Discussion by Athanasios Orphanides, Governor of the Central Bank of Cyprus, at the Fifth High-Level Seminar of the Eurosystem and Latin American Central Banks on "Global Rebalancing, Asset Prices and Policy Responses"
Madrid, 9 December 2010
It is a great pleasure to discuss this paper on the changes, achievements and challenges of central banking in Latin America by Klaus at this Eurosystem and Latin American central banking gathering. I’m going to start by noting that Klaus and I have something in common: we are both Rudi Dornbusch students. As a student of Rudi during the 1980s, I learnt quite a bit about the monetary experiences of Latin America. Rudi routinely used current events and present crises as case studies to explain the success and failure of monetary policy regimes and institutional arrangements. He also used examples to contrast politically induced short-termism against policies that would enhance welfare over the long haul. During the 1980s there were indeed plenty of examples. However, not all of the examples were from Latin America. Some concerned what is now the euro area. For example, after researching debt and deficit data for Rudi at the end of the 1980s, it was hard not to be impressed by the challenges then facing Belgium. I mention this to cheer us up a little bit because it/s never pretty when events force us to refresh our knowledge of debt deficit dynamics, but it’s reassuring to know there are examples of countries that managed to come back from the brink.
Turning back to Latin America, the focus of Klaus’ paper, some knowledge of historical experiences is, I think, very useful to gain a proper appreciation of the truly remarkable advances in central banking practice that have been observed in the region over the last two decades. In his survey, Klaus does an excellent job of explaining key changes and achievements, and then ends the paper by discussing some challenges for central banks both in Latin America and elsewhere. In my discussion I will briefly focus on two areas: first, on the sources of the improvement seen in Latin American central bank practice that Klaus discusses, and the inflation targeting framework of monetary policy that he uses quite often in that regard; second, I will talk a little bit about some of the challenges he identifies going forward.
Regarding the changes in central banking in Latin America, Klaus uses the inflation targeting (IT) framework of monetary policy as an organising device for much of his discussion, even for non-inflation targeting countries. He considers it to be a useful device to that end and I find this quite helpful as well. But I must stress that I have doubts about the view that the IT framework is necessarily superior to other frameworks. For example, Klaus suggests that IT may even be superior to the price stability approach followed by the ECB. I would say that what matters most is an institutional design that retains some crucial characteristics that we can all identify as being essential for good policy practice. What I think is going on is that those crucial characteristics that Klaus does identify are actually shared both by the ECB approach and the canonical inflation targeting framework that he discusses. Here I very much agree with Klaus on the bottom line in describing the changes and achievements in Latin America: namely, that we have observed improvements in central banking practice that have placed many Latin American central banks in the group of central banks that can claim to be very close to the benchmark of best practice, precisely because these key characteristics have been embedded in the policy framework. This couldn’t have been further away from being the case back in the 1980s and is thus quite an achievement.
Now what are the most important elements for best practice? In my view there are two. Central bank independence and the clear commitment by both the state and the central bank that price stability is the primary mandate of the central bank. Independence must be both in its legal and economic form, and Klaus presents very nice evidence on how this has worked for Latin America. He has transparency in policy high on his list, and I agree with him. However, I will give you a somewhat different interpretation of the aspect of transparency that is most important. Klaus interprets transparency as policy bound by ex ante rules and monitored with ex post accountability. I very much favour rules as guides for policy discussion, but I believe that the key here is a clear explanation of the monetary policy framework, including the goals of policy and how this is going to be achieved. This interpretation of the rule is a little bit broader than what Klaus has in mind. The key focus should be on the one predominant objective that monetary policy can achieve. That is, the crucial feature regarding transparency is what the price stability objective is for the central bank, and how the central bank pursues it. This is the broader definition that I interpret as equivalent of what Klaus calls the rules approach. The most important element of good policy practice by an independent central bank, regardless of whether it is called an inflation targeting central bank or not, is a monetary policy framework that focuses maximum attention on the objective of price stability and indeed forces close monitoring of current and prospective aggregate prices, both as a means to guiding current policy and as a means to evaluating past policy action. I take this to be the most distinguishing characteristic of inflation targeting and also of the ECB strategy, even though the ECB is not an inflation targeting central bank. By encouraging an ongoing open dialogue between the central bank, the government, the public and financial market participants, the inflation targeting approach leaves little room for neglecting price stability, further reinforcing its unique focus. It is for these reasons that inflation targeting may be particularly effective as a monetary policy framework for central banks that are institutionally challenged in some way before they adopt it. For instance, because they lack a history of political independence or because they have an impaired credibility in pursuing monetary stability-oriented policies. Latin America in the 1980s fits this description—the initial condition that would benefit most by moving in that direction.
The intrusion of politics into monetary policy decisions as well as the pursuit of multiple and possibly conflicting objectives are potential sources of such impaired credibility with regard to a central bank’s commitment to achieving and maintaining price stability. Inflation targeting helps guard against these forces. Klaus argues, and I am in full agreement, the reason the focus on a clear definition of price stability is crucial is that by being so transparent on this, the central bank can help anchor inflation expectations in the best possible manner. This in turn has multiple benefits for the central bank. For example, it allows the central bank flexibility to respond to other disturbances in the economy thus not only achieving and maintaining price stability but also contributing to economic and financial stability. This action is what Klaus demonstrates with the evidence he provides when he compares the response of central banks in Latin America and economic outcomes that are associated with the 2008 crisis and with the others that occurred ten years earlier. The comparison is very instructive on the benefits of having these institutional improvements. As he shows domestic policy helped soften the blow to the economy in the most recent episode, and that was not the case at all in the episode ten years earlier. This is a very useful demonstration that best practice with an independent central bank, and with a clear definition of price stability as its mandate, does indeed deliver better outcomes.
I am somewhat more ambivalent about Klaus’ use of adjectives to describe inflation targeting, i.e. "partial" versus "fully fledged" IT, even what he calls "flexible" IT. I also disagree with his claim about another adjective – "explicit". Specifically, I do not agree with him when he says "that explicit IT dominates other successful monetary regimes … like those pursued by the U.S. Federal Reserve, the European Central Bank …". Hence to the extent that a useful metric of success is how well-anchored inflation expectations are, I do not interpret the evidence as suggesting that the performance of IT central banks dominates that of the ECB. My greatest concern here is with the so-called flexible inflation targeting framework, used to target not just inflation but multiple objectives at the same time. Klaus recommends that central banks discuss all the unobserved variables someone would put in a model forecast to design optimal control and IT-type policies. This includes very explicitly the natural rate of interest, the natural rate of output, the natural rate of unemployment, the natural or equilibrium exchange rate, the natural or equilibrium commodity prices, and so on. My fundamental disagreement here is that I would argue that robust policy should stay clear from such concepts as much as possible to achieve best results regarding price stability. The problem with an approach that allows or even encourages policy measures to try to stabilise output and unemployment in addition to maintaining price stability is that it can easily transform into the sort of fine-tuning approach that achieves none of its multiple targets well. Let us remember that we experienced such failures in the past and, indeed, a starting point for understanding the useful elements of inflation targeting for me is to comprehend the sources and magnitude of the failure in some of the countries that adopted it. I find the case of New Zealand, the pioneer of the approach, particularly instructive.
Prior to the mid-1980s, New Zealand had the unenviable record of one of the highest rates of inflation in the OECD. Inflation exceeded 10% per annum for almost an entire decade. According to Don Brash, the Governor who implemented inflation targeting at the Reserve Bank of New Zealand, the problem with the price stability mandate was that it was only one of several goals. As he put it: "The legislation under which we operated required us to have regard for the inflation rate, employment, growth, motherhood and a range of other good things" (Brash, 1999, p.36). The Reserve Bank was also hampered by its lack of operational independence.
These weaknesses were corrected by the Reserve Bank of New Zealand Act, 1989. The price stability remit was given prominence. As section 8 of the Act states: "The primary function of the Bank is to formulate and implement monetary policy directed to the economic objective of achieving and maintaining stability in the general level of prices". No ifs, no buts, no other things. Section 9 of the Act requires a numerical target for inflation to be agreed between the Governor and the Minister of Finance, and section 10 ensures the Bank’s operational independence.
The Reserve Bank of New Zealand Act, 1989 thus describes the two defining characteristics of the economic and inflation targeting approach where we started from. First defining a hierarchical mandate for the central bank with price stability, a clear definition of price stability, and second, providing the central bank with the independence to pursue this objective. And this is what many others followed in one way or another. These are precisely the two characteristics that Klaus stressed as crucial and which are common to the institutional design of the ECB.
Let me now turn to Klaus’ discussion of some challenges for the future. I would focus on just three elements.
First, the implications of the zero lower bound on the short-term nominal interest rate. Here Klaus seems to take it for granted that monetary policy is constrained by the zero lower bound, and thus explores ways to reduce the likelihood of hitting the lower bound. For example, he talks about the possibility of raising the numerical definition of price stability which, as he recognises, creates other problems. On this point the underlying premise is, in my opinion, flawed. It is based on the so-called liquidity trap, suggesting that no additional monetary policies can be implemented once the short-term nominal interest rate is close to zero. But as we know, for example from Brunner and Meltzer’s arguments from the 1960s (Brunner and Meltzer, 1968) and by others since, this is a conceptual mistake. The liquidity trap concept is seriously flawed because it can only be proven to hold in a model under some unrealistic assumptions, and is understood to be false once those assumptions are relaxed. Such academic exercises can be useful to sharpen discussions, but they can become quite harmful if they permeate into policy debates. In practice, non-standard measures can be employed to engineer additional monetary policy easing if and when needed at the zero bound, and there is no need to abandon a definition of price stability that has otherwise been deemed to be perfectly fine just in order to protect against the possibility of the zero bound.
The second element is whether a central bank should lean against the wind to protect against nascent imbalances. Here the discussion in the paper needs some clarification. The answer may be different depending on whether the central bank has, in addition to monetary policy, a role in prudential supervision or not. Klaus actually alludes to this, making the distinction between macro-prudential measures and monetary policy. In my view, central banks should have banking supervision responsibilities also, but not all central banks do. For those central banks that do have the additional tools that can be used for macro-prudential purposes, then clearly those tools must be incorporated into the broader stability-oriented policy design.
There are two examples that help illustrate how the impact of the credit boom prior to the most recent crisis could have actually been reduced by central banks employing prudential supervision tools. One is the practice of dynamic provisioning, i.e. asking banks to raise provisions in good times. The boom may thus be tempered and bigger shocks can be absorbed in bad times. This worked well in Spain. The second is macro-prudential adjustment of loan-to-value ratios. This was adopted with some success in my own country, Cyprus, in 2007 and 2008. In July 2007, one month before the onset of the crisis, while the real estate party was in full swing, we tightened conditions on real estate loans, capping the maximum loan-to-value ratio, except for owner-occupied housing, to just 60%. Some developers took the hint and cut back on their plans. Others complained quite a bit and we faced a lot of criticism, which was evidence that the policy was effective. This policy action was one of the reasons why our banking system was shielded from the shocks and aftershocks of what happened following September 2008. Although we had a real estate price boom it was tempered and the banking sector was protected from it.
I will conclude with a remark about another challenge raised by Klaus: fiscal policy. Here I urge Klaus to flesh out more the successes in Latin America. I find particularly important the discussion regarding the establishment of fiscal councils and fiscal rules. Klaus is certainly right; the way forward is conceptually straightforward but politically fraught with difficulties. This appears to be a crucial missing link in designing an institutional framework that ensures stability going forward.
-------------------------------------------------------------
References:
Brash, Donald (1999) "Inflation targeting: an alternative way of achieving price stability", address on the occasion of the 50th Anniversary of Central Banking in the Philippines, Manila, 5 January.
Brunner, Karl and Allan H. Meltzer (1968) "Liquidity traps for money, bank credit and interest rates", Journal of Political Economy, 76(1):1-37.
Reserve Bank of New Zealand (1989) Reserve Bank of New Zealand Act.