Foreign buyers again

I was a bit surprised that Labour thinks that the latest Statistics New Zealand release on foreign home buyers is evidence in favour of their ban. 

Bernard Hickey writes:

Yesterday Statistics New Zealand reported that 8.0 percent of buyers across New Zealand were either not citizens or held only temporary visas, while 19 percent of buyers in the old Auckland City Council area (now known as the Waitemata local board) were not citizens. See chart above. Across Auckland, the percentage of foreign buyers was 7.3 percent, while it was 9.7 percent in the Queenstown Lakes District.

The figures are significantly higher than LINZ's figures, which focused on tax residency rather than whether buyers were citizens.

Economic Development Minister David Parker said the figures vindicated the Government's looming ban on foreign buyers of residential properties.

"I think it's clear that these statistics back up our decision to ban overseas buyers of existing New Zealand homes. We've no doubt that buyers are having an affect on New Zealand housing market," Clark told RNZ.

"How significant that price is no one really knows but in those suburbs where there's an 18 percent participation by foreign buyers buying that number of houses it must be having a significant effect I would have thought."

I'd had a look through these numbers when they were released yesterday.

First up, a trivial bit: I'm pretty sure that Bernard's transposed the wrong figure. Table 1 of the release has 8.0% of home transfers in which at least one NZ resident visa holder (but no citizens) are buyers, but 3.3% of buyers being neither citizens nor residents. So when Susan and I sold our place in Christchurch, and bought our place in Wellington, we'd have been in the "At least one NZ resident visa (but no citizens)" category* because we have permanent residence but have not yet applied for citizenship. I'm pretty sure that Hickey's 8% is an error that substantially overstates the extent of foreign buying - and that I hope doesn't get repeated elsewhere.

But here are the more substantial takeaways.

First up, I remember there being a big deal made about the "affiliation unknown" category last year and how that could be skewing things. In December quarter 2016, when they started collecting these stats, that category was very large: multiples of the figure for which they had citizenship data. Some folks suspected that the unknowns were disproportionately hiding foreigners.

The quarterly number of "affiliation unknown" buyers dropped from 36,237 in December 2016 to 4,287 in March 2017, to 450 in June 2017, and ratcheted down to 39 in the current March 2018 quarter. So the 'unknowns' problem was largely solved by September 2017.

Over that period, the proportion of "No NZ citizens or resident visas" among those buyers for whom affiliation is known went from 2%, to 2.1%, to 2.4%, to 2.3%, to 2.9%, and is now 3.3%. The biggest jump was from September 2017's 2.3% to the current quarter's 3.3%. But by September 2017, they'd basically identified all the buyers' affiliations.

So either there was never any substantial hidden foreign buyer contingent in the "affiliation unknown" category, or there was and many of them exited the market. The increase in the proportion of foreign buyers was really after they'd sorted out identifying foreign buyers.

Next up, the recent surge. I expect this is best explained by folks rushing to get ahead of the ban. Anybody who is here on a work visa and is trying to sort out residence, or who hopes to renew their visa, and has any ability to buy a house, will want to get that sorted absolutely immediately before they're banned from buying anything. The growth in sales by foreigners has not seen the same recent surge - although that might yet come if living here becomes sufficiently uncomfortable for foreigners.

It's also worth having a very good sense of proportion here. That 3.3% of buyers, across the country as a whole, in the March 2018 quarter, represented 1,083 house purchases. In the same quarter, non-citizen non-residents sold 501 houses. So, on net, in March 2018, "foreigners" (which will include people living here on non-resident visas, whose residence visas may be in progress) on net acquired five hundred and eighty-two houses in that quarter.

If we include all sales over the year to March 2018, we have 3,834 home purchases, across the whole country, by non-citizen, non-residents - and 1,899 home sales. So, on net, foreigners owned 1,935 more houses in New Zealand at the end of the year March 2018 than they did at the start of it.

Labour has implemented a ban on all house sales to foreigners on the basis of this. Just under two-thousand houses - some of which could have been new builds financed by foreign buyers in the first place.

Sales to foreigners are relatively concentrated in two places: Auckland and Queenstown.

In Auckland, in the year ended March 2018, 2,307 purchases (5.7% of purchases)** were by non-citizen non-resident buyers. That category also sold 1,050 homes (2.6% of sales). So net foreign ownership in Auckland, for the whole year, increased by 1,257 homes. In a city of 1.5 million people.

As a proportion, some of the figures can look bigger. If you look for the place with the biggest proportion of non-citizen, non-resident buyers in the year to March 2018, it's Auckland's Waitemata. 12.9% involved non-citizen, non-resident buyers. That's 573 houses. Non-citizen, non-residents there also sold 384 homes (8.7%).

Now if you think effects in Waitemata are particularly important, then maybe Labour could just have banned foreigners from buying houses and apartments in Waitemata. Now that would be silly because it would just encourage people to shift outside of that boundary line - but that also reveals the problem in asserting stuff like "Well, those horrible foreigners must have had a big effect in Waitemata! Look at their proportion there!" Auckland's market is bigger than that. An increase in demand in Waitemata will have some folks with weaker locational preferences pick spots outside of Waitemata. So any effect would be spread across the Auckland region.

In Queenstown-Lakes, foreigners made 7% of purchases (129 homes for the whole year) and 5.1% of sales (93 homes for the whole year).

Can foreigners on net acquiring thirty-six more houses in Queenstown and 1,257 more houses in Auckland in a single year really blow up the the housing market in those cities? If it does, shouldn't we be terrified of the financial stability risks inherent in Auckland? It should not be possible to break the housing market in a city of 1.5 million people by buying a couple thousand houses. If it is possible, then how many mortgages would rapidly be underwater if demand dropped by 3,000 houses in a year for some reason?

It all suggests to me that we don't need a ban on foreign buyers but rather reform to the supply side to unfreeze that blade of the supply-and-demand scissors so that quantity can start adjusting.

During the election campaign, Labour liked to note restrictions on sales to foreigners in other countries, and then Labour turned that into a justification for a ban. But foreign examples are almost never a ban. Denmark has something close to a ban on sales to non-EU citizens, and Austria will make you get permission from local government. But look at the US, the UK, Canada - heck Portugal will give you a residence visa if you buy a house there. Like, pick your favorite model country, and check whether a Kiwi could buy a house here. This site seems as good as any for checking that, though it might be missing some recent changes.

I guess I just have trouble seeing how it makes sense to ban sales to foreigners, across the whole country, to prevent 2,307 sales to foreigners in Auckland in a year whose figures will be somewhat inflated by folks rushing to get in ahead of the ban. And remember that the ban isn't costless to locals either.

Update: Hickey's Monday newsletter, at the very bottom, after his collection of interesting links, notes that the correct figure is not 8%. 

* Or we would have been had those stats then been collected.

** Note that I am here figures from Table 3. They differ from the Table 2 figures that Bernard Hickey is using. I am not sure why they differ, but I am using Table 3 because it provides the absolute figures rather than just the proportions, and because it includes the better regional breakdown - Table 2 does not have a separate breakdown for Queenstown-Lakes, for example. Since all of this is small number stuff, small changes in small numbers can make for bigger looking differences in proportions.