Monetary policy as we know it today – discretionary choices made by central banks (or Ministers) – is a relatively new thing. Of course, money has been around for a very long time, and the state has often had a role in specifying the metal content of various units of money, and (not unrelatedly) what money was acceptable in settlement of tax obligations. But there was no such thing as “monetary policy” in, say, the 16th century – when prices rose across Europe, it was because the additional gold and silver being mined in South America, adding to purchasing power of (first) Spaniards and then more generally. The gold rushes of the mid 19th century – in which New Zealand had a small part – had qualitatively similar effects.
Even in the heyday of the classical Gold Standard – the few decades prior to World War One – when central banks did exist in a growing number of countries (although not, for example, the US, Australia, Canada or New Zealand, there wasn’t much to monetary policy. Convertibility into gold, at a fixed (government-set) parity, was at the heart of the regime, and variations in official interest rates – eg by the Bank of England or the Banque de France – were largely about managing pressures on gold reserves. If there was a net loss of reserves from the UK, the Bank of England would typically raise its interest rates. It wasn’t a mechanical process, and central banks would at times borrow from each other to tide themselves over what were thought to be purely temporary pressures. In places without central banks – New Zealand was an example – banks were obliged by law to convert their notes into gold on demand. Banks themselves had to manage pressures on their own reserves – whether gold, or balances held with banks in London – by altering the interest rates they offered, and varying their credit standards (tightening credit would reduce demand for imports).
Across much of the world, World War One disrupted these arrangements. New Zealand suspended gold convertibility on the outbreak of the war, and never restored it. Much of the world attempted to go back onto gold in the 1920s – the UK famously restoring convertibility in 1925 – as part of trying to restore normalcy and monetary stability. But the restoration didn’t last. Many authors see the attempt to return to gold, in a somewhat hybridized manner, as a key cause of the Great Depression, and by the mid 1930s the Gold Standard had been largely abandoned. The imperatives of short to medium-term macroeconomic stabilisation displaced the belief in fixed parities. Voters – this was the new age of universal suffrage – demanded that governments “do something”. And the evidence is pretty clear that countries that went off gold earliest – or devalued earliest – recovered soonest from the Depression. The imperative of doing something about really severe cyclical unemployment drove monetary actions and monetary regime choices, including the establishment in 1934 of the Reserve Bank of New Zealand.
There was an attempt after World War Two to re-establish something that looked like a system based on gold (the US offered governments – only – convertibility into gold, while other countries had fixed exchange rates to the US dollar), but it was a much different beast. Most countries had quite tight controls on cross-border capital flows – of the sort that had not previously existed in peacetime in democratic societies – which allowed countries to use counter-cyclical domestic fiscal and monetary policy to do more to promote something akin to full employment, without too readily putting the exchange rate pegs in jeopardy. It didn’t end the business cycle (of course), and didn’t avoid periodic exchange rate crises but for a time – a couple of decades – it more or less worked. But pushed too far, under pressure of various political and demographic shocks, it broke down into the era of the Great Inflation – loosely, from around the mid 1960s to the mid 1980s.
And thus monetary policy as we recognise it today really only dates back a few decades. The major Western economies floated their exchange rates in the early 1970s, New Zealand and Australia did in the mid 1980s, and some advanced OECD countries (eg Sweden and Norway) only did so in the 1990s. Tiny Iceland only floated in 2001. Inflation targeting – whether formally (as pioneered by New Zealand) or less formally – makes sense only in the context of a pretty flexible (probably floating) exchange rate. It is a regime that exist with twin goals in mind, whether or not they are written down in statute book somewhere or not.
Floating your currency allows a country to choose its own inflation rate. That was a big consideration in the 70s and 80s: if, like Switzerland, you wanted to maintain low inflation, you couldn’t do so with a fixed exchange rate to the rest of the world that was running an inflation rate of 10 per cent. But it also allows your country to cope better with severe adverse real economic shocks; in particular shocks specific (or more intense than typical) to your particular country. I wrote last week about the Finnish situation in the late 80s and early 90s. We had it pretty tough here during that period, but it was nothing like as bad as the Finnish experience – despite big structural reforms going on at the same time – because we had our own monetary policy and could allow interest rates to fall, and the exchange rate, when times turned tough. We didn’t give up on inflation – in fact this was the period we were getting inflation to target for the first time – but we had an institutional arrangement that provided a better mix of low inflation and somewhat-mitigated real economic downturns. (In fact, it wasn’t so different back in the 1960s when, faced with a big fall in the terms of trade, New Zealand chose to devalue its nominal exchange rate – an active monetary choice – rather than attempt to force the adjustment through lower domestic prices and wages. Most observers – then and now – would have thought that alternative would have been much more costly, in unemployment and lost output.)
All of this so far is really a rather long prelude to articulating a disagreement with an eminent former colleague.
Last week, Reuters ran an article under the heading “New government in New Zealand could spell changes for pioneering central bank”, with a particular focus on what a Labour-led government might mean. The article quoted various people (including me – my own thoughts were elaborated here) but the comments that caught my eye were those by Arthur Grimes. These days Arthur is a researcher at Motu – mostly focused on issues other than macroeconomics – a professor (of wellbeing and public policy), and generally one of the “great and the good” of New Zealand economics (president of the Association of Economists etc). But in his younger days he spent 15 years or so at the Reserve Bank, rising to Head of Economics and then Head of Financial Markets before leaving for the private sector and academe. In that time, he was one of those closely involved from the Bank’s side in the design of the Reserve Bank Act, and was also involved in the practical development of inflation targeting (the two developed in parallel). Later, he ended up on the Reserve Bank’s Board, serving as chair of the Board until about four years ago. As chair of the Board, he probably should be seen as having had prime responsibility for the appointment of Graeme Wheeler as Governor. In many respects, were he to be interested, Grimes could have been the best of the status quo candidates to replace Wheeler permanently.
Grimes is pretty deeply committed to the status quo on monetary policy (I’m not sure what his views now would be on single decisionmaker vs a committee, although interestingly he has been a longserving Board member of the Financial Markets Authority, where decisionmaking responsibility rests with the Board not the chief executive).
And that commitment to the status quo was on display in the Reuters article.
“It’s a huge change. We’ve had over 25 years of an extraordinarily successful monetary policy that has been copied around the world,” said Arthur Grimes, RBNZ’s chief economist in the early 1990s and Board Chair between 2003 and 2013. Any change without careful consideration and analysis would be “extraordinary”, he added.
For 28 years, New Zealand’s central bank has had the single aim of keeping inflation between a set range. But Labour wants to add employment to the bank‘s mandate, a goal shared by NZ First which also wants to broaden the Reserve Bank of New Zealand’s (RBNZ) focus to include greater management of the local dollar’s value against other currencies.
Grimes, however, argues that history proves monetary policy cannot have a sustained impact on employment.
“It would be like having someone who is running for health minister argue for a cancer drug to be used for heart issues,” said Grimes
I’m not sure what benchmarks Grimes is using to describe New Zealand’s monetary policy as “extraordinarily successful”. There is no doubt that inflation has been much lower and more stable than it was in the 1970s and 1980s – although not much different than it was in the 15 years prior to 1967 – but that is true almost everywhere. So if one is going to argue that the specifics of the way the New Zealand target/Act are specified have been “extraordinarily successful”, and need protecting, one would need to show that that particular specification has led us to have better outcomes than, say, other advanced countries that did inflation targeting differently, that specified things (formally) less tightly, than put less emphasis on formalised accountability mechanisms, or which even kept “dual mandate” types of language in their statutes and official communications/rhetoric. The United States and Australia might be obvious cases to look at. But it would be impossible, as far as I can see, to demonstrate such superior New Zealand performance.
Now it is no doubt true that the New Zealand (and near-parallel Canadian) early experiences with inflation targeting did influence other countries’ choices to some extent. But it would be flattering (and fooling) ourselves to suppose that the specifics of the New Zealand model have been widely copied at all. Indeed, in several important areas – including governance/accountability – what is striking is how few countries have gone the same way we did. We remain, I think, the only inflation targeting country to have (a) twice changed its target, and (b) where monetary policy has been an election issue for some parliamentary parties or other every single election since the 1989 Act was passed. Even on the specification of the mandate, a Reserve Bank Bulletin article a few years ago highlighted just how a wide a range of ways mandates and overarching goal for monetary policy are specified even among advanced country inflation targeters. It is not, after all, as if what the Labour Party has proposed involves tossing out inflation targeting. That would indeed be extraordinarily bold – not necessarily wrong, as there are plausible alternatives bruited about internationally – without a lot more work. But simply adding a formal statutory recognition that we have active discretionary monetary policy because of concerns about shocks that can take unemployment well away from its full employment (non-inflationary) level isn’t radical or extraordinary at all.
Analogies can be powerful rhetorical devices, so it is always worth testing analogies that people propose to see if they capture a useful and valid point or not. And it was Grimes’s analogy that really prompted this post. He suggests that adding something – and recall that all of us are reacting to a general point not specific proposed wording – about unemployment to the Reserve Bank’s statutory monetary policy mandate
“It would be like having someone who is running for health minister argue for a cancer drug to be used for heart issues,” said Grimes
And that is simply an invalid analogy (assuming, as I imagine both Grimes and I do as non-medical laymen, that there is no connection between cancer drugs and heart issues). It is generally recognised that monetary choices can have output and employment consequences and that, at times, those effects can be large, and persistent enough to be troubling for individuals and (voting) populations. Of course, no one argues (I think) that monetary policy choices today will affect the unemployment rate 15 years hence. If there are problems there, you need a different set of tools (labour market reforms, welfare reforms etc). But a succession of monetary policy choices today can, if mistaken, leave the unemployment rate away from a long-term sustainable rate for some considerable time. One could mount a plausible argument – for example – that the fact that the New Zealand unemployment rate has been above all official estimates of the NAIRU for some years now, while at the same time core inflation has been below target, might be one of those examples. Choices – risks taken, or not, under uncertainty – have consequences.
And that sort of example (demand shocks, or surprises) is the easy case – after all, getting inflation back to target and getting unemployment down work in the same direction. For plenty of shocks it works the other way. A big boost to oil prices tends to raise CPI inflation. Attempting to prevent, or reverse, those inflation effects can only be done by monetary policy actions that would raise unemployment. The Reserve Bank – and those setting its specific goals – have always considered that would (normally) be an inappropriate response. In those circumstances, we allow a bit more inflation temporarily – and a permanently higher price level – to avoid unnecessary departures of the unemployment rate from its sustainable level. We articulated that logic in public right from the very first days of inflation targeting (it is explicit in the first Monetary Policy Statements – which I wrote and Grimes commented on).
Don’t get me wrong. There are some arguments for not including the unemployment references in the Reserve Bank Act. I was persuaded by them for a long time, even if I no longer am. But they are fine judgements at the margin, nothing remotely like the suggestion of snake-oil peddling implicit in Arthur Grimes’ medical analogy. Price and wage rigidities – that exist for rational and efficient reasons – mean that in the short to medium term, targeting inflation and targeting unemployment are inextricably linked (not mechancially, but inextricably).
Other people recognise this. I’ve linked previously to the writings of leading academic in the field, Lars Svensson (also former monetary policy board member in Sweden, and former independent reviewer of New Zealand monetary policy), who favours explicit statutory recognition of the role that deviations of unemployment from a long-run sustainable rate play in monetary policy. The Reuters article quotes Phil Lowe, current RBA Governor, in defence of such language in the RBA Act (although I’d argue that the RBA experience illustrates that words make less difference than people). Janet Yellen and Ben Bernanke have similarly been comfortable in the United States, and although Alan Greenspan was a well-known hard money man (favouring, at least in principle, a “true zero” inflation average), (a) he was never that keen on inflation targeting itself, preferring to keep a considerable measure of discretion to himself, and (b) he was not averse to talking about unemployment (“we are keenly interested in what we can do to maximise sustainable employment growth and to reduce unemployment” as his biographer records him noting at a Jackson Hole conference at which Don Brash was one of the speakers).
So, of course, the specific wording a Labour-led government might seek to introduce – should things go their way – should be carefully scrutinised. But it wouldn’t be extraordinary at all to make such a change, rather it would be a pretty straightforward translation into statute of the reasons why we have a discretionary and active monetary policy in the first place. If we didn’t care about the output and employment consequences of adjusting to shocks, we might as well just go back to the Gold Standard, or a fixed exchange rate.
Not a word of this would be particularly surprising to Arthur Grimes. A few years ago, on leaving the Reserve Bank Board, he gave a series of lectures in the UK on central banking. They were a pretty robust defence of the status quo, broadly defined. In some places, I thought he claimed too much, but the underlying economics wasn’t much different than anything I’ve articulated here. There was, for example, a nice piece on the exchange rate system headed “A floating exchange rate is the worst exchange rate regime (except for all the others that have been tried)”. I’d agree with him entirely. And what reasons does he give? This is from his conclusion
Third, dynamics do matter. The evidence shows that countries that adopt a hard peg may experience greater persistence in economic cycles than those with a floating currency. If domestic prices and costs can adjust easily, a hard peg may not be problematic. But in a country with sluggish domestic price adjustment, the hard peg can result in persistent real sector imbalances as we have seen both in the upward and downward direction for several Euro-zone countries.
If we rule out a soft peg as being the worst of all worlds, how should a country decide whether to adopt a hard peg or a floating rate? The trade-offs are complex: How flexible is domestic price adjustment? How diverse are the country’s trading partners, and hence what are the effective currency impacts of pegging to a specific country or bloc? How likely is it that a government will adopt sensible economic reforms under one or other regime?
In the end, a floating rate appears to have advantages, especially in relation to persistence of real sector variables, over a hard peg. However, if the political economy is such that a country with weak policies is more likely to adopt reforms under a hard peg than under a float, then it may be better for it to retain a hard peg and be forced to reform its other policy settings.
Ultimately, in terms of long run economic performance, the choice of regime does not matter much, so we cannot expect substantive changes in long term outcomes through a change in the exchange rate regime. But while the long term destination may not change, the quality of the ride does differ depending on the chosen vehicle.
Ignoring unemployment in choosing monetary policy regimes, and conducting monetary policy, might be more like caring only about the speed at which one drove from Auckland to Wellington, not the comfort or the safety of the journey. No one does. In practice, no one ignores unemployment in monetary policy design either. The question is whether explicit recognition of that fact, in statute or even in the Policy Targets Agreement, in conjuction with the appointment of a good Governor, might (a) assist communications, around what the Reserve Bank really exists for, and (b) at times, produce better outcomes, and better accountability for performance against the unemployment dimension of what we care about in monetary policy management. Reasonable people can reach different conclusions on that point, but whichever side one lands it simply isn’t a terribly radical choice. And, on the other hand, the status quo – whether around the Bank itself, or short to medium term, economic outcomes, or the ongoing political debate around these issues- isn’t so obviously superior that we should not explore alternatives.