The IMF Report: saving and vulnerability

The IMF released its latest Article IV report on New Zealand yesterday.  There are also some background research papers released with the report, and I might come back to look at them later.

There aren’t material surprises in the IMF’s views in the full report, which builds on the Concluding Statement released at the end of the staff mission here last November.  My comments about that statement (here) were fairly critical, noting both the marked change in the messages from one review to the next, and the fairly limited evidence base for many of the policies the Fund was recommended –  quite a few of which were a considerable distance removed from the core business, or expertise (macroeconomic policy and financial stability) of the International Monetary Fund.

Today I wanted to focus just on the Fund’s claim that there is a major policy problem as it affects savings in New Zealand, a proposition on which much of the rest of the report rests.  The Fund talks of a “chronically low national saving” rate, and worries about the vulnerability that allegedly results from a net international investment position (NIIP) of around -65 per cent of GDP.  In the Fund’s Board discussion, we read that “Directors agreed that raising national, and in particular private, saving is critical to reducing external vulnerabilities from the still heavy reliance of offshore funding”.  Note the strong words:  action is “critical”.

I’ve shown a version of the following chart before.  It is common to present charts of net national savings as a per cent of GDP, but to do so involves two errors: first, the numerator is net but the denominator is gross (the difference is depreciation), and second, the numerator refers to savings of New Zealanders and the denominator refers to economic activity occurring within New Zealand, whether owned by foreigners or New Zealanders.  A more conceptually meaningful approach is to do as I do here: compare the net savings of New Zealanders with the net national income of New Zealanders.  Here it is shown all the way back to 1970.

net savings to nni feb 2016

Savings of New Zealanders (public and private combined) as a share of income have been consistently lower than the median of the whole group of OECD countries.  But there is a very diverse range of experiences, and cultures, within the OECD group.  I’ve also shown the comparison with the median of the five other, probably more culturally similar, Anglo countries (US, UK, Canada, Australia, and Ireland).  Over the 45 year period, mostly we don’t look much different from the Anglo median –  we did worse in the years of very large fiscal deficits in the late 70s and early 80s, but other than the experiences are pretty similar on average.     Where is the evidence of a chronic savings problem?  And it is no defence simply to focus on private, or (worse) household savings: first, the boundaries between household and business savings are blurred, and second, the private sector takes account (typically implicitly) the savings behavior of governments over time.  New Zealand governments have done less badly than most for some decades.

The IMF makes no systematic effort to identify reasons why national savings rates might have been as they are.  Instead, they mostly repeat old lines that don’t have much basis to them.  For example, they assert that an overwhelming proportion of household assets are in the form of housing, even though new Reserve Bank estimates –  published almost a year ago –  make clear that that claim was never justified.  After all, relative to population, there is now a consensus that we have too few houses not too many.  The Fund also asserts that there is something wrong with the tax treatment of housing, but appears to make no effort to illustrate, whether in a cross-country or time series context, how that has contributed to national savings behavior.  They urge changes to Kiwisaver and the NZS, but again make little effort to illustrate how policy parameters in these areas explain savings behaviour.  All in all, it seems like a rather weak basis on which to rest quite strong policy recommendations.

The other aspect of this issue which they just seem to take for granted is the alleged vulnerability that the NIIP position gives rise too.  Buried deep in an Annex to the report, they do produce a chart making clear that there has been no worsening trend in the NIIP position for over 25 years –  the negative position tends to widen in boom times and narrow in downturns, but has fluctuated around a pretty stable trend level.  At present, the negative NIIP position is actually below (less negative) than the average since 1988.

The report has no analysis of the nature of the vulnerability that this NIIP position gives rise to –  even though addressing this vulnerability is considered “critical” by staff and Board.  And it gives no example of any country, anywhere, ever,  in which a stable (but quite high) negative NIIP position over 25-30 years has been (causally) followed by a crisis.  I’m pretty sure there are none –  and remind that IMF that for most of its first 100 years, New Zealand’s negative NIIP position was materially larger than it has been over the last 25 years, again without ending in crisis.  There are plenty of cases where a rapid worsening in the NIIP position over a few years led to trouble –  Spain, Ireland, and Greece are three recent advanced country examples –  but that is a very different situation from the New Zealand situation in recent decades.  As has been the case for many years, the IMF simply seems to struggle to come to grips with New Zealand.

Most of the negative NIIP position is represented by (net) banking system borrowing from abroad.  But that creates serious macroeconomic risks only if the assets that are financed by the overseas borrowing are of poor quality.  Often that is the case when foreign debt is rising quickly  –  but that hasn’t been the New Zealand story.  Perhaps the Fund believes that the New Zealand banking system is shaky?   But Directors noted that “the banking system is resilient and well-supervised” –  the resilience conclusion is certainly backed by the Reserve Bank’s own stress tests, which I discussed at length last year.

New Zealand deserves better quality analysis and insights from its membership of the IMF than it has had in the main part of this report.

The report also contains brief sections headed “Authorities’ views” –  the wording of which will have been approved by our Treasury and the Reserve Bank.   I was surprised to find that “the authorities agreed that raising national saving was an important policy objective”, going on to state that the authorities would “consider measures to boost private saving….in the future”.  There isn’t much elaboration of the point, but the statement itself was something of a surprise.  Last I had heard, the Minister of Finance was very sceptical that there was a “national saving problem” in New Zealand, and particularly that there was an issue materially amenable to policy remedies.  If one can’t convincingly identify policy distortions that materially lower national saving rates relative to those in other countries, it is hard to see a good case for policy interventions to encourage people to save when, on their own, they would not.    It would be interesting to know what was behind this latest, apparent, change of view.




National savings

The annual national accounts data were released a few weeks ago by Statistics New Zealand. They got little media attention, which isn’t surprising, but I like fossicking in the spreadsheets. Apart from anything else, they provide an annual update on some of the longest official time series data we have. Australia has full national accounts data back to 1959, and the United States provides official data back to 1929, all on current methodologies. By contrast, we have real quarterly data only back to 1987, and annual nominal national accounts data back to 1972.

The (flow) national savings rate has had a lot of focus in the New Zealand debate over the years. Indeed, early in the term of the current government, there was even an official Savings Working Group. A lot of discussion focuses on household savings, but I prefer to focus on national savings (ie the savings of New Zealanders, New Zealand-owned companies, and the New Zealand government). It provides a good basis for international comparisons, and isn’t messed up by the somewhat-artificial boundaries between households, corporates, and governments.

I also prefer to use net savings data rather than gross savings (the difference is the estimate of depreciation, or “consumption of fixed capital”).  Net savings is the real resources added to wealth.  And if I’m using net savings data I need to use net national income data.

As I highlighted a few weeks ago, our national savings rate has been relatively low by the standards of the typical OECD country. And it is really quite low when compared with the net national savings rate in Australia – but it has been for decades, including the period well before Australia introduced compulsory private superannuation savings. On the other hand, our net savings rate has been strikingly similar to median of the other Anglo countries.

This is what the chart looks like, starting in the year to March 1972, and end in the year to March 2015.

net savings to nni nz
Of course, the sharp fall in the series at the start of the period really catches the eye. But the other thing that strikes me is just how stable average the savings rate has been over the subsequent 40 years, fluctuating around 5 per cent. As you’d expect, it falls quite sharply in recession (see 1991 and 2008/09) – corporate profits tend to fall in recessions, and fiscal deficits widen – but since 1975 there has been no trend in the series at all [1] .

Which creates difficulties for those looking for explanations for our relatively modest national savings rate:
• Some reckon tax incentives might help. But actually we had a very generous tax treatment of superannuation and life insurance until the late 1980s, and a rather ungenerous one (defenders would say “neutral”) since. But the difference isn’t visible in the aggregate data.
• Some reckon a liberal approach to New Zealand Superannuation might explain something. But in the years to March 1975 and 1976 we had a compulsory private scheme, then we had very liberal universal NZS at 60, then we had means-testing and a fairly rapid increase in the age of eligibility. None of it is very evident in the data.
• Some talk about “wealth effects” from rising house prices dampening savings. But the biggest house price bust in modern New Zealand history was after 1974, and the biggest boom was over 2003 to 2007. None of it is very evident in the data.
• The (non-superannuation) welfare state has got bigger over the period, while tertiary education went from being largely “free” for a small group of people, to really rather expensive for a huge number of people. None of it is very evident in the data.
• Some reckon financial liberalisation will have dampened savings, enabling people to bring forward consumption in ways they couldn’t previously. The real freeing-up of the system didn’t start until the mid 1980s. But the difference isn’t obvious in the aggregate data.

• Kiwisaver hasn’t been compulsory, but the take up was sufficiently large that if advocates had been told in advance that it would be that high most would have thought it would have boosted national savings rates. But neither in the more formal research nor in a simple chart like this is it particularly evident.

I’m not suggesting none of these factors made any difference. I’m sure in many cases they did, and (for example) the increase in the NZS eligibility age helped put the government in the position of running large surpluses in the years leading up to the 2008 recession (which was also the peak in the national savings rate). But it isn’t easy to point to a single factor, or even an identifiable set of factors, to explain New Zealanders’ savings choices. An alternative way of saying that is that it is not easy to point to what one might change if one were convinced (which I’m not) that the national savings rate is a policy problem. 40 years of a constant mean is really quite a long time.  More-formal modelling might shed some light, but I wouldn’t be optimistic.

Discussions of savings often focus on households, and then secondarily on the government’s own finances. But they tend to ignore the role of business savings. I’ve wondered whether the modest rate of national savings partly reflects the perceived lack of profitable opportunities in New Zealand. As I’ve pointed out before, business investment as a share of GDP has been quite low in New Zealand for decades, and less than one might expect in a country with quite a fast-growing population (Austria or Belgium need to devote a smaller share of their income each year to adding new shops and offices etc than, say, New Zealand or Australia do). Firms might save more if the growth prospects were better – if, say, real interest rates were nearer those in the rest of the world, and if the real exchange rate had been lower. But in that case, savings rate wouldn’t be the cause of any problems, but just another symptom.

It is one of those areas where better data might help shed a little further light. What was going on with that fall in the national savings rate in 1974/75? It looks a lot like the impact of the collapse in the terms of trade.  But the savings rate has never recovered, and we don’t even know if it was exceptionally high in the early 1970s.  Contemporary estimates suggest that business savings were almost half of private savings – from perhaps a third a decade earlier. Unfortunately, the earlier estimates aren’t compiled on the same basis as the modern national accounts. For what it is worth, here is a chart for the full period since 1954/55, using data published in the New Zealand Official Yearbooks (in this case the 1975 one). There is a hint of national savings rates rising in the late 1960s and early 1970s, but it is hard to know, and hard to know whether the average savings rate for the last 40 years is really lower than it was in the earlier post-war decades.

net savings to nni 2
Surely we should be funding Statistics New Zealand – or at a pinch some good academic researcher – to produce longer backdated series of our national accounts. Better data on its own probably wouldn’t answer all our questions about New Zealand’s longer-term economic performance, but it surely couldn’t hurt. Would it provide value to the plumber from Masterton? Hard to tell, but good data at least opens the possibility of better policy.

NB: Before anyone comments, this post is dealing entirely with the flow rates of savings from current income.  It is not dealing, at all, with stock measures of wealth, or how they might aggregate to some sort of national balance sheet.

[1]  In the years of high inflation and high public debt, the story is a little complicated because much of what is recorded as interest is in effect a principal repayment.  Grant Scobie (and co-authors) looked at that effect here.

Compulsory enrolment in Kiwisaver

I’m puzzled.

Fresh from removing the sign-on bonus for new KiwiSaver members, Bill English is now talking about compulsorily enrolling everyone (well, all employees I assume) in KIwiSaver.  This was National Party policy to do at some stage, when the government’s books were in surplus, but it has now got cheaper to do because people could now be forced to join, at no upfront cost to the government.

It is far from clear what problem the Minister of Finance thinks he would be addressing.   Everyone who has started a new job since 2007 has already been auto-enrolled, and many others have chosen to enrol.  My guess would be that at least half of those who were in labour force in 2007 have changed job since then, and a large number have entered the labour force for the first time. All those people were auto-enrolled.   Most of those who have never joined, or have opted out, or subsequently taken a contribution holiday, have presumably made a considered choice.  Perhaps they can’t afford to be in the scheme now.  Perhaps paying off the mortgage is a higher priority.  Perhaps they are among that large group for whom NZS will already more or less maintain their pre-retirement standard of living.

I had hoped that the government might now leave KiwiSaver to wither on the vine.  After all, there is not much evidence that the scheme so far has boosted national savings, and there isn’t an elderly poverty problem.  Enhancing the retirement income of the relatively comfortable doesn’t seem like an obvious public policy priority.

The Minister did once regard New Zealand’s relatively low rate of national savings as a problem, but Google throws up no references along those lines in the last 3-4 years.  If he no longer regards national savings as a policy problem,  I happen to agree with him (I’ll get to savings as I carry on discussing the reasons behind NZ’s high interest rates).

Surely provision for retirement savings can’t be the problem either?   After all, this government has ruled out any change to the age of eligibility to NZS, and if they have not ruled out lowering the rate or means-testing they have certainly shown no appetite for doing anything.  And, as is well-known, New Zealand’s rate of poverty among the elderly is low, absolutely and by comparison with other countries.  Of course, high house prices may make that picture less rosy in a few decades’ time, but the government knows that freeing-up housing supply is the answer to that one.

And, in any case, although it is proposed to auto-enrol all employees, they could still choose to opt out.  Indeed, given that it is almost exclusively people now over 25 who have not been auto-enrolled already, one might reasonably assume that a large proportion of those who would be compulsorily enrolled if the Minister’s proposal went ahead might choose to opt out.  Yes, I know all the “nudge” literature, but recall that this experiment would not be on the population as a whole, but on a self-selected group of whom many had already deliberately chosen not to join.

And if income adequacy in retirement were really the concern, surely (a) compulsory contributory membership (not just initial enrolment), and (b) something that encompassed the whole population (business owners, welfare beneficiaries, stay-at-home parents) not just employees, would be the way to go.  But, fortunately, that doesn’t seem in prospect.

The National Party’s website says that it believes as follows:

The National Party seeks a safe, prosperous, and successful New Zealand that creates opportunities for all New Zealanders to reach their personal goals and dreams.

We believe this will be achieved by building a society based on the following values:

  • Loyalty to our country, its democratic principles, and our Sovereign as Head of State
  • National and personal security
  • Equal citizenship and equal opportunity
  • Individual freedom and choice
  • Personal responsibility
  • Competitive enterprise and reward for achievement
  • Limited government
  • Strong families and caring communities
  • Sustainable development of our environment

At least three of those values –  “individual freedom and choice”, “personal responsibility”, and “limited government” look inconsistent with auto-enrolling everyone in KiwiSaver.  Perhaps they might justify it under “personal security”, but I’d assumed that has to do with crime, and anyway, as noted already, elderly poverty just is not a major problem in New Zealand.

Compulsory enrolment of all employees in KiwiSaver smacks of something from The Treasury, and its “living standards framework”. But last time I looked, that framework put no independent value on things like individual freedom and choice, except insofar as they served some other end.

The Minister also appears to be using the dubious argument that Kiwisaver is an excellent investment.  If it were really so, you would think that after eight years most people might have got the idea.  But in fact, it is true only on very shaky assumptions.  Certainly, for the time being there is an annual government subsidy of up to about $500.    That is worth having, but it is hardly transformative.  In fact, for most people with a mortgage and contributing to Kiwisaver, it won’t even cover the tax wedge.  So the Minister’s claim is true only if the employer’s contribution to Kiwisaver would not otherwise be paid to the employee.  In the short-term, for any particular job or employee that might be true –  that (and the subsidies) was the main reason I joined Kiwisaver.  But in the medium to long-term surely the Minister of Finance does not believe that returns to labour will be materially affected by whether people take their income in the form of Kiwisaver contributions or in straight wages and salaries?

So New Zealand

  • Does not have a national savings policy problem (although the national savings rate does raise some interesting questions)
  • Has little or no evidence that KiwiSaver so far has made any material difference to national savings anyway
  • Does not have an elderly poverty problem
  • Has a government which is firmly committed to the current NZS system.
  • Has most people already in KiwiSaver, while  those in the workforce who aren’t in KiwiSaver will already mostly have made an active choice to stay outside.
  • Has a governing party that proclaims commitment to personal responsibility and individual freedom and choice.

And for most people in their middle years, Kiwisaver involves a very nasty tax wedge.

So quite why would the Minister of Finance think it was good policy to compulsorily enrol the rest of the employees in the country in KiwiSaver?