Some months ago I wrote about the NZIER’s Shadow Board, a panel of expert and informed observers who are each asked prior to each OCR review to provide a probability distribution as to what OCR is “appropriate for the economy”.  It isn’t quite the same job as the Reserve Bank has –  the Bank has to follow the PTA, and in principle the panel members might, say, agree with the two NZIER economists who recently argued for nominal GDP targeting.

There isn’t usually much information in the results of the Shadow Board exercise.  They usually track remarkably closely with the Governor’s own choice about the OCR, so they are right or wrong about as often as the Governor is.

That pattern continues this month.  The mean expectation across the respondents has dropped a little, from 2.70 per cent in October to 2.67 per cent  this time.  Whichever way the Governor goes tomorrow the Shadow Board will have been close.

I’ve been more struck by the lack of much diversity in the views (across panel members) and the high degree of confidence with which each panel members appears to hold his view.  This time the lower quartile expectation is 2.5 per cent and the upper quartile is 2.75 per cent.  According to these respondents there is only a 12 per cent chance that something other than 2.5 or 2.75 is the right OCR for the economy.

I don’t really understand how anyone can be that certain, given the uncertainties about the current state of the economy, the future, and about the connections between real activity and inflation (or other nominal variables monetary policy can target).

I told an interviewer this morning that I thought the economy would be better off with the OCR at 1.75 per cent than at 2.75 per cent.  I hadn’t distributed my probabilities then, but I’ve put them in the chart below (and compared them to those of the Shadow Board).  My numbers effectively say that I think there is roughly a 30 per cent chance that the consensus view is correct.

It would be interesting to know the probabilities the Governor and his chief advisers would assign.  We’ll know the mean tomorrow, but almost certainly will learn little or nothing about the distribution.  Having information of that sort –  whether as a table like the Shadow Board provides, or as fan charts  –  would provide useful information on how these senior officials think about the economy.

Tempting as it is to write about the Ombudsman’s weak report on the OIA, or to carry on looking at the Wellington airport proposal (taxpayer subsidies for long haul holidays for Wellingtonians, the return of a Think Big mentality, as Brian Easton suggested here), it is time to get back to macro.

Tomorrow is the final Reserve Bank Monetary Policy Statement for the year.  The focus, of course, is on whether or not the Governor chooses to cut the OCR, but it is worth briefly looking at what Parliament requires from the Reserve Bank in its MPSs.  That is set out in section 15(2) of the Reserve Bank Act.

The policy statement shall be signed by the Governor and shall—

(a) specify the policies and means by which the Bank intends to achieve the policy targets fixed under section 9:

(b) state the reasons for adopting those policies and means:

(c) contain a statement of how the Bank proposes monetary policy might be formulated and implemented during the next 5 years:

(d) contain a review and assessment of the implementation by the Bank of monetary policy during the period to which the preceding policy statement relates.

As I’ve noted previously, this bit of the legislation needs updating.  But it is the law, and it isn’t typically followed very closely by the Bank.  MPSs tend to be full of data analysis –  important and sometimes interesting –  but light on policy.  There is never much, for example, on the reasons why the Reserve Bank is adopting one particular approach to policy rather than another.   And when there is such discussion it is never very serious –  it is almost always a caricatured or straw man alternative  Scrutiny and accountability involves, in part, the ability to assure ourselves that powerful policy officials have thought seriously about the alternatives.  For example, at the start of last year, the alternative of not raising the OCR.

MPSs never look five years ahead, but then there isn’t much they can usefully say about such a horizon.  Whenever core inflation gets back to target they should envisage keeping it there, absent the unforeseen and unforeseeable shocks.  As I said, the legislation needs amending.

And MPSs don’t often contain much in the way of serious review and self-critical assessment of past policy.  In one respect that is understandable.  Bureaucratic incentives don’t encourage open self-scrutiny. But that is precisely why Parliament puts such provisions in legislation, to lean against the natural self-protective tendencies of powerful agencies.  We’ve heard defensive lines from the Governor –  “of course I was right” –  but little to give us (public, Board, FEC) much confidence that the Bank has really thought hard about its monetary policy performance in the last few years.   Given the target Parliament set for the Governor, it looks as though mistakes have been made.  Contra Donal Curtin, it isn’t a “gotcha culture” to suggest as much, simply a recognition of the difficulty of doing monetary policy consistently well.

But what of tomorrow?    Most economists apparently expect the Governor to cut the OCR, even if market pricing is less certain.  I don’t claim any insight into what he will do, but I am clear that the OCR should be cut.  And that the likelihood of further cuts next year should be foreshadowed.

Liam Dann had a piece in the Herald the other day looking at the pros and cons of cutting now.  I was a bit surprised to see him listing the prospect of a drought this summer as a reason to cut now.  I disagree.  Apart from anything else, it is too late to make any material difference: the biggest impact of further OCR cuts now will be seen next summer, and I doubt anyone has any particular insights on what the weather will be like then (or even on what the after-effects, eg on pasture or stock condition, of this year’s drought might be).

But also, even if a drought were to temporarily knock real GDP, it is not clear it would make much difference to incomes (nominal GDP) –  Reserve Bank research done at the time of the last drought found, somewhat surprisingly, that New Zealand droughts tend to raise world diary prices.  If so,. some people will be worse off, but some will probably be better off.  Monetary policy can’t usefully deal with distributional issues.

And, of course, the fall in real GDP associated with a drought is also a temporary reduction in the supply capacity of the New Zealand economy, so it doesn’t have very obvious implications for the level of excess capacity or of inflation pressures.

Droughts can be nasty things for rural producers, but mostly they are best looked through by monetary policy makers.

I was also a bit surprised by one of the items on Dann’s lists of reasons to hold.  This was the suggestion that the economy had been doing quite well, in maintaining real GDP growth of around 2 per cent, perhaps even picking up to around 3 per cent next year.  I highlight this not to pick on Dann, but because it is such a common line.  People seem to lose sight of the fact that Statistics New Zealand estimates that New Zealand’s population grew by 1.95 per cent in the year to September, faster than almost any advanced country.  In other words, 2 per cent real GDP growth would represent no per capita growth at all. Perhaps we are living in an age of diminished expectations, but since when was almost zero per capita growth a mark of a reasonably well-performing economy?

As a reminder, the unemployment rate has risen back to 6 per cent this year.  With inflation as low as it has been, monetary policy should have been set looser.  If it had been not only would inflation have been a bit higher, but the economy would have been growing rather faster.  That would have been a good thing, not a bad one.  Over the 17 years 1992 to 2008, real GDP growth averaged 0.8 per cent per quarter.  Since then it has averaged 0.5 per cent.    Monetary policy can’t materially influence the longer-term structural performance of the economy, but when inflation is so low we should be expecting to see growth rates materially above potential.  Monetary policy choices have meant that hasn’t happened.

Here is another way of looking at the disappointing performance of the economy in the last few years. The trend line is the path the economy would have followed if it had sustained the average growth rates of 1992 to 2008, and the red line is the actual path.  The gap between the two is now equivalent to around 15 per cent missing GDP.

There is also the argument that the OCR should not be cut because of property price inflation in Auckland (perhaps Hamilton and Tauranga too).  Reasonable people can debate the merits of whether house prices (or perhaps credit growth) should be part of the goal for monetary policy.  But they are not at present.  And the job of the Governor is to implement policy in accordance with the Policy Targets Agreement.

My view is that house prices should generally not be part of what monetary policy is targeting.  High house prices, particularly in a single city (even the largest) are largely a real relative price phenomenon –  in this case, the interaction of supply restrictions and policy-induced population growth.  Monetary policy is singularly badly suited to trying to deal with uncomfortable relative price movements –  doing so involves throwing the whole rest of the economy round to make up for some other microeconomic policy failures.  We should always be vigilant about the possibility of emerging financial stability threats, while being modest about how much we (or the Reserve Bank) really know about those risks.  But if policy responses are needed to contain those risks, there are perfectly good conventional options open to the Reserve Bank –  increasing the risk weights on housing loans and/or increasing overall capital requirements.  Such approaches are effective in limiting the potential damage to the financial system if things go wrong, while imposing minimum distortionary effects on the rest of the economy now.   And, of course, while its gets boring to say so, monetary policy works in part by lifting the demand for, and price of, long-lived assets.

A final line I’ve seen repeated several times in recent days is the suggestion that in any case monetary policy can’t do much to raise inflation.  I’ve not seen any very serious analysis or argumentation in support of that view, and it seems to simply to reflect the fact that inflation is low relative to target in much of the advanced world.  But in most of the advanced world, the scope of conventional monetary policy to do more has been long since exhausted –  interest rates have been at or near zero for years.  Neither we, nor Australia, are in that position.  We can cut the official interest rate.  And even if some are sceptical that lower domestic interest rates will do much to boost domestic demand, lower interest rates would almost certainly lower the exchange rate.  A lower exchange rate will, all else equal, boost domestic prices to some extent. And, more importantly, it will over time encourage more investment, production and employment in the tradables sector of the economy.  There is no reason to believe that something closer to full employment of domestic resources would not tend to lift core inflation.

As I’ve noted repeatedly, core inflation has been below the target midpoint –  the number explicitly highlighted in the PTA –  for several years.  There is no sign that anyone really fully understands quite why or (hence) that the Reserve Bank has been able to adequately correct its models and forecasts to avoid a repetition of this outcome in the future.   Given that, it would be prudent for the Reserve Bank to be acting now in a way that it believes would actually deliver core inflation in the upper part of the target range.  Act on the basis of forecasts that you think will deliver, say, 2.5 per cent core inflation 18 months hence.  If the Bank did that, they might be right and core inflation might end up higher than the midpoint.    If so, as it became more certain that was the case, there would be plenty of time to tighten gradually.  But if the forecasting  (or understanding) errors of the last six years continue, it is likely that core inflation would end up somewhere near 2 per cent.  The latter would be an unambiguously good outcome –  good in its own terms, and also good for the unnecessarily unemployed.

At present, by contrast, and unless the MPS tomorrow reveals some startling new analysis, we have simply to take on faith the Bank’s view that the current approach to policy and forecasting  is enough to get back to 2 per cent, even though it has not been enough for years now.  With the zero-bound no longer that far away, and with nothing in the domestic or global environment suggested any sustained acceleration in growth or inflation pressures any time soon, it is an approach that should have been seriously considered.  And if it isn’t adopted, perhaps in scrutinising the MPS tomorrow, FEC members might ask the Governor about the reasons he has chosen to adopt one policy approach over the other.

I urged again the other day that there should be an open process of research and debate leading towards the negotiation of the next Policy Targets Agreement in 2017.  These documents matter.  Monetary policy is the main tool for short-term macroeconomic stabilisation, so the PTA sets the “rule” (well, loose guide) for how the short-term fluctuations in the economy will be managed.  The Reserve Bank –  and the Treasury and Minister –  has often had a tendency to treat deliberations around the PTA as technocratic in nature (which in some ways they are), and hence not something with which to trouble the natives.  The standard Reserve Bank response to any suggestion of greater openness was “but we already tell them what we want them to know”.  But open government is not just about releasing finished products, after the event, in bureaucratically-approved formats.

Two other former Reserve Bank staff, Kirdan Lees and Christina Leung, both now at NZIER, have made a useful contribution to a debate about the future of New Zealand’s monetary policy.  They put out a note the other day headed Time to reassess inflation targeting, which concludes with a pretty strong leaning towards adopting nominal income targeting instead.  I don’t think they will get far with the current Governor on that one – he used to bristle and react very frostily whenever anyone so much as mentioned nominal income targeting  – but he won’t be Governor for ever, in the end the Minister of Finance calls the shots, and whether the Governor likes debate or not, it is an important part of good public policy processes.

However, I’m not convinced by the Lees/Leung argument.  In particular, I’m not persuaded that the form of the rule makes a great deal of difference to assessing the appropriate stance of monetary policy now.  Nor am I convinced it would have made a great deal of practical difference over the pre-recession years.  And if we were going to move away from inflation targeting, I’m not convinced that nominal income targeting is the alternative I would adopt.

Lees/Leung have a number of strands to their argument.

First, they argue that “supply shocks” have become more important relative to “demand shocks”.  Perhaps, but where is the evidence for that proposition in New Zealand or in other countries?  They seem, in part, to be arguing from the presence of a number of phenomena (fracking, the internet etc) which are improving productivity.  All of them are real, but in aggregate productivity growth has been materially slower in the last half dozen or so years than in the previous decade.  And, in any case, the issue for monetary policy would not normally be the trend rate of productivity growth, but shocks –  surprises, which can go either way.   There is, of course, one area where supply shocks have become more important for New Zealand –  terms of trade volatility has been much greater in the last decade than in the previous 15-20 years (apparently driven mostly by the fairly extreme dairy price volatility).  We’ll come back to terms of trade shocks.

Second, they point out that many advanced countries have seen inflation undershoot respective targets.  That is, of course, true, but most of the countries on their chart have largely exhausted the potential of conventional monetary policy.  Interest rates are basically at zero, and have typically been so for quite a few years.  There are reasonable arguments that a different target might make it a little easier to get out of the current “trap”, but they aren’t relevant to New Zealand at present.  Our Reserve Bank has undershot the inflation target not because it couldn’t cut interest rates enough, but because it chose not to.  That failure probably wasn’t wilful –  largely it was because they misread the data.  They (and other central banks) misread the data on the other side during the boom years.  Forecasting is difficult, but it is a problem that bedevils any of the rules under discussion.

Third, they point out the well-known proposition that, in principle, nominal GDP targeting can generate better short-term macroeconomic performance (eg less output variability) in the presence of supply shocks.  In the example they cite, faced with drought, an inflation targeting central bank will tend not to adjust policy (since inflation, and especially core inflation, won’t change much) while a nominal GDP targeting central bank will tend to ease policy to lean against the drought-induced fall in GDP.  But, in fact, whichever rule was adopted, the central bank would almost certainly be reacting to forecasts (whether of inflation or nominal GDP), since monetary policy only works with a lag.  Droughts typically aren’t recognised by central banks until we are in the midst of them, and when they are recognised they are typically assumed to be shortlived.  Faced with the prospect of a drought this summer, the Reserve Bank will typically (and reasonably) assume that next summer will be normal, and since monetary works with a lag they wouldn’t change policy under either regime.

Fourth, they argue that the difference between inflation targeting and nominal GDP targeting is quite material for where the OCR should be set right now.    It is certainly feasible that in some circumstances there could be quite a difference, but they don’t make a persuasive case that this is one of those times.

They compare inflation rate targeting with nominal GDP level targeting.    Either prices or nominal GDP can be targeted in rate of change terms (inflation rates) or in levels terms.  No country in modern times has adopted levels targets for either prices or nominal GDP.  The Bank of Canada looked quite carefully at the option of price level targeting a few years ago, and concluded that it would not represent an improvement over inflation targeting.  One reason levels target don’t appeal to practical policymakers is that if one makes a mistake and prices or nominal GDP rise unexpectedly strongly, one can’t just treat bygones as bygones –  one has to tighten to drive the level of prices (or nominal income) back down again.    Whatever the theoretical appeal of such an approach, it seems unlikely to command much public enthusiasm or support –  and hence seems unlikely to prove durable.

Much of the older literature around nominal GDP targeting was done in terms of rates of change (nominal GDP growth rates).  But since the 2008/09 recession there has been renewed interest in the idea of a level target for nominal GDP.  The argument made, most prominently by US economist Scott Sumner, has been that a target for the level of nominal GDP would have (a) prompted an earlier easing in monetary policy, and (b) would underpin expectations (especially in the US and Europe) that interest rates would stay low for a long time.

Lees/Leung acknowledge that the current inflation targeting framework invites further cuts in the OCR  (we’ll see next week whether the Governor agrees, although recall that it is a forecast-based framework, so OCR cuts aren’t warranted if the Bank can convincingly show that core inflation is heading back to 2 per cent reasonably soon on current policy).  But then they suggest that using nominal GDP levels targeting “interest rates are about right”.

They appear to base that observation on this “illustrative example”.

In this chart, they appear to have simply drawn a trend line through actual nominal GDP since 1998 and then calculated the difference between the trend line and actual. That difference is small.

But to adopt a nominal GDP levels target, one would need to define an appropriate trend period first.  And it isn’t clear to me why this is the right one.  Most advocates of nominal income targeting at present argue for using something like the pre-recession trend (since the arguments are about whether policy has been sufficiently loose in the year since 2008).  In a New Zealand context, in both 1996 and 2002 policymakers decided that New Zealand should have a faster trend rate of nominal GDP growth (since they revised up the inflation target).  Alternatively, a common approach in New Zealand has been to look at the entire period since low inflation (and lowish nominal income growth) was established, around 1992.

I’m not sure that a trend starting from 2002 to, say, 2008, is that enlightening.  After all, there was a common view that monetary policy was too loose over at least several years of that period (Alan Bollard has openly acknowledged as much).  But if we used that as the trend, this is what the picture looks like (using logged data).

Nominal GDP is well below that pre-recession trend (as it is in most countries), arguing for looser monetary policy now as well.

Or we could use a trend done over 1992 to 2008 and one ends with a similar gap.

Levels targeting does require identifying a starting level (which is neither easy nor uncontentious).  But what if we just look at nominal GDP growth rates?

Not only has nominal GDP growth averaged far lower since 2008 than it did over the previous 17 years, but the most recent observation (annual growth of 3.9 per cent) is right on the average for the post-2008 period.    If we are happy with something like 2 per cent inflation (few have argued for lowering the target) and have a population growth rate of almost 2 per cent per annum, then 5 per cent nominal GDP growth might be a reasonable benchmark.  Current nominal GDP growth is well below that, just as current inflation (headline or core) is well below the 2 per cent inflation target.

So, shifting between CPI or nominal GDP based rules, levels or rates of change, looks as though it would not make much difference to how one thinks about appropriate monetary policy at present, at least on the current data.

But as I noted earlier, central banks aim to base policy on forecasts, so the issue is not so much where inflation or nominal GDP is right now, but where the central bank thinks it will be in a year or two’s time.  My proposition is that most of the mistakes central banks have made in the last decade or two have been forecasting mistakes rather than policy rule mistakes.  Monetary policy wasn’t tightened soon enough during the boom years partly because Alan Bollard was a dove, but partly because the Bank –  and most other forecasters even more so –  recognised the immediate inflation pressures, but forecast that they would soon dissipate.  They were wrong, and as a result inflation and nominal GDP growth were higher than forecast.  Similarly in the last few years, central banks have underestimated how weak both inflation and nominal GDP growth have been.  If one could forecast nominal GDP more reliably than inflation, perhaps the case for change would be stronger, but outside recessions the big source of fluctuations in New Zealand’s nominal GDP is international commodity prices.  They are highly volatile, and the volatility dominates any trend movements over the sorts of period relevant to monetary policy.

An international conference was held in Wellington a year ago this week to mark 25 years of inflation targeting, and the papers have recently been published.  Several academics presented a paper looking at how inflation targeting compared with nominal GDP targeting for New Zealand.  They looked at a variety of different time periods, including the pre-liberalisation period, the transition to a more liberalised economy, and the current period.  The authors were sympathetic to the case for nominal GDP targeting.  I was asked to be the discussant, bringing a practical policy perspective to bear on the issues raised in the paper.  In my remarks, I set out some of the reasons why I’m not convinced that a practical nominal GDP rule would represent a material advance over (practical) inflation targeting.

One of the attractions of nominal GDP targeting is that it prompts a monetary policy tightening when export commodity prices rise, even if there is no immediate rise in consumer prices. But as I noted one needs to think specifically about the characteristics of the particular economy.

In thinking about an export price shock, it might also be important to understand the transmission of the shock across the rest of the economy. A highly open economy, in which a generalized export price shock affected firms across an employment-rich wide-ranging export sector, might look considerably different than a sector-specific shock in a moderately open economy where the commodity production sectors employ relatively little labor (the story in New Zealand dairy, and much more so in Australian minerals and gas extraction). If New Zealand experiences a surge in dairy prices, and much of the proceeds are saved by farmers—perhaps because they are very conscious of the volatility of prices—why would one want to tighten monetary policy against that lift, if there was little or no apparent spillover to domestic (wage or price) inflation? Perhaps if the shock destabilized wage expectations there could be a basis, but there has been little sign of that sort of wage-setting behavior in response to recent export price shocks. The issues are even more stark in Australia, where most of the profit variability in the face of export price shocks accrues to non-Australian owners of capital (whose consumption choices are likely to put few pressures on domestic resources in Australia).

Partly for this reason, over several years I have been drifting towards the conclusion that if one were to replace inflation targeting with another rule, in New Zealand’s case nominal wage targeting might have rather more appeal.   I noted

Much of the academic discussion of inflation targeting focuses on the idea of stabilizing the stickier prices in order to minimize the real costs of adjustment to shocks. Since, as this paper agrees, wages are typically among the stickier prices, perhaps we should be more seriously considering the merits of nominal wage targeting, as Earl Thompson argued decades ago. I have noted elsewhere (Reddell 2014) that such a rule could even have financial stability advantages. Nominal wages are the prime basis for servicing the nominal household debt that dominates the balance sheets of our banks. Faced with adverse shocks, and especially deflationary ones, nominal debt is arguably the biggest rigidity of them all. It would be interesting to see such a rule evaluated in a suitable model.

But…..

If productivity shocks were the dominant source of dislocations in New Zealand, such a wage rule could also have considerable appeal— shifting the variability into the price level rather than into (sticky) nominal wage inflation. As it is, over the last twenty years, wage inflation has followed a rather similar path to core CPI inflation— and does not look much like fluctuations in the path of nominal GDP (or in NGDP per capita, or NGDP per hour worked). So perhaps, at least over that period, policy should have looked very little different under a wage rule than under the CPI inflation targets that successive ministers and governors have agreed upon.

Of course there might be considerable political/communications difficulties with wages-targeting.  But this would be nominal wage targeting: actual real wages and the labour share of income would still emerge from the market process.   But given these communications difficulties, the case for change would have to be stronger than it is right now (although for what it is worth, current wage inflation also probably argues for looser monetary policy –  just like the CPI or nominal GDP).

I have little doubt that inflation targeting is not the “end of history” for monetary policy.  But the choice between inflation, nominal GDP, or wage targets –  in levels or growth rate terms –  doesn’t seem to be the biggest issue we face in designing monetary policy and the related institutions.  In practical terms, each would rely on forecasts, and our forecasts simply aren’t very good.  And each still faces the issue of the near-zero lower bound.  There are arguments that levels targets might help alleviate the ZLB, but only zealots think that in isolation it would make a huge (or sufficient) difference.  We need much more energy being applied to either removing the ZLB constraints (which are essentially regulatory in nature) or raising the target for inflation (or nominal GDP or wages growth) sufficiently so that the zero bound is no longer likely to be binding.  The Bank of Canada is right to be looking at this issue.  Other central banks and finance ministries need to be doing so.

And I still think the other issue is one of just how much accountability there can actually be for autonomous central banks implementing monetary policy.  As I have noted recently in both the New Zealand and US contexts, in practical terms there is very little.    In the United States, John Taylor has argued for legislating something like a Taylor rule as a benchmark against which the Federal Reserve’s judgments can be formally evaluated, requiring the Fed to explain deviations from the recommendations of that rule.  Some on the political right argue for a return to the Gold Standard.  I don’t think either would be desirable, but in a sense both are reactions against the delegation of too much unchecked power to central banks.  The original conception in New Zealand was of a high degree of effective accountability –  an easy test as to whether or not the Governor has done his job. Money base target ideas had a similar conception –  plenty of delegation, but plenty of effective accountability.  It turned out not to be so easy.  But if we cannot meaningfully hold these powerful independent agencies to account –  in ways that mean real consequences for real people –  I suspect the debate will begin to turn again as to whether the power should be delegated to unelected officials at all.  Citizens can vote governments out of office, and that has real consequences for real decisionmakers.

There have been a couple of odd comments this week about the use of macroeconomic policy tools in New Zealand.

In his weekly column yesterday, Brian Fallow suggested that it had been unwise to have put so much emphasis on getting the budget back to balance, and that it was time for more fiscal stimulus.  Of course, there is nothing sacrosanct about getting back to balance by any particular date, but if anything I thought our government had been rather too slow to get there.  With the benefit of record –  and probably unsustainable –  terms of trade – there wasn’t really much excuse for having run deficits in the last few years.       And a tighter stance of fiscal policy should, at the margin, have eased the upward pressure on demand, the OCR, and the exchange rate.

Fallow draws on the generalised advice of the IMF to advanced economies.  But most of those advanced countries can’t do anything much with conventional monetary policy even if they wanted to.  Quite a few advanced countries (including the euro area) already have negative policy interest rates, and many of the others –  US, UK, and Japan among them – are essentially at zero.  Perhaps new rounds of QE might make some difference –  I rather doubt they could do much – but to all intents and purposes monetary policy options are exhausted.    That is partly the fault of central banks and finance ministries that have done nothing material over eight years to remove the near-zero lower bound on nominal interest rates but, choice or not, it is the situation today.

If there is still excess capacity in many of those countries, and if many of them face widening output gaps if world activity growth continues to slow, the appeal of looking to fiscal policy for stimulus is understandable.  In general, I’m not sure it is a call that should be heeded to any great extent, as most of the larger countries already have rather sick fiscal positions –  made considerably worse when those countries resorted to fiscal stimulus, to loud cheering from the IMF, in 2008/09.

New Zealand –  and some other advanced countries such as Sweden, Finland, Estonia, Australia, and Switzerland –  is in the fortunate position of having a low level of public debt.  That means we do have some room  to use fiscal policy if such stimulus is required.  But even that potential isn’t limitless. Faced with another severe recession, even allowing the automatic stabilisers to work would add materially to the government’s debt over several years.  In such a downturn, it is probable that the government’s speculative investment vehicle – the New Zealand Superannuation Fund –  would lose a lot of money.  And for those who worry about the financial stability risks of the house prices more than I think warranted, bear in mind the potential need to bail out banks and their creditors.   If there is a severe downturn, we need the political room to allow those buffers to work, not to have to resort to pro-cylical fiscal policy

Fallow –  and many international commentators –  have favoured additional government spending because interest rates are low.  But remember that interest rates are low for a reason –  it isn’t just some number thrown up by a random number generator.  I’ve argued previously that the effective cost of capital the government should be using in deciding on even good quality projects is probably in excess of 10 per cent (the sort of standard private businesses use), not something close to the government bond rate.   And all this is before the questions that must be asked about the poor quality of too much government spending.

There are distinct political limits to how much fiscal stimulus any government can do, even in a crisis.  Why fritter away that potential now, when our OCR is still 2.75 per cent?  New Zealand has far more monetary policy headroom still open to it than most other countries do.  There are real macroeconomic issues in New Zealand –  as Fallow points out, the high and rising unemployment rate suggests that the economy continues to run below capacity  –  but as Eric Crampton noted in his response to Fallow yesterday, it is not as if monetary policy has been tried and failed.  Rather, because of the repeated mistaken calls by the Reserve Bank, monetary policy has barely been tried.

ANZ advocated fiscal stimulus back in July.  I set out here the reasons why I thought that was the wrong call. I don’t think I’d resile today from anything in that piece.

But the other odd comment on New Zealand macroeconomic policy came from someone who really should have known better.    In his speech on Wednesday, the Governor of the Reserve Bank

It is important also to consider whether borrowing costs are constraining investment, and the need to have sufficient capacity to cut interest rates if the global economy slows significantly.

I’ll largely ignore the first part of the sentence (although if inflation is low and unemployment still high, isn’t more private sector investment generally likely to be a good thing?).  It was the second half of the sentence that really reads oddly –  and arguably, worse than oddly.

This idea of keeping some powder dry in case there is a renewed sharp slowdown pops up from time to time in international commentary.  We’ve even seen the argument made that the Federal Reserve should raise interest rates now so that it has room to cut them if there is a future slowdown.  But it is a deeply flawed argument.  It may have some merit on the fiscal side –  higher public debt now leaves less room to run up more debt later on  – but in respect of monetary policy it is just wrong.

Monetary policy that is tighter than strictly necessary (in terms of the PTA) now, is likely to both weaken the economy (relative to the counterfactual) over the coming 12-18 months, and further low inflation and inflation expectations.  Lower inflation is undesirable (in terms of the PTA itself) and low inflation expectations are deeply problematic.  Lower inflation expectations, all else equal, raise real interest rates for any given nominal interest rate.  The experience of advanced world since 2007 says that one of the biggest macro management problems of a sharp slowdown in the presence of low inflation is getting real interest rates low enough.    It was easy to get real interest rates materially negative in the high inflation 1980s but it is almost impossible to do so now.  In other words, holding up nominal interest rates now increases, perhaps materially, how much one might need to cut rates if a severe downturn happens, while doing nothing to create that extra space.

The Governor has rather reluctantly come to acknowledge that global deflationary risks.  The last thing anyone in his position should be doing right now is making choices that would make it harder to handle the next sharp slowdown.  But that is what he appears to be set to do.

If anything, those countries that still have monetary policy room to move should be doing so now.  Real interest rates should be as low as possible, consistent with the PTA –  and perhaps especially in a country that starts with the highest real interest rates in the advanced world.  Rather than core inflation of 1.5 per cent or less, the Governor should be rather more comfortable with core inflation around 2.5 per cent.  The best way to get that sort of outcome would be to have cut the OCR over the last couple of years, not raised it.

As a commenter here the other day put it, this idea that we should hold the OCR up now so that it can be cut later “is like keeping your shoe laces tied so tightly that it cuts off your circulation, just so it feels good to loosen your laces later.”

Between the failure to do anything about the zero lower bound –  which the Governor now (belatedly)  implicitly acknowledges to be an issue –  or, absent that, to consider a higher inflation target, the Governor and the Minister have left New Zealand less well placed than it could have been if there is a new sharp global slowdown in the next few years. But the decisions that keep on delivering such unnecessarily low rates of core inflation (and high unemployment) are those of the Governor alone.

Plenty of market economists have commented on yesterday’s inflation numbers.  My only contribution is a simple chart of a new series Statistics New Zealand has just started publishing.

Non-tradables inflation has long been the focus for analysis of the underlying inflation position.  Tradables inflation is thrown around by short-term swings in international oil prices and level shifts in the exchange rate, and non-tradables inflation should provide a better guide to underlying inflationary pressures. But non-tradables inflation is made harder to read because of repeated tax increases (notably tobacco taxes) and changes in government charges –  which don’t reflect anything about the state of the domestic economy.   It is quite common internationally for statistical agencies to publish series excluding taxes and government charges.  And now SNZ has provided us with this series for New Zealand.  It has the advantage, over the Bank’s sectoral core factor model measure, that it is not prone to revisions.

Note that this measure of non-tradables inflation is running at only 1.6 per cent, barely above recessionary low.  Non-tradables inflation should be expected to run above tradables inflation on average over time (there is typically more scope for productivity gains in tradables than in many non-tradables).    Indeed, if CPI inflation was going to average around 2 per cent –  the Bank’s target  –  non-tradables inflation shoiuld probably average somewhere in the 2.5-3 per cent range (and perhaps tradables might be in the 1-1.5 per cent range).     Non-tradables inflation is extremely low in New Zealand –  it is too low and should, as a matter of active policy, be raised.

No doubt the Governor and his economists will say that that is what they have been trying to do.  But if so, they have repeatedly failed.  Becoming reluctant to cut the OCR further because of the housing market (one of the channels through which lower interest rates work –  a buoyant housing market is a desired feature not a bug) or for fear of hitting zero in a global downturn is a recipe for continuing the mistakes of the last few years.

To repeat Eric Crampton’s line: monetary policy has scarcely been tried.  It should be.

I learned something from Graeme Wheeler’s speech this morning. Having not gotten round to reading Andrew Graham-Dixon’s Caravaggio, which has now lingered on our bookshelves for five years, it hadn’t occurred to me that the Gericault painting The Raft of the Medusa was influenced by Caravaggio; in Graham-Dixon’s words, it is a “modern secularised version of an altarpiece by Caravaggio”.

It is a striking painting, but this was a strange speech – and nowhere more so than in the concluding reference to The Raft of the Medusa.

There was plenty of routine material I agreed with. It gets boring to say it, but central bankers have to go on making the point that monetary policy has no material impact on longer-term real interest rates, longer-term average real exchange rates, or the longer-term real growth performance of the economy. For 25 years, real interest rates and the real exchange rate have averaged too high here, and real per capita growth has been too low. There are things that could have been done to change that, but changing the monetary policy framework isn’t one of them.

But beyond that it all got rather strange. He started his speech with the references to Caravaggio to play up the dark and turbulent nature of the last few years (as he sees them). How difficult it all is for small country central banks. More so, he suggests, than previously.

I guess Graeme Wheeler is a latecomer to central banking. He’s been Governor for only three years and spent most of his career doing stuff other than macroeconomic policy. So perhaps recency errors should be pardoned. But it isn’t clear why he thinks, or wants us to think, that his job is harder than that of his predecessors (Don Brash, the challenges of disinflation, or Alan Bollard, and the massive credit boom). Self-pity is rarely an attractive quality, and perhaps especially so when it comes from highly-paid powerful public officials. The Governor has been given a relatively straightforward job to do by Parliament, and – unlike many of his offshore peers – he still has all the conventional tools at his disposal. He has made some mistakes along the way, as his predecessors did (and his successors no doubt will too), but it just is not that hard to get it roughly right.

There are some analytical puzzles to be sure, but we (taxpayers) don’t pay Wheeler to answer all those. We pay him to keep medium-term trend inflation near 2 per cent. And he hasn’t done that job very well. Inflation has surprised on the weak side for several years. It has surprised both the Bank and the markets, but the Bank has the job of delivering inflation near target. Having fairly consistently failed to do so, a reasonable rule of thumb might have been something along the lines of “core inflation has been below the midpoint of the target for so long, and we don’t fully understand why, so we’ll hold fire, and not raise interest rates until (say) core inflation is actually back to around 2 per cent”. It isn’t a perfect rules – in an imperfect world there are no such rules – but it would be better than what we’ve had.

Part of Graeme Wheeler’s defensive strategy – which shouldn’t really be needed, because there should be no great shame in recognising a mistake and owning up to it – is to cloth himself in the problems of other countries. Many other advanced economies have also struggled with at best modest (per capita) economic recoveries, and surprisingly weak inflation. But most of those countries had little or no conventional monetary policy ammunition left. Interest rates were at zero. I don’t think Wheeler’s speech even mentions the point.

By contrast, New Zealand’s OCR at 2.75 per cent means the Governor (and his predecessor) had plenty of room, if he (they) had chosen to use it, to secure a rather more conventional recovery in New Zealand. As I’ve pointed out previously, in conventional New Zealand recoveries we see a couple of years of 4-5 per cent GDP growth. We’ve seen nothing like that since 2009, despite the very rapid population growth in the last year or two. The number of people unemployed has stayed high, and has recently been rising again. That isn’t because of the travails of the rest of the world, but because of the Governor’s misjudgements. Inflation wasn’t rising. He didn’t need to raise interest rates.

As he winds up his general observations, Wheeler includes a couple of other curious paragraphs. He claims that “economic management has come a long way since Paish described it in the 1960s as like driving a car with a brake and an accelerator and only being able to look through the back window.” He backs this claim, that things have come a long way, by reference to sophisticated models used by central banks today [and when is that expensive model, developed at taxpayers’ expense, finally going to be published?]. But then he changes course midstream, concluding that really the models can’t tell one much and are more use for posing questions than delivering answers. I entirely agree with that final observation, but then it is a puzzle as to why the Governor thinks that economic management has improved a lot since the 1960s. As the Governor found with his ill-fated tightening campaign last year, forecasting remains hard, especially about the future. And even the rear-view mirror is prone to mislead at times.

And then Wheeler winds up the whole speech with his paragraph about The Raft of the Medusa, the “forlorn and exhausted sailors” and the “wild seas” which are apparently “symptomatic of the world central bankers are trying to navigate”. It is all bizarrely self-pitying, especially for a Governor with instruments at his disposal. The steering wheel isn’t broken.  The hull isn’t holed.

Here is the Wikipedia summary of the story of the painting.

it is an over-life-size painting that depicts a moment from the aftermath of the wreck of the French naval frigate Méduse, which ran aground off the coast of today’s Mauritania on July 2, 1816. On July 5, 1816 at least 147 people were set adrift on a hurriedly constructed raft; all but 15 died in the 13 days before their rescue, and those who survived endured starvation and dehydration and practiced cannibalism. The event became an international scandal, in part because its cause was widely attributed to the incompetence of the French captain

Great painting, but of a terrible event, for which the captain had to take blame.  The French public knew that.

The mismanagement of New Zealand’s monetary policy in recent years springs to mind. Fortunately, no one dies from those sorts of mistakes, but thousands of people are unemployed today who would not have been if the Reserve Bank of New Zealand – and specifically its Governor – had not made repeated choices – and it was pure choice, unlike the situation in ZB countries – to hold interest rates persistently higher than they needed to be. The apparent indifference of New Zealand’s elites – and of the Governor and his Board – to the awfulness of the involuntary unemployment (“forlorn and exhausted”?) is something I still struggle to understand.