RBA Annual Conference – 2000 Discussion

1. Colin Rogers

In this paper Robert Leeson offers what he calls a ‘dynamic framework’ for policy analysis. He uses the term ‘dynamic’ not in any technical sense but, in a sense akin to what McCloskey (1994) called the ‘rhetoric of economics’ – the battle for the hearts and minds of fellow economists and policy-makers. We are introduced to a world of cynical, campaigning and revolutionary scholars, who study strategy and tactics in an effort to influence their peers. They also interact in a political market place with those most unstable of characters – politicians. Some of the ideas in the paper also intersect with the work by Boland (1979) and Hoover (1984) on monetarist methodology.

Hence, as I read it, the paper is mainly about the art of persuasion as it applied to monetarism and the natural rate hypothesis. It deals with the influence that these ideas had on macroeconomic policy in Australia, largely in the 1970s and 1980s. This is an important dimension to the policy debate and deserving of serious attention – particularly if it provides some insight of lasting value about potential pitfalls in debate about macroeconomic policy. The question I found myself asking then was this: does use of the ‘dynamic framework’ lead to durable and fruitful insights into the Australian policy debate? My answer on completion of the paper was ambivalent – maybe.

My assessment is ambivalent for two reasons. First, because the conclusions presented on the unreliability of ‘conventional wisdom’, the limitations of mathematical formalism and econometric analysis and the susceptibility of economists to self-deception are generally well-known. I don't believe we need the ‘dynamic framework’ to reach them. Second, although I found many of the stories and insights interesting, the emphasis was a little too historical and not sufficiently analytical for my taste. From a topic such as this I would be looking for some basic principles that might be fruitfully applied to the 1990s and beyond. What lessons can we learn from these debates that will enable us to avoid sterile arguments in future?

In that respect, and despite my ambivalence, I think there is an important issue to which much of the discussion in this paper alludes but which it does not quite succeed in bringing into focus. That is the question of conceptual congruence between analytical concepts and their real world institutional counterparts. Academic input into policy debate is often plagued by this problem and that appears to be especially the case with the topics discussed in this paper. To make the point I will briefly provide two examples of what I mean from the recent literature:

  1. The question of inflation bias by governments and/or central banks and the associated issue of central bank independence.
  2. The evolution of monetarism from academic monetarism, via pragmatic monetarism to inflation targeting. The latter was surely one of the most significant changes to Australian macroeconomic policy to occur in the 1990s.

Inflation bias

The question of supposed inflation bias by central banks is an issue of the 1990s that seems to me to be ideally suited to scrutiny from Robert's ‘dynamic framework’ perspective. I have in mind here McCallum's (1997) claim that this literature suffers from two fallacies that arise from conceptual rather than logical errors. McCallum's point is that these fallacies arise because of inappropriate mappings between analytical constructs and real world institutions. The message is obviously not new but it is a timely reminder that although academic economists are often strong on logic they can be rather vague on institutional detail. McCallum argues convincingly that central bankers will simply not behave as postulated in the model. In the opinion of McCallum (and others) inflation bias on the part of governments and/or central banks is a pseudo problem (at least in democratic societies). The empirical evidence on the relationship between central bank independence and macroeconomic performance can hardly be described as robust (see Fuhrer (1997)).

Evolution of monetarism

Another area where much the same concerns arise is the interaction between academic and pragmatic monetarism. It could be argued that the evolution of monetarism from academic (money base, fixed money growth rates), to pragmatic (monetary targets), and ultimately to inflation or price level targeting, also illustrates the problem of inappropriate mappings from analytical constructs to real world institutions identified by McCallum.

For example, in a recent manuscript, Pepper and Oliver (2000) define a ‘pragmatic monetarist’ as ‘… someone who, as time has passed and practical experience has been gained, whilst still accepting the theory of monetary control, has concluded that the theory cannot be turned into working practice’. This is tantamount to conceding that monetarism works in theory but not in practice – a concession that is usually regarded as the hallmark of a ‘poor’ theory. The qualifier ‘poor’ is attached precisely because the conceptual mapping from Inline Equation to monetary policy has always posed problems for central bankers who inhabit a world where interest rates are the instrument of policy. Fortunately it is now generally recognised that it is not operational (feasible given existing institutions) to ask central bankers to fix M or its growth rate, and most modern macroeconomic models have no role for Inline Equation (see for example, Romer (2000) and Henckel, Ize and Kovanen (1999)). Instead these models contain a simple monetary policy reaction function in which interest rates are adjusted to achieve some nominal target (inflation or nominal GDP).

The evolution of these ideas clearly raises interesting questions about the relationship been monetarism and inflation targeting. What is the relationship if any? Some would argue that the question of M control is the essence of monetarism and consequently it is not useful to describe either inflation-targeting or modern macroeconomics as monetarist. In that respect, DeLong (2000) recently suggested that the lasting analytical contribution of monetarism is not the proposal for monetary control but the natural rate doctrine and policy-makers' aversion to old-style Keynesian fine-tuning. On this view monetarism lives on under another name – even in New Keynesian economics!

Whatever one's views on these matters, and I don't intend to adjudicate the debate here, I would have expected the ‘dynamic framework’ presented in the paper to focus attention on this type of issue. In my view both of these examples illustrate the importance of examining the conceptual mapping between theoretical constructs and real world institutions. (As a matter of interest Friedman's as if methodology probably made a negative contribution to debate on this issue.) These issues are important because they have implications for the way academics and policy-makers sell their policies to politicians and the public. Much frustration in debate between academics and policy-makers might thereby be avoided.

To sum up. In my view this paper has a lot in common with the literature on the rhetoric of economics and the methodology of monetarism. I think that the underlying idea behind the paper is interesting but the promise is largely unrealised. In particular I don't think the ‘dynamic framework’ does enough to highlight the type of important conceptual problem illustrated by the two examples provided above. After reading the introduction I expected the paper to have a comparative advantage in that area and I think much of the discussion would benefit if interpreted from this perspective. These sorts of conceptual problems are important because they have befuddled macroeconomic debate over the past few decades and will no doubt continue to do so. In a sense they are just as important as the technical theoretical and econometric issues so I think Robert is looking in the right place even if the paper doesn't take the line I was expecting.

References

Boland LA (1979), ‘A Critique of Friedman's Critics’, Journal of Economic Literature, 17(1), pp 319–337.

DeLong JB (2000), ‘The Triumph of Monetarism?’, Journal of Economic Perspectives, 14(1), pp 83–94.

Fuhrer JC (1997), ‘Central Bank Independence and Inflation Targeting: Monetary Policy Paradigms for the Next Millennium?’, New England Economic Review, pp 19–36.

Henckel T, A Ize, and A Kovanen (1999), ‘Central Banking Without Central Bank Money’, IMF Working Paper 99/092.

Hoover KD (1984), ‘Two Types of Monetarism’, Journal of Economic Literature, 22(1), pp 58–76.

McCallum BT (1997), ‘Critical Issues Concerning Central Bank Independence’, Journal of Monetary Economics, 39(1), pp 99–112.

McCloskey DN (1994), Knowledge and Persuasion in Economics, Cambridge University Press, Cambridge.

Pepper GT and MJ Oliver (2000), ‘Monetarism Under Thatcher – Unfinished Business’, unpublished manuscript.

Romer DH (2000), ‘Keynesian Macroeconomics Without the LM Curve’, National Bureau of Economic Research Working Paper Series No 7461.

2. General Discussion

The general discussion of Leeson's paper centred mainly around two issues: the relevance of the ‘dynamic analysis’ framework for policy analysis, and the influence of monetarism on macroeconomic policy in Australia.

There was general agreement with Leeson's basic proposition that careful examination of the historical origin and evolution of ideas would enrich contemporary policy analysis. Many participants, for instance, noted that a better understanding of the nature of past intellectual debates did indeed provide valuable insights into new policy agendas. However, some felt that Leeson had unfairly underplayed the usefulness of formal analysis for policy. It was also noted that the limitations of formal and econometric models highlighted by Leeson are generally well understood and accepted in the profession, and that the dynamic analysis framework did not necessarily add value in this regard. One participant noted that since most economic relationships are complex, a lack of formal analysis could lead to incorrect policy inferences. Another agreed that formalist models might better explain economic relationships, but noted that these models nonetheless have limited impact on policy, especially microeconomic policy. It was suggested that one reason for this might be that academics are unable to effectively communicate abstract concepts to policy-makers. Taking issue with the implicit assumption in this discussion that all academics are interested in policy, one participant made the point that to some, research is an intellectual exercise which need not have direct policy application.

In discussing the influence of monetarism on Australia's monetary policy, one participant remarked that Australia's adoption of the monetarist framework in 1976 reflected a response to the perceived ineffectiveness of the earlier Keynesian approach. The monetarist approach appealed to the authorities at the time as it seemed well-suited to re-establishing price stability at as low cost as possible to economic activity. It was noted that monetary targeting was eventually abandoned in 1985 as increased instability of the velocity of money led to a breakdown of the relationship between money and nominal income. One participant made the observation that along with the floating of the dollar in 1983, the gradual transition from monetary targeting to inflation targeting had been one of the most significant developments in the Australian macroeconomy over the last few decades. Another noted that the shift to the inflation-targeting framework did not necessarily constitute a total departure from the monetarist approach, but what had changed was the articulation of the explicit objective of monetary policy. The question of whether the inflation-targeting framework was more robust than monetary targeting was also raised. Inflation targeting was seen by many as being more robust as it does not rely on the stability of a single simple relationship, and focuses directly on the final objective of monetary policy. A few participants argued that the answer to this question depended on the nature of the shocks, and that the inflation-targeting framework is relatively better at dealing with demand shocks.