I calculated the change in population using the post enumeration survey. I counted up the births and deaths from the vital statistics from Infoshare. The difference should be international net migration. In the 2001-2006 period there was a difference, but a small one.

In the latest census, the gap is really quite large. Implied net migration is around 500 people per year, compared to the data that is normally used for international migration (7,500 per year).

Does anyone know why the difference is suddenly so large? Or what I am doing wrong?

Also, sorry I have been missing in action from the blog. Work has been crazy and I was writing this book to, um, agitate people.

While there are reasonable and often cited risks in his analysis, the substance is lacking. You will find any number of economists, including in the RBNZ and Treasury, highlighting the risks from high Auckland house prices, high household debt and concentration risk in exports (from our increasing exposure to emerging markets, China in particular).

Here are his 12 reasons and why I think there is reason not to panic:

1)     Property prices have doubled since 2004

In Auckland and Canterbury. They have fallen elsewhere.

2)     New Zealand has the world’s third most overvalued property market

Yes. Auckland is.

3)     New Zealand’s mortgage bubble grew by 165% since 2002.

Bit selective. Household debt to income has actually been going sideways, if a little down, in recent years. Although not paid down the rapid accumulation in the 2000s.

4)     Nearly half of mortgages have floating interest rates.

Actually 73% by value. But you can fix if you want to. Which borrowers have done in the past. This is not to say that rising interest rates wont bite, but they will be spread over a long period of time.

5)     Mortgages account for 60% of banks’ loan portfolios.

I haven’t verified this number, but presumably this is bad if this will lead to high defaults. New Zealand does not have the legal structure to allow borrowers to walk away from their debts. Also, even during the recession of 2008 and the early 1990s mortgage default rates in NZ were relatively small.

6)     Finance, not agriculture, is New Zealand’s largest industry.

Like any advanced economy services are a big part of the economy. This is not surprising. Although he misstates the data. The top 10 industries as per GDP are:

• Information Media and Telecommunications    7%
• Professional, Scientific and Technical Services                   7%
• Owner-Occupied Property Operation (National Accounts Only)                  7%
• Rental, Hiring and Real Estate Services                  6%
• Wholesale Trade              5%
• Construction      5%
• Health Care and Social Assistance           5%
• Retail Trade        5%
• Transport, Postal and Warehousing       5%
• Financial and Insurance Services              5%

Agriculture details are:

Production approach

• Ag 3.2%
• Fishing 0.6%
• Forestry and logging 1.1%
• Food manufacturing 4.6%
• Wood & paper product manufacturing 4.6%
• Total 11.1%

Export approach

Or Ag related exports are 43% of total exports and 14% of expenditure GDP.

Like most advanced economies the services sector is a large share of the economy. Ag, forestry and fishing is around 7% of GDP as at 2010, compared to the OECD average of around 2%. Finance & insurance directly account for around 10%, while the OECD average is around 6%. Most debt in NZ is intermediated by the banking sector (smaller equity market etc). Details available at oecd.stat

7)     New Zealand’s banks are exposed to Australia’s bubble

Not really. Banking regulation in NZ separates our banks from direct exposure. Although our cost of funding may rise if tarnished with a Aussie housing bust. Real economy threats too from a recession in our second biggest trading partner.

8)     Australian and Chinese buyers are inflating the property bubble

Really? I still haven’t seen evidence of this. So cant comment.

9)     New Zealand has a household debt problem

Same as number 4.

10)  Government overseas debt has nearly tripled since 2008

Government net debt to GDP is less than 30% of GDP and denominated in NZD. Whats the issue?

11)  The New Zealand dollar is overvalued

NZD is high relative to history. Export share of GDP is at a historically high level. A fall in the NZD would spur exports and reduce imports.

Here is what to expect when New Zealand’s economic bubble truly pops:

The property bubble will pop.

Sure in Auckland.

Banks will experience losses on their mortgage portfolios.

Like they did in the GFC? More than half of the housing market in NZ crashed in the GFC and bad debts peaked at a very low amount. Don’t buy this argument.

The country’s credit boom will turn into a bust.

We haven’t had this already?

Over-leveraged consumers will default on their debts.

Why? Its not America. You cant walk away from your debts. This did not happen in the GFC.

Stock and bond prices will fall; the New Zealand dollar may weaken.

Good.

Economic growth will go into reverse.

Ok. Thats what happens in a recession.

Unemployment will rise.

Ok. Thats waht happens in a recession.

“Could it be that Brisbane feels more at like home for people from the rural North Island and South Auckland than Southland or Auckland City do?”

It got me thinking about why people migrate.

I am currently reading “Exodus” by Paul Collier. (I am only a few chapters in, but it is a good read.) He writes that there are three big drivers of migration. I am paraphrasing from page 38:

1. It is an economic response to the gap in income. All else equal, the higher the income gap, the stronger the pressure to migrate.
2. There a number of impediments to migration (economic, legal and social). Costs must be borne before benefits can be reaped. The poor are least able to meet the cost of investment and act as a counter to 1.
3. Costs of migration are reduced by the presence to a diaspora.

Using this model, it seems pretty clear why New Zealanders choose to migrate to Australia. They reinforce each other:

1. Australia’s wages are typically around 30% higher than in New Zealand. So the income gap encourages migration.
2. The impediments to moving to Australia are relatively low. While not eligible for social welfare, New Zealanders are able to work and live in Australia with few legal impediments. The cultural and social gaps are also rather small – even if it is made out to be rather wide. Speaking from person experience – and this is not mean to offend anyone – I found living in Melbourne not too dissimilar to living in Wellington (in a more grown up way). The actual cost of flying there isn’t high either.
3. New Zealand has a large diaspora in Australia. According to the Australian Census, around 500,000 New Zealand citizen live there.  The most popular are QLD (40%), NSW (24%), VIC (17%) and WA (15%). The New Zealand diaspora encourages migration.

The framework in Paul Collier’s book suggests that migration to Australia will accelerate over time, because the key drivers are reinforcing each other.

I had a quick read through the first one, which is on “Recent Unemployment Experience in New Zealand”

It’s an interesting paper and worth a read. But they reach a surprisingly strong conclusion, where I think a more nuanced interpretation is required:

“… the rate of unemployment can be explained entirely by economic growth outcomes, and do not seem to reflect any structural change in the labour market. This suggests that there are not any impediments to the rate of unemployment falling back to levels that existed in the mid- 2000s.”

I am a little surprised at this rather strong conclusion, as this paper did not explore the reasons for job losses.

The 2013 Census shows that jobs losses have been widespread across regions and industries. This was a combination of:

• structural (eg hollowing out of manufacturing, some of which may never come back)
• cyclical (eg construction downturn, which is likely to come back).

The recovery since then has also been uneven (services and metro city dominated). This could indicate some difficulty in getting the unemployment rate back to the lows in 2000s – especially as those who lost jobs may not have the skill set to match the skills for the new jobs created in the recovery. If you are applying for unemployment you can check out the virginia unemployment commission here.

I would suggest the Working Paper is a good start to look at the evolution of the unemployment rate. But they need to do work on understanding the job losses, and which are likely to come back and what skills may be redeployed elsewhere to reach a convincing conclusion.

I am a big fan of these Working Papers and really enjoy reading them. This is not a criticism of the work, rather a suggestion to extend it.

A did a quick decomposition of employment data by age group from the Census. It shows that the employment rate, employed share of the working age population, has fallen from 65.0% to 62.3%. This 2.7%point decline can be roughly broken down to:

• roughly 1/3 is from the age composition effect (1.0%pt)
• remaining 2/3 is within age group and other factors. (1.8%pt)

Looking within the age groups it is clear to see there are large changes within age groups. Older people are still less likely to work than young, but the current generation of oldies are staying in employment for longer.

Between 2006 and 2013, the employment rate rose for only those aged between 60-84. The rest were worse off.

At some stage I got cornered on what would happen to Auckland house prices. Having no good answer I resorted to the Keynes line that markets can remain irrational for longer than you can remain solvent.

This morning I thought I would have answered it more as The Economist writes:

“Manias can last much longer than investors think, as many contrarians discovered to their cost during the dotcom boom of the late 1990s. Nor do investors know whether a bubble will be resolved through a sharp fall in prices or a long period of stasis, in which inflation erodes prices in real terms.”

And the folly of forecasting asset prices is neatly encapsulated by Irving Fisher (Professor of Economics at Yale University), in 17 October 1929, soon before the crash:

“Stocks have reached what looks like a permanently high plateau.”

Intercensal employment growth (compounding rate) in the Census and HLFS aren’t the same, but the difference looks awfully big this time around (perhaps because the Census was delayed by two years due to the Christchurch earthquakes).

A rough and ready ‘taper’ to make the HLFS and Census consistent could mean that the level of employment looks very different after revisions next year. The regional estimates will also change considerably.

I am reiterating from yesterday, but comprehensive data  from the Census give us a better understanding of the economy and rebase surveys and other measures to reduce errors. Any attempts to discontinue the Census should carefully consider the many costs of not doing it.

The reason for putting the policy in place is financial stability. That is to reduce the accumulation of high-risk debt in the banking system. By this test, and this should be the reasonable test, this is bad policy.

By this exemption the RBNZ is saying that it is less risky to borrow to build a new house than to buy an existing house. I disagree that new house prices move less than existing house prices. So, the RBNZ is now exercising a policy of exclusion – against those high LVR borrowers who want to buy existing homes.

The justification that we need to build more houses, surely is retorted with, is this the right policy to address that problem? It also raises the question of why was this not a consideration before the LVR policy was implemented? What did the RBNZ actually know about how high LVR loans were being used, by whom and where?

By increasing highly leveraged credit flow to new builds will not solve issues to slow land release and planning restrictions.

By seemingly bending to industry and political pressure, the RBNZ has tarnished its shield of independence. I am fearful of ongoing lobbying and political interference it invites heading into the 2014 election.

I assumed a NAIRU of 5% for illustration; blue is less than NAIRU and red is higher. It shows the impact of a number of secular forces (ageing, urbanisation, contrasting industry performance, etc). The unemployment rate is generally lower in the South Island than in the North Island – with labour in relative scarcity and better utilised in the South Island compared to the North Island.

Given this information is not possible to glean from the Household Labour Force Survey (HLFS), it reinforces the need to do the Census at regular intervals (in addition to some surprises on population size, location, age, etc).

I was part of a panel that spoke briefly at the launch. Donal at Economics New Zealand has written about it already.

The question posed to me was: Are we doing better? It’s a loaded question. It depends on who you ask and what dimension you measure.

Putting it in the context of the economic cycle, we can describe a recession where economic activity, employment and various other indicators fell. They have subsequently recovered. But the recession and the recovery were shared unevenly, across regions, industries, age, ethnicity etc:

• A young, unskilled, Maori or Pasifika solo parent from the provinces probably felt the recession harshly.
• A middle aged, Pakeha male in a professional job in Auckland probably barely felt the recession.

Looking at aggregate measures of economic data is very helpful to describe general trends in the economy, but it also behoves us – as economists – to look at the implications for real individuals beneath the surface.

Macro data can help the focus our attention. Atomised and distribution data provide additional powerful information to properly define a problem, formulate the likely solution set and, ultimately, measure the impact of the chosen policy intervention(s).

Preferences matter

But at the same time, relying entirely on one lens to make policy is dangerous. What should the policy intervention be for the following two hypothetical people?

1. Chooses to move from a well-paid job Auckland to Wanganui. Then is surprised to find lower wages and fewer employment opportunities.
2. A young solo parent with limited skills in Wanganui. Who wants the opportunity to access education and move to Auckland for a job.

Many characteristics may look the same between these two people, but one is an issue of preference and the other is about persistent poverty and access to opportunity – they are not the same. Based on my values, I feel no moral duty to help person 1, but I see every reason to help person 2.

So, are we doing better?

I strayed far from the question at hand. Are we doing better? It depends on who you ask and how you measure it.

At the aggregate level we are doing better. Social indicators enrich our understanding of the economy, but they are not without their weaknesses. The dispersion of outcomes in the GSS reinforces that the picture for the anonymous average is not the same as the varied experiences of the individuals who constitute society.