In 1999, having been out of office for nine years, the Labour Party campaign platform included promises about monetary policy. They undertook to change the Policy Targets Agreement – and they did, adding the words (still) requiring the Bank to “seek to avoid unnecessary instability in output, interest rates and the exchange rate”.
But they also promised an independent inquiry into the operation of monetary policy. It was then 10 years since the Reserve Bank Act had been passed, and we’d gone through both a wrenching but successful disinflation, and through one full business cycle since something like price stability had been established. Some of elements of the management of that cycle hadn’t been the Reserve Bank at its finest: use of the Monetary Conditions Index to guide short-term policy management had given us a (relatively short) period of quite astonishing interest rate volatility, not helped by being slow to appreciate the significance of the Asian financial crisis.
I don’t suppose Michael Cullen was ever a great fan of Don Brash’s. But Brash had already been reappointed for a third term in 1998 (arguably fortunate that the reappointment was done before the nature of the MCI debacle was fully appreciated). And Cullen was clearly uneasy about the volatility in New Zealand interest rates, and about the big cycles in the exchange rate. There were also suggestions that he was a bit uneasy about the rule of a single unelected technocrat at the Reserve Bank of New Zealand, and Labour at times seemed to look longingly across the water at the Reserve Bank of Australia (with a higher target, more flexible rhetoric, and a reputation for being a steady hand). And, of course, Labour was coming into government with Jim Anderton as Deputy Prime Minister. Anderton had still not been reconciled to the Reserve Bank Act framework at all. So it was, all round, opportune to have an inquiry.
But of course whenever one sets up an independent inquiry, the name of the person appointed to conduct the inquiry tells one a lot about what the appointer is looking for. There were all sorts of names bandied about at the time, including (for example) Bernie Fraser who had until recently been Governor of the Reserve Bank of Australia, and whose centre-left sympathies were not exactly unknown. But the government settled on Swedish academic Lars Svensson. Perhaps being Swedish – home of centre-left big government – lulled some on the left of New Zealand politics. But, more importantly, Svensson was also a leading academic author on aspects of the (then still relatively new) theory and practice of inflation targeting. He’d also spent some time in New Zealand a couple of years earlier, as the Reserve Bank professorial fellow. In other words, it was never likely to be a terribly radical report.
And it wasn’t. Which is not to say that it wasn’t a useful exercise, or that Svensson did not make some useful recommendations. He did. Some of the less important recommendations – eg around the make-up of the Bank’s Board, and the publication of a Board Annual Report – were even adopted. Some others that should have been adopted – for example, the introduction of a monthly CPI – still, unfortunately, haven’t been. Svensson also proposed legislating for committee decision-making for monetary policy, but his proposal of a committee of insiders (including the role I then held) went nowhere: among other reasons no doubt, Michael Cullen hadn’t come into politics to give statutory power to more Reserve Bank pointy-heads.
I was quite heavily involved in the review, both in contributing to the Reserve Bank’s own substantial submission to the inquiry and – along with a couple of Treasury economists – as part of the secretariat to the inquiry itself. For an inquiry into the Bank, it was a bit of an odd arrangement – shortly after the inquiry began I was promoted into one of the half dozen top policy/management roles in the Bank and did the two roles in tandem – but I guess it is a small country, and there was never much doubt about the overall favourable stance Svensson was likely to take. He was a big fan of Don Brash, and the conclusions were his not those of the secretariat (in fact, flicking through notes stapled in my copy of the report yesterday I noticed that in some places we – and the Bank – urged Svensson to toughen up his comments, lest the report look in places a bit like a whitewash.)
But the main point of this post isn’t about history. It was initially prompted by an observation in a column in the Sunday Star-Times the other day, in which their resident right-wing columnist was quoting Svensson from 2001. Damien Grant, in commenting sceptically on Grant Robertson’s proposals,
Robertson might find it useful to know that when Cullen became finance minister he commissioned a review of the Reserve Act by Swedish economist Lars Svensson who concluded:
“It is beyond the capacity of any central bank to increase the average level or the growth rate of real variables such as GDP and employment.”
The understanding that monetary policy can only influence the value of money and nothing else is one of the few untarnished successes of modern economic thought. It is deeply disturbing that Grant Robertson does not seem to appreciate this.
As a commenter observed yesterday, no mainstream economist believes that monetary policy can change the long-term level of employment/unemployment/real GDP or whatever. In the long-term monetary policy can only affect nominal variables. Svensson certainly believed that then and believes it now. I don’t know whether Grant Robertson does, but I expect so.
But equally, not many mainstream economists believe that active monetary policy typically has no effect (short to medium term) on real variables. There is pretty general acceptance, I think, that the depth and severity of the Great Depression was in substantial part a matter of monetary mismanagement. That’s a deliberately extreme example, but it both illustrates the point and (historically) provides some of the backdrop to modern more active discretionary monetary policy. In earlier decades, adjustments in central bank interest rates (where central banks existed at all) were mostly about maintaining the gold (or silver) convertibility of the currency. Domestic economic conditions didn’t play much of a role. Really bad experiences like the Depression, along perhaps with the rise of universal suffrage (the more marginal got a say in politics), helped change that focus.
Writing in 2001, reviewing New Zealand policy, Lars Svensson had no doubt about the importance of real variables in the management of monetary policy. He didn’t question section 8 of the Reserve Bank Act – the focus on price stability. But his articulation of flexible inflation targeting – what the Reserve Bank saw itself practising – involved short-term trade-offs between pursuit of the inflation target, and the variability of the real economy. At the time, as an academic, he focused explicitly on the trade-off with variability in the output gap (the gap between actual and potential output), and devoted several pages of the report to a discussion of the issue, describing it not just as a very short-term matter, but as a “short and medium term” issue. (For anyone interested, the full report and associated documents are here.) What he was talking about wasn’t at all inconsistent with the 1999 addition to the PTA, quoted above, about seeking to “avoid unnecessary instability in output”). And there was only one tool – the OCR.
Standing back from his more theoretical perspective, there was good reason why one might want explicit consideration of real variables in the official articulation of what an independent central bank was asked to do. One doesn’t need active monetary policy if all one is concerned about is long-term stability in the general level of prices – something passive like the Gold Standard would do it. But the Reserve Bank Act – and other comparable legislation abroad – was about a regime for governing the discretionary active use of monetary policy. We had – and have – such a policy because it was believed that discretionary monetary policy could make a difference, over meaningful horizons, to real economic outcomes (GDP, unemployment or the like) even if not to the trend or potential levels of those variables.
Some years later Lars Svensson himself became a policymaker, as a fulltime member of the executive board of Sweden’s Riksbank. The Executive Board makes the monetary policy decisions in Sweden. Many of my old Reserve Bank colleagues don’t agree, but I think Svensson proved to be an ideal person to have on a monetary policy decisionmaking committee. He had strong expertise in the subject – albeit initially at a rather abstract level – and a cast of mind which meant that he wasn’t just going to fall into line with the preferences of the Governor and the long-term staff advisers. He strongly and opened argued against the Riksbank’s strategy, adopted several years back in the wake of the global recession of 2008/09, of trying to use monetary policy to lean against the accumulation of household debt, even at the expense of inflation undershooting the target (and unemployment remaining very high). It was a costly failed experiment, which the Riksbank eventually abandoned.
His experience as a policymaker led Svensson to recraft how he thinks about the objective of the central bank and explicit role that unemployment should have in that thinking. He hasn’t, of course, changed by one iota his belief that in the long-term the level of real variables is determined by a whole bunch of regulatory, demographic etc factors, but not by monetary policy. He reflected on these issues a couple of years ago in a lengthy lecture, Some Lessons from Six Years of Practical Inflation Targeting (of which only the first 10 pages are directly relevant to this post), and in another article How to weigh unemployment relative to inflation in monetary policy?
Flexible inflation targeting involves both stabilizing inflation around an inflation target and stabilizing the real economy. A clear objective for monetary policy contributes to monetary policy being systematic and not arbitrary. Furthermore, for central-bank independence to be consistent with a democratic society, it must be possible to evaluate monetary policy and hold the central bank accountable for achieving its objective. This requires that the degree of achieving the objective can be measured. A numerical inflation target allows target achievement with regard to inflation to be measured and the central bank to be held accountable for its performance regarding inflation stabilization. But if monetary policy also has the objective of stabilizing the real economy, that part of the objective must also be measurable, in order for monetary policy to be evaluated and the central bank be held accountable. Given this, how should stabilization of the real economy be measured?
Stabilization of the real economy can be specified as the stabilization of resource utilization around an estimated sustainable rate of resource utilization, accepting the conventional wisdom that the sustainable rate of resource utilization is determined by nonmonetary factors and not monetary policy and therefore has to be estimated. But how should resource utilization be measured? More precisely, besides inflation, what target variable (or variables) should enter the monetary-policy loss function? One can answer this question by interpreting the legislated mandate for monetary policy and by examining what economic analysis suggests about a suitable measure of resource utilization.
In Sweden, the Riksbank’s own act mentions only price stability. But
The Riksbank’s mandate for monetary policy follows from the Sveriges Riksbank Act 1988:1385 and the preparatory works of the Act, the Government Bill 1997/98:4 to the Riksdag (Swedish Government 1997) that contained the proposal for this legislation. In Sweden, the preparatory works of laws carry legal weight, since they contain guidance on how the laws should be interpreted. According to the Riksbank Act, the objective of monetary policy is “to maintain price stability.” The Bill further states (p. 1): “As an authority under the Riksdag, the Riksbank should, without prejudice to the objective of price stability, support the objectives of the general economic policy with the aim to achieve sustainable growth and high employment.”
(I didn’t know this when in 2014 we wrote a Reserve Bank Bulletin article on the statutory goals for monetary policy in a range of countries, the Swedish entry in which thus should thus be discounted, or read in the light of these Svensson comments.)
The idea in the Bill is hardly that there is any conflict or tradeoff between sustainable growth and high employment. Furthermore, for many years Swedish governments have emphasized full employment as the main objective for general economic policy. Also, in this context, high employment should be interpreted as the highest sustainable rate of employment, if we accept that monetary policy cannot achieve any level of unemployment and that the sustainable rate of employment is determined by nonmonetary factors. According to this line of reasoning, the Riksbank’s mandate for monetary policy is price stability and the highest sustainable rate of employment.
In practice, he argues that the unemployment rate – and in particular the gap between the actual unemployment rate and the long run sustainable rate of unemployment (LSRU, determined by those non-monetary factors) should be the focus. 15 years ago his focus was on the output gap but
What does economic analysis say about the output gap as a measure of resource utilization? Estimates of potential output actually have severe problems. Estimates of potential output requires estimates or assumptions not only of the potential labor force but also of potential worked hours, potential total factor productivity, and the potential capital stock. Furthermore, potential output is not stationary but grows over time, whereas the LSRU is stationary and changes slowly. Output data is measured less frequently, is subject to substantial revisions, and has larger measurement errors compared to employment and unemployment data. This makes estimates of potential output not only very uncertain and unreliable but more or less impossible to verify and also possible to manipulate for various purposes, for instance, to give better target achievement and rationalizing a particular policy choice. This problem is clearly larger for potential output than for the LSRU.
Compared to potential-output estimates, estimates of the LSRU are much easier to verify, more difficult to manipulate and can be publicly debated. Independent academic labor economists can and do provide estimates of the LSRU and can verify or dispute central-bank estimates. Several government agencies have labor-market expertise and provide verifiable estimates of the LSRU. One could even think of an arrangement where an independent committee rather than the central bank provides an estimate of the LSRU that the central bank should use as its estimate, to minimize the risk of manipulation by the central bank. Furthermore, unemployment is better known and understood by the general public than output and GDP.
Most importantly, it has much more drastic effects on welfare. As expressed by [academic labour economist, and former Bank of England Monetary Policy Committee member] Blanchflower (2009):
Unemployment hurts. Unemployment has undeniably adverse effects on those unfortunate enough to experience it. A range of evidence indicates that unemployment tends to be associated with malnutrition, illness, mental stress, depression, increases in the suicide rate, poor physical health in later life and reductions in life expectancy. However, there is also a wider social aspect. Many studies find a strong relationship between crime rates and unemployment, particularly for property crime. Sustained unemployment while young is especially damaging. By preventing labour market entrants from gaining a foothold in employment, sustained youth unemployment may reduce their productivity. Those that suffer youth unemployment tend to have lower incomes and poorer labour market experiences in later life. Unemployment while young creates permanent scars rather than temporary blemishes.
When unemployment rises, the happiness of both workers and non-workers falls. Unemployment affects not only the mental wellbeing of those concerned but also that of their families, colleagues, neighbours and others who are in direct or indirect contact with them.
Thus, I think there are strong reasons to use the gap between unemployment and an estimated LSRU as the measure of resource utilization that the central bank should stabilize in addition to stabilizing inflation around the inflation target.
Svensson proposes reduces all this to a “loss function”, to which, in principle at least, central bank monetary policy decisionmakers can be held to account, with formal weights attached to each of the inflation gap (from target) and the unemployment gap (from the LSRU).
Personally, I think he is rather unrealistic in supposing such a formulation is possible, at least as the basis for formalised accountability. But if it is practically challenging (or even impossible), the sort of analysis he advances here isn’t unorthodox or out of the mainstream. It is simply one plausible extension of the conventional economics of modern monetary policy, from one of the leading contributors to the academic literature (and someone who has himself been exposed to the real world challenges of policymaking).
I don’t know specifically what Svensson would make of the current debate in New Zealand, or of what the Labour Party (at quite a high level of generality) is proposing. What we do know is that Labour is proposing nothing nearly as specific or formal as Svensson argues for: there would be no numerical unemployment target or an official external assessment of the NAIRU (or LSRU). My impression would be that his reaction would be along the lines of “well, of course the unemployment rate – and short to medium term deviations from the long-run level, determined by non-monetary factors – should be a key consideration for monetary policymakers; in fact it is more or less intrinsic to what flexible inflation targeting is”. He might suggest there are already elements of that in the PTA, but that making it a little more high profile, with an explicit reference to unemployment, might be helpful. I might be wrong about, but it could be worth Robertson or his advisers getting in touch with Svensson – who retains an interest in New Zealand, and gave a paper here only a couple of years ago – and asking.