Probably like most people I’d got a bit tired of the foreign buyers’ tax revenue estimates issue. At best I can tell, something like the estimates we produced seem more plausible than National’s numbers. In the grand macroeconomic scheme of things, the differences are fairly second order (0.1 per cent of GDP vs the 2.7 per cent of GDP gap between revenue and expenses this year in the PREFU). But it has not been reassuring that National has simply refused to release their modelling/calculations or authorise the release of the modelling said to have been done for them by Castalia. It is an evasiveness that doesn’t speak well of a party that has come out in support of a state Policy Costing Office, and whose leader just this morning is reported to have championed an overhaul of the OIA (presumably to encourage more transparency/accessibility). No one can make them publish, but if they don’t we can draw our own conclusions. As a straw in the wind on that point, I saw National’s recent responses to questions on this issue from one media organisation, in which Willis seemed more interested in casting doubt on our independence, than in elaborating on their own work. (My earlier two posts are here (including the underlying paper) and here.)

Much as I quite liked the modelling exercise we did, in some ways I took at least as much comfort from what I could learn of the Canadian experience, most notably the revenue numbers from their foreign buyers’ tax in British Columbia – Vancouver having been a magnet for (in particular) Chinese purchases, and the tax having been applied at all price points. But I hadn’t made much effort to look more broadly, not really being that familiar with the data sources in other countries.

But this afternoon someone sent me a link to an Australian Tax Office insight note, with data, on foreign purchases and sales of residential real estate. Under the Australian system (accessible summary here) foreigners other than New Zealand citizens need FIRB approval to buy a house in Australia (this includes people on temporary work visas, but not permanent residents).

The report has data for four (June) years, to June 2021. The numbers appear to be for those classes of people who require FIRB (ie ex NZ purchases).

You might be inclined to discount 2019/20 and 2020/21 a bit, having been disrupted by Covid, including the border closures for much of those periods. But even if one starts from the 2018/19 numbers, it is a total of 7679 sales of dwelling (new and existing) in a full year, at all price points. Around 80 per cent of all sales (by number) were at price points under A$1m (this will include the vacant land sales, but they only make up about 20 per cent of the total).

The total value of foreign purchases in 2020/21 was A$4.2 billion. In 2018/19 – when the volume was higher – the value of sales in total was A$7.5bn. (Interestingly, whereas foreign purchases have been dropping, sales of such properties by foreigners – a much smaller number – were rising over this period.)

Had a 15 per cent tax been applied to that 2018/19 higher level of sales it would have raised only just over A$1.1bn even in 2018/19. Australia is, of course, five times the size of New Zealand (and the proposed tax here will exclude Australian – and Singaporean – buyers, as New Zealand buyers are excluded from the Australia numbers), so if it were really a like for like comparison this would be consistent with not much more than about $200m of annual revenue here. On the assumption that – unlike National’s proposal – purchases at all price points were covered.

However, this is not a clean comparison. The Australian land transfer market is riddled with complexity and costs. Not only does the FIRB process appear more constraining (especially on non-resident purchasers of other than investment properties – as distinct, say, from people living in Australia on work and study visas) than what National is proposing for properties selling for more than $2m, but there are stamp duties and the like. In Victoria, for example, foreign buyers in this period faced this additional transfer duty

All of which means I would not put too much weight on this report – comprehensive as it is in the Australian context – for insights on prospects for New Zealand if National’s policy were to be implemented. But in the grand scheme of things, Sydney and Melbourne are rather more “global cities” than Auckland (and about as expensive, in price/income terms), and Australia has – per capita – about as many migrants as we do. And what none of us know with any certainty is how many of the houses being bought by foreigners before our ban – outside Queenstown, small in the overall scheme of things for non-Australian buyers – were bought by people living here at least part of the time, as distinct from holiday homes

The most I would say is that it is just one more straw in the wind – but no more – that the National revenue estimates still seem implausibly high.

But perhaps things would look different if we saw National’s own or Castalia’s modelling.