Fortunately, Singapore and Hong Kong - and some cities in China - have been down this path before. We can draw on their experience. Singapore imposes a 15 per cent stamp duty on non-resident purchasers of residential property, as does Hong Kong. In China, in order to restrain the property boom in 2011 cities such as Beijing and Shanghai - amongst others - imposed a blanket ban on non-resident purchases of residential property, which in practice blocked purchases by Chinese residents from other provinces.
I believe we should take the lead from Singapore and Hong Kong and impose a 20 per cent stamp duty on non-resident purchases of Auckland property...
The proceeds of the stamp duty could stay within the Auckland economy, with the duty receipts contributing to a sinking fund for Auckland infrastructure. As the Chinese leadership likes to say, this would create a win-win for Auckland - and for New Zealand.Now I didn't argue for a 20 percent stamp duty, but a much more modest 2.5 percent (similar to Australia), which I think is much more feasible and less costly on the economy to collect. For instance if foreign buyers make up 40% of Auckland buyers (Labour suggested 40% of buyers had Chinese ethnicity, but some of those will be New Zealanders; this will be offset by foreign buyers from other countries, so 40% may be illustrative), then a 20% stamp duty on 40% of the market raises about three times as much revenue as a 2.5% stamp duty on the entire market (ignoring for simplicity the decreased demand that would result from the stamp duty raising prices). However, the deadweight loss (the loss of economic welfare) generated by the tax is likely to be much more than three times larger. Consider the diagram below (the diagram shows a specific tax, which is a constant per-unit amount of tax, rather than an ad valorem tax, but the principle is much the same in either case). Without the tax, the price is P0 and quantity Q0. There is no deadweight loss. With a small tax (tax1), the per-unit value of the tax is the vertical distance AB, and the deadweight loss is the triangle AEB. With a tax that is three times larger per unit (tax2), the per-unit value of the tax is the vertical distance CD (about three times the vertical distance AB), and the deadweight loss is the triangle CED. It is clear that CED is much more than three times larger than AEB (with linear demand and supply curves and a specific tax, the deadweight loss is nine times larger when the tax is three times larger).
Neither did I suggest that stamp duty should only apply to non-residents. Taxes that generate the same income but apply to only some sub-markets also generate larger deadweight losses than taxes that apply to the market as a whole. Again, thinking about the diagram above a tax that generates the same income from only one sub-market would need to be larger in per-unit terms, leading to a larger deadweight loss (even when you consider that there would be no deadweight loss on the sub-markets that are left un-taxed).
Moreover, I'm not convinced that the distinction between residents and non-residents is so easy to enforce in practice - many foreign buyers will have contacts (or children) in New Zealand that can execute purchases for them. Applying stamp duty to only some purchasers will increase the bureaucratic costs associated with its collection.
However, I do agree with Jones that the proceeds of a stamp duty should be used for infrastructure provision, or housing development. An increase in supply would act to dampen house price increases in Auckland. Bring on stamp duty!