Auckland housing derp

Every economist, at one time or another, has been driven to desperate measures by a lack of data. But few economists have gone as badly wrong as whoever did the Labour party’s most recent “research” on Auckland house sales data.

Labour then grouped buyers into ethnicities based on surnames using electoral roll data.

Following the surname analysis, Labour found buyers of Chinese descent accounted for 39.5 per cent of sales in Auckland during the period, Twyford said.

The figure is more than four times the 9 per cent level of ethnic Chinese in Auckland’s population who are New Zealand residents or citizens.

Oh my goodness. If it was that easy, wouldn’t you think someone whose job this is – say, Treasury or MBIE – would have done it already?

The inference is supposed to be that there are too many buyers with Chinese surnames for them all to be residents, so some of them must be from offshore. But that inference is rubbish, as pointed out by Shamubeel Eaqub in the same article:

The increase of property buyers with Chinese and Indian surnames corresponded with the inflow of recent migrants, he said, adding that recent migrants were largely Chinese and Indian.
The fact that Chinese and Indians were buying property fitted with the life stage and financial position of those migrants, Eaqub said.

And by Keith Ng in a post on Public Address over the weekend:

What Phil Twyford has done is just a sleight-of-hand with percentages:

39.5% of house buyers are ethnically Chinese…
…but the resident Chinese population in Auckland is only 9%.
9% of residents can only buy 9% of houses…
…so 30.5% must be non-residents! Ta da!
Here are the same numbers, in absolute terms:

3,500 house buyers are ethnically Chinese…
…but the resident Chinese population in Auckland is 126,000.
126,000 residents can only buy 126,000 houses…
…so, uh, yeah.
It is entirely plausible that 126,000 people can buy 3,500 houses.

Keith’s 3500 number is based on 39.5% of recent REINZ house sales in Auckland. In today’s Herald article I see that Labour’s analysis was based on a dataset of only 3922 houses, meaning that the absolutes we’re dealing with here are even smaller, and Labour’s inference even less reliable, than first appeared.

Along with comments about the statistical and economic flaws of reasoning in this piece, comments also appeared pointing out the racism of the approach. That gave Labour housing spokesman Phil Twyford the opportunity to try for the reasonable corner:

Others, notably in the Twittersphere, have been quick to accuse me of racism because I have talked about a particular ethnic group.

But here’s the thing. We do need to have a mature public debate about Chinese foreign investment in New Zealand real estate…

It is simply not good enough to try to shut down an important public debate with allegations of racism.

Nice try, but those accusations are coming from such reasonable voices as Eaqub’s – hardly a hysterical denizen of “the Twittersphere”:

Eaqub said it was hard to understand the legitimacy of Labour’s data without knowing who supplied it and the conditions surrounding the figures.

“It draws this line across race and ethnicity, which is very damaging for a multi-cultural, welcoming place like New Zealand.”

Labour presumably know that their research is rubbish. Their aim is to accomplish something, not to produce useful analysis, and Twyford tries to make it clear what they’re trying to accomplish:

Twyford agreed that the government needed to make data relating to foreign buyers accessible.

There would not be a debate about the legitimacy of the data if the government gathered data about foreign buyers and made it publicly accessible.

It was “cynical” and “irresponsible” of the government to deny this was an issue and to refuse to make the data available…

Again, nice try, but the data availability issue is already addressed:

The government has refused to set up an overseas buyers register like Australia uses, but said in this year’s budget that foreign buyers would need to have an IRD number and a New Zealand bank account from October.

Eaqub said these conditions, coupled with transparency of the data, would paint a more accurate and legitimate picture of who was buying property in New Zealand.

The problem of housing affordability in Auckland isn’t about Chinese buyers. We don’t even really know whether it’s about speculation or offshore investors or supply. The question is hard, and that doesn’t have much to do with data availability – it’s just plain difficult to separate one cause from another when studying the behaviour of a market.

One thing we do know, though, is that shrill and divisive commentary like Twyford’s does nothing to help answer the questions. Instead it does everything to make it harder to get possible legitimate answers heard.