, but not so appropriate that the discussion has focused on the size of the duty instead of its base. Reducing the size of the duty must have a budgetary impact, but changing its base need not.

Stamp duty is a transfer tax on the total value of a property including the building(s) and the site (i.e. the land). By taxing buildings, the duty deters construction and thus reduces the supply of housing, driving up rents and prices. Furthermore, by separately taxing every transfer in the supply chain, it impedes the process of bringing new homes to market (and discriminates against owners who need to move frequently).

The solution is to apportion the duty to the increase in the site value since the last transfer of title, with a deduction for any cost incurred by the taxpayer in contributing to that increase (e.g. infrastructure built by a private developer and transferred to public authorities). Under this arrangement, buildings and other capital works are not taxed, and each transfer does not create a new tax liability, but merely realizes an already accumulated liability (and the discrimination against frequent movers is eliminated).

The change would put stamp duty on a rational, equitable, efficient base: the unearned increase in the site value.

For administrative convenience, the modified stamp duty would need to be payable by the vendor, not the buyer. But this would not affect its final incidence, which would be decided by the market: in the end, any transfer tax will be shared between the vendor and the buyer in inverse proportion to their bargaining power, regardless of who actually “pays” it. To avoid any suggestion of “retrospectivity”, vendors who bought before the change can be given the option of paying tax as if they had sold and bought back their properties on the last day before the change.

A similar reform can be implemented at the local level. At present, local governments require developers, as a condition of development, to pay per-lot lump-sum “development levies”, ostensibly to meet the cost of infrastructure headworks. The economic and moral justification for such levies is that the provision of headwords, together with permission to develop, increases the value of the developers' land. But the justification should be made official — by apportioning the levy to the unearned increase in the lot value.

As there may not be an official site valuation between acquisition and disposal of a property, how should the increase in the site value be assessed? One option is technology. If valuations are based on a computerized database of transactions, it should be possible to recompute valuations after each transaction is entered into the database, and to extrapolate established trends to obtain “up to the minute” valuations — although “up to the day” would probably suffice. Failing that, one can estimate the increase in the site value from the vendor's capital gain and building depreciation (as calculated for income-tax purposes). In the case of the development levy, the initial value of each lot could be estimated from its area and the initial value of the entire estate; the division of the total initial value among the lots would have little affect on the total tax bill.

Releasing more land for development, as the property industry keeps demanding, will not remove the bottlenecks further downstream. To do that, one must rationalize stamp duties and development levies.

[First posted at grputland.blogspot.com. Relocated Dec.30, 2009.]