RBNZ is still capable of producing worthwhile research

I’m so used to seeing sloppy and ill-thought-out analysis coming out of the Reserve Bank these days that I confess when a paper on housing landed in April, I went looking for faults.

Like the time when RBNZ said it needed to focus on climate change as part of its role in ensuring financial stability because mortgages are of such long duration.

But its own data showed mortgages reprice in about three or four years while insurance reprices every year.

And RBNZ tried its darndest to suppress a paper one of its staff wrote in 2018 that found climate change posed little threat to financial stability because of such repricing.

And another paper on house price inflation a while back didn't mention the impact of monetary policy.

Because among the myriad factors that influence house prices, interest rate levels are one of, if not the major influence.

As is employment – even if house prices fall so much that a recent home buyer is staring at negative equity, if they still have a job and income, the means to pay their mortgage, they usually sweat it out until house prices recover.

The April paper by Andrew Coleman was of much higher calibre and focused squarely on the role of interest rates, as well as many other factors.

Tax effect

As some solid work by Dominick Stephens , now deputy Treasury secretary, showed, changes in the tax treatment of housing can also have profound impacts.

Raising the top rate of income tax tends to encourage investment in property, although other tax changes in recent years have reduced the advantage.

Immigration can be another – its recent surge has probably contributed to the earlier-than-expected bottoming of house prices, which are now starting to rise again, particularly in Auckland, the largest magnet for new migrants.

Housing is such a fundamental underpinning of household wealth and the economy that it isn't surprising economists spend so much time on trying to understand what drives changes in values.

The gold standard measure is Quotable Value's house price index, but because it uses house settlement data, which are registered months after actual transactions, it is mainly useful for historical analysis, and not at all useful for showing what's happening currently.

QV has tried to get around that by preparing a three-month rolling average index and has done its best to convince people it's timely.

But while it will announce an index number for a given month, for example the July number will actually be measuring transactions that mostly occurred between March and May.

REINZ numbers

The Real Estate Institute, whose data measures unconditional sale and purchase agreements, still emphasises its median house price – that used to be the most current data available, but now it's like choosing to use a sledge hammer when a much better precision tool is now available.

RBNZ worked with the institute to develop that organisation's house price index and then later refined it to take into account the array of vastly different types and sizes of dwellings – comparing a one-bedroom apartment to a five-bedroom house isn't very useful on its own – which is now the best available indicator of what's happening in the housing market now.

Much of the political and prudential focus has been on the bottom end of the quality spectrum and on the cheapest available houses, because that's the part of the market in which both investors and first-home buyers tend to be found.

Indeed, the government's modification of its remit for the monetary policy committee issued in March 2021 specifically stated government policy was “to support sustainable house prices, including by dampening investor demand for existing housing stock, which would improve affordability for first-home buyers.”

That's where the government's removal of mortgage interest deductibility for landlords and the RBNZ's imposition on property investors of the greatest loan-to-valuation restrictions for bank lending came from.

Andrew Coleman's paper

Coleman's April paper adds a further dimension to our understanding of what drives house prices, focusing on the links between a structural fall in interest rates, the desire of people to improve the quality of their homes and the impact this has had on house prices.

As Coleman notes, as people's affluence changes – either through higher incomes or lower interest costs, or both – they tend to want a nicer house, either through renovating their existing house or moving to a better one or a more desirable location.

“If many households want to upgrade at the same time, however, it requires a change in the entire quality profile of the housing stock,” his paper says.

But it takes time to build sufficiently higher quality houses or to renovate existing stock to meet that demand, meaning that house prices can remain unsustainably high for very long periods until construction catches up.

This is no simplistic analysis. It considers everything from location, land availability, geographic restrictions, cost, efficiency and availability of transport, degree of population density, distance from amenities and regulations, such as density rules, that go into determining the quality of housing beyond its physical features.

It throws up a few gems of information, such as the fact that in 1999, Auckland and Christchurch had the highest amount of parking per person in the world. (Being one of the few people my age who never learned, or wanted to drive, I have a particular bugbear about the amount of space the country devotes to cars and parking.)

Quantifying the quality impact

Coleman concludes that between 30% and 40% of demand for housing between 2000 and 2020 can be attributed to the demand for better quality homes and that improvements in quality may have accounted for about half the new housing construction over that period.

As he notes, “builders tend to construct high quality houses before low quality houses as they are more profitable.”

The average size of newly constructed houses has been rising significantly from 111 square metres in the 1970s to 184sq in the 2000s and the number of houses larger than 200sm has jumped from 23% of new builds in 1991 to 43% between 2007 and 2014.

But most of Coleman's analysis covered the period since 1990 and I was curious as to why.

“1990 is a natural choice to split the analysis of the NZ housing market and the NZ economy more generally as it is close to the break-point between two eras,” he explained.

A number of massive economic changes that occurred in the 1980s, such as the introduction of GST and other tax changes, the floating exchange rate, and the passing of the Reserve Bank of New Zealand Act in 1989, which both granted it independence and charged the governor, then Don Brash, with a sole focus on quelling then rampant inflation.

Is it over now?

That started a long-term trend of structurally lower interest rates – in its latest monetary policy statement, RBNZ raised its assessment of a neutral official cash rate from 2% to 2.25%, suggesting that trend may now be over.

Coleman says unemployment peaked in the early 1990s and the reforms of the previous decade took effect, “in combination with international trends that saw manufacturing shrink substantially in most countries, subsequently real wage growth started to increase.”

Pile onto that changes in the availability of mortgage credit which “culminated in the widespread availability of modern, flexible mortgages in the early 1990s.”

Earlier research Coleman had done showed all these forces meant a significant change in house prices from about 1990.

Between 1975 and 1990, there had been a real – after inflation – annual decline in house prices of 0.8%. Between 1990 and 2000, the annual real increase in house prices was 2.5% and that accelerated to 4.2% in the years through to 2014.

“International comparisons using data from the Federal Reserve Bank of Dallas suggest that NZ had the third-slowest increase in real (inflation-adjusted) house prices in the OECD from 1975 to 1990 but the fastest increase from 1990 to 2020. This further suggests something significant changed in NZ in 1990,” he says.

Note: In case you're wondering about the timing of this column, I had planned to base a column on Coleman's paper shortly after it was published and while I was still at BusinessDesk, but my departure turned out to be more precipitous than I'd expected. And then I had to get through a three-month restraint of trade period. But I thought it deserved to be highlighted.

Correction: The sentence under the Is it over now? subhead has been corrected to show RBNZ estimates the neutral OCR is now 2.25%.