Unprecedented…in a sample of four

The Reserve Bank had an interesting brief issue of the Bulletin out yesterday, reporting the results of some fairly straightforward data analysis as to how house prices (more accurately, house plus land prices) have behaved over the last fifty years or so, and how prices in Auckland relative to those in other parts of the country have behaved going back a few decades.  It is almost wholly a descriptive piece, offering no views on why things happened as they did, and little on what those patterns might mean.

The author –  Liz Kendall –  has done to work to construct a quarterly series for each TLA and a range of regional indices, using QVNZ data.  It would be nice to have those indices generally available, but perhaps restrictions on the use of the QV data precluded that?

The author identifies six distinct upswings in real national house (+land) prices since 1965, and as she notes the recent one has been relatively muted (for the country as a whole) –  despite all the fevered talk of low interest rates driving prices higher, ignoring the fact that interest rates are low for a reason (weak demand at any higher rates).

There wasn’t much of a upward trend in real New Zealand house prices, as far as we can tell, until the last 15 or 20 years.

rb housing 2

There was a huge boom in the early 1970s, as some combination of very low real interest rates and very rapid inward migration drove prices up.  But that increase was fully reversed over the following few years as credit conditions tightened, real incomes came under pressure, and significant net outward migration relieved pressure on the housing (house+land) stock.  The lack of any strong trend is consistent with what we see in the advanced countries that have a much richer longer collection of historical data –  including Australia and the United States.

Kendall illustrates that house prices in Auckland have not always been that highly correlated with those in the rest of the country.  There is clearly a common element to house prices in the country as a whole, but there are times when each region- including Auckland – “does its own thing”.  For example, in the 2002 to 2007 boom real house prices rose by materially less in Auckland than they did in the rest of the country (taken together).  And in the last three or four years, Auckland prices have risen much faster than those in the rest of the country.

The punchline of the article appears to be this chart

RB housing 1

For the period since 1981 –   only 35 years –  it shows the ratio between Auckland house+land prices and those in the rest of New Zealand, and a moving average trend in this ratio.  The trend has no economic significance –  it simply smooths through the ups and downs in the actual data, and the more recent observations might end up being quite materially revised (up or down) as new data emerge (there are always “end point problems” with these sorts of filters).  The ratio is currently 22 per cent above this particular trend line, but when we look back 10 years hence who knows how large we will estimate that gap to have been.

I had just a few other thoughts on the article:

First, it was a shame that the authors did not look at Christchurch separately.    They look at an entity called “Greater Wellington” (Wellington, Upper and Lower Hutt, Porirua and Kapiti) but it is puzzling that they don’t even break Christchurch (an urban area of a similar population) out separately (just bunching it with Nelson, Dunedin and Invercargill), let alone look at a combined “Greater Christchurch” entity (Christchurch, Selwyn and Waimakariri).  At least in recent years, it has only made sense to think of the three TLAs together –  and price behaviour in that area has been distinctively different from, say, the rest of the South Island).

Second, it might have been interesting to see whether the differences between prices in other TLAs and the rest of New Zealand (perhaps excluding Auckland) were similar to, or different than, the pattern we observe in Auckland, including whether those patterns have been changing over the relatively short period under study.

Third, I would go easy on the word “unprecedented”.  In a short article, it is stressed several times that the gap between the increase in Auckland house (+land) prices and the increase in prices elsewhere in the country is “unprecedented”.  But they have regional data only since 1981, and in that period there have been a grand total of four upswings.  It is true that there is no precedent in the data, but there isn’t much data.

Relatedly, the article highlights how ill-served New Zealand is with historical economic and financial data.  That isn’t the responsibility of the Reserve Bank –  largely a policy and operational agency –  but it is a limitation that all users face.  Between the continued underinvestment in historical official statistics, and the lack of academic economists working on New Zealand economic history, and doing the leg-work to develop longer-run analytical series (as has been done for house prices in Australia, the US and various other countries), we have a relatively poor sense of what is normal or abnormal.  For example, it would be interested to know whether similar divergences occurred in the early 70s, in the post WW2 house (+land) price boom (when wartime controls were lifted), to be able to see what sort of regional divergences occurred during the Great Depression, and to be able to understand 19th century regional house price shocks (for example, the impact of the gold rushes on Dunedin prices, or of the land wars on Auckland prices).

The subtext in the Reserve Bank article is that what goes up comes down again.  In this article, it isn’t a particularly powerful point (and, in fairness the article doesn’t make much of it directly), since there will always be times when one region or another (even the largest) lags behind house price inflation in the rest of the country.  But there is no natural equilibrium  relationship between Auckland prices and those in the rest of the country –  as experience in the US demonstrates, it is mostly a matter of policy choices.  With weaker immigration and more liberal land-use policy around Auckland, real Auckland house+ land prices could be much lower absolutely, and relative to the rest of the country than they have been in recent years.  Many cities much bigger than Auckland in the US have house prices much cheaper, relative to surrounding areas, than Auckland does (I stayed recently with some friends in a small college city in the middle of the US where house prices were higher than those in the nearest, fairly prosperous, city of a million or more).

Finally, on the housing sector, SNZ released building consent data yesterday, completing the data for 2015.  The authors of the press release pointed out that the number of residential consents was the ninth highest ever –  which isn’t very impressive, since most macro series reach their highest or maybe second highest level each and every year.  What wasn’t pointed out was the way New Zealand’s population has increased –  not only is stock of people living here much higher than it was in previous housing booms, but the population increase over the last year or so has been larger –  even in percentage terms –  than at any time for decades.  Unfortunately, we don’t have an official population series that goes back prior to 1991, but using one of those from the international databases, here are residential building permits per capita since the series the mid 1960s.

residential building permits per capita


Last year’s building permits per capita were only just back to the average for the last 35 or so years –  even though the fast rate of population growth might normally have suggested, in a less regulatorily-impeded market, that consents per capita might have been considerably higher than the average over that period.