RBA Annual Conference – 1990 Discussion

1. Ian R. Harper^[*]

In any assessment of developments in the Australian macro-economy in the 1980s, the Australian monetary and financial system must figure prominently. By the same token, to describe and analyse the rich tapestry of events, in the world of ideas as well as the world of affairs, over this period is a task of Herculean proportions. We are indebted to Ross Milbourne for his labours; but, impressive though his stature is as a monetary economist, he is no Hercules. This, of course, is fortunate for me. Had the Augean stables been completely clean, I might have had difficulty justifying the expense of bringing me here to inspect them!

Ross Milbourne's paper spans a wide range of issues. I will provide my general impressions and then address a number of specific points (i) which I feel have not been adequately canvassed in the paper or (ii) with which I disagree or (iii) which I feel are particularly well handled by Ross.

To begin, I confess that my overall reaction to the paper was one of frustration. I feel that Ross has been too even-handed, that he is reluctant to indicate where he stands on an issue and that on too many occasions he simply leaves the reader to make up his or her mind on the basis of the logic and evidence he presents. This is not to say that the informed reader will not in any case form an opinion based on his or her personal assessment of the various arguments; nor is it to suggest that Ross's brief was other than to write an impartial survey of the field. But even for a reader familiar with most of the literature traversed in the paper, it would have been helpful to have a clear statement of Ross's position on a number of issues, if only as a pole to react against.

In addition, I feel that the paper places a little too much emphasis on debates within the theory of monetary policy, especially as informed by experience outside Australia, at the cost of a more detailed discussion of the formation of monetary policy in Australia. Much of the richness of recent monetary history in Australia, especially the uncertainty and excitement associated with the transition from the regulated to the deregulated era, seems to be lacking. The story as Ross presents it, at least to my taste, is too clinical; the revolutionary nature of our monetary experience post-Campbell is understated.

In making this point, I do not wish to imply that the theoretical debates were of secondary importance or that they did not inform developments in the formation of monetary policy. I merely wish to express mild disappointment that the opportunity to delve into the policy debate in greater detail has not been taken up by Ross on this occasion. My disappointment is assuaged to some extent by the fact that Jeffrey Carmichael in his paper to this conference critically assesses the role of monetary policy and the attitudes of the monetary authorities in the formation of anti-inflation policy in the 1980s.

I might illustrate the type of issue which I believe Ross could usefully have explored in greater depth by moving to my first specific point of omission. It is generally acknowledged that financial deregulation has strengthened the hand of the Reserve Bank in the implementation of monetary policy. It has also created conditions under which the Reserve Bank can act with greater independence from the Commonwealth Treasury. The whole question of central bank independence, including, in particular, the suggestion of Peter Jonson^[1] and others that the Reserve Bank Act be re-drafted to guarantee greater statutory independence for the Bank, is ignored in the paper.

This is especially curious given the recent bold move by the government of New Zealand to grant a large measure of statutory independence to the Reserve Bank of New Zealand. Some of the first fruits of that experiment are now being tasted.^[2] An examination of this experience from an Australian perspective would have made fascinating reading,^[3] as would a discussion of the changing roles of the Reserve Bank and the Commonwealth Treasury in the formulation and implementation of monetary policy.

My second specific criticism concerns Ross's discussion of the factors leading up to financial deregulation. This discussion fails, in my opinion, to grasp the heart of the matter. While the paper mentions the pressure brought to bear on the government to deregulate, pressure both from the banks and from the monetary authorities, nowhere is there any suggestion that this was a surprising or unusual outcome. Indeed, the fact that the recommendations of the Campbell Committee were adopted so completely and within such a short time-period (in spite of a change of government and the establishment of a second committee of inquiry) passes without remark. It does not strike the author as worthy of comment that the experience of the Campbell Inquiry stands in utter contrast to that of every other major committee of inquiry in recent Australian history, not least in the degree of unanimity amongst the different interest groups, private and public, appearing before it.

I believe the experience of financial deregulation, both in Australia and around the world, owes a great deal to the endemic capacity of financial markets to generate close substitutes for existing financial products and processes, especially with the assistance of microprocessor technology and the goad of onerous regulations.^[4] This observation has implications for the capacity of the authorities to re-impose financial regulations at some future time and for the long-term effectiveness of monetary policy, points raised briefly by the author towards the end of the paper.

A further omission from the paper is a detailed discussion of the impact of deregulation on the techniques of implementing monetary policy. The author certainly mentions the prominence necessarily assumed by open market operations and discusses at a general level the implementation of policy through the cash market. But, in my view, the revolutionary nature of the change in operating procedures is inadequately recognised, especially the reliance the Reserve Bank now places on its statutory role as the monopoly supplier of cash. The future effectiveness of monetary policy rests on the Bank's ability to preserve its monopoly status and, in particular, to guarantee a downward-sloping demand curve for its liabilities.

This highlights the importance of institutional developments for monetary policy. Traditional discussions of monetary policy were largely dismissive of the role of institutional factors in the transmission mechanism. Deregulation has brought institutional arrangements to the foreground. For example, an understanding of the purely institutional distinction between banks and non-banks and between different classes of bank liabilities is crucial for the correct interpretation of recent developments in monetary aggregates. Fortunately, the traditional disdain for institutional detail evinced by monetary economists appears no longer to prevail within the Reserve Bank, as a number of recent papers indicate.^[5]

I feel that Ross has been kinder to the Bank than might be warranted on certain issues. For example, there is no discussion of the steady increase in the extent and frequency of intervention in the foreign exchange market by the Reserve Bank since the floating of the Australian dollar in December 1983. This is a controversial issue and the Bank no doubt has its side of the story to tell. But such a seeming contradiction of earlier grounds for abandoning exchange rate management deserves critical investigation and analysis – not least in light of the work of McTaggart and Rogers^[6] who suggest that such intervention has given monetary policy a pro-cyclical bias.

Then there is the ‘check list’; in my opinion, a piece of political subterfuge of which Sir Humphrey Appleby would be proud! Charles Goodhart in his recent survey of monetary policy commented as follows:

“Supporters would describe it [the ‘check-list’] as sensible pragmatism; detractors as a reversion to a muddled discretion, which, once again, allows the authorities more rope than is good for them, or us.”^[7]

Again, I feel that Ross has missed an opportunity to make some salient points about the influence of such obfuscation on the credibility of monetary policy and the role of the Reserve Bank.

I am afraid I cannot accept Ross's easy dismissal of the relationship between the money base and nominal magnitudes. I concede that it may be difficult to establish empirically, although I do not believe that the velocity of base money is as unstable as Ross implies, at least not since the float. Ross bases his argument on a comparison of the quarterly growth rate of the money base with the quarterly growth rate of its income velocity (the exact definition of income used is unclear). He shows that these quarterly changes are almost perfectly negatively correlated. But the growth rate of velocity is simply the difference between the growth rates of nominal income and the money base. So if the growth rate of velocity is equal to minus one times the growth rate of the money base, as Ross claims, this implies that the growth rate of his nominal income variable is zero. Furthermore,^[8] if the growth rate of the nominal income variable is zero (or close to zero), any monetary aggregate would display the behaviour illustrated by Ross in his Figure 17.

McTaggart and Rogers^[9] show that the velocity of the money base has been much more stable since the float and that it is not markedly less stable than the velocity of higher aggregates. In any case, it cannot be denied that the rate of growth of the money base must in the longer run determine the rate of growth of nominal magnitudes. The unit in which the money base is denominated, viz., the dollar, is the unit of account. So long as demand for the money base is stable (and this is where institutional arrangements play a major part), controlling the supply of base money will control the price level. Ross does point out that the demand for base money may not be assured (and this is important) but he gives us no reason to believe that the demand varies so closely with supply as to render control of the price level and nominal magnitudes impossible. At least he is prepared to concede (p. 260) that, “ … the lack of a relationship [between the money base and] prices and income is surprising”!

His concern over the relationship between the money base and nominal magnitudes colours his subsequent discussion of money base targeting. Concern over the ability to influence prices and nominal income in a timely fashion should strengthen rather than weaken the case for money base targeting. The points he raises in regard to the capacity of financial innovation to undermine the demand for base money are well taken but he ignores the role of reserve requirements. This is surprising in view of the prominence given in the theoretical literature to reserve requirements as instruments of monetary policy in a “deregulated” environment.

Let me conclude on a positive note. I thoroughly endorse Ross's identification of the fundamental tension between monetary policy and the wage-fixing mechanism in Australia, of the tendency of Australian governments of both persuasions to rely on monetary policy as a substitute for more far-reaching and politically-demanding micro-economic reform, of the unlikelihood that monetary policy can do anything to improve the current account and external debt “problems” and of the absurdity of the Reserve Bank's perceived need to justify its actions month by month. Ross's discussion of these points is masterly.

I have been harsher in my criticism than I had hoped to be. But this should in no way diminish the paper as a signal (even if not Herculean!) contribution to the discussion of monetary policy in Australia.

Footnotes

University of Melbourne. The author gratefully acknowledges helpful discussions with Jeffrey Carmichael, Doug McTaggart, Jim Perkins and Peter Stemp during the preparation of these comments. Sole responsibility rests with the author, however. [*]

Jonson, P.D., “The Case for an Independent Reserve Bank”, Address to the Sydney Institute, 26 October 1989, Sydney. [1]

Frew, W., “NZ Govt loses control of monetary policy to Reserve”, The Australian Financial Review, Thursday, June 14 1990, p. 3. [2]

A useful starting point for such an exercise is Buckle, R.A. and Stemp, P.J. “Reserve Bank Autonomy and Government Objectives in New Zealand: Can Tensions be Resolved?”, Centre for Economic Policy Research Discussion Paper Number 215, Australian National University, August 1989. [3]

See Harper, I.R. “Why Financial Deregulation?”, The Australian Economic Review, 1st Quarter 1986, pp. 37–49. [4]

Grenville, S.A., “The Operation of Monetary Policy”, The Australian Economic Review, 2nd Quarter 1990, pp. 6–16. [5]

McTaggart, D. and Rogers, C., “Monetary Policy and the Terms of Trade: A Case for Monetary Base Control in Australia?”, The Australian Economic Review, 2nd Quarter 1990, pp. 38–49. [6]

Goodhart, C., “The Conduct of Monetary Policy”, The Economic Journal, Volume 99, Number 396, June 1989, p. 334. [7]

I am indebted to Doug McTaggart for this observation. [8]

See footnote 6. [9]

2. Peter J. Stemp^[*]

This comment highlights two subjects that have been central to monetary policy in the 1980s. First, important issues that have arisen post deregulation are analysed. Secondly, monetary targeting – which has received overwhelming rejection in Milbourne's paper – is examined. I believe his assessment has been too harsh and that the approach warrants further consideration.

(a) The Implications of Financial Deregulation

At the time of the 1979 Reserve Bank Conference,^[1] innovations in financial markets were making regulations less effective. In this climate, it was the degree of financial deregulation that was being debated, rather than whether the plethora of financial regulations should remain. One particular point of concern was whether or not financial deregulation would prove to be destabilising.

With the economy having survived greater volatility in the foreign exchange market and a world-wide stockmarket crash in October 1987, there is now little concern about the destabilising aspects of financial deregulation. At the end of a decade of change, the process of financial deregulation is all but completed. New problems have, however, arisen in the wake of financial deregulation.

Breakdown of traditional transmission mechanism

There has always been considerable uncertainty about the lags involved in using monetary policy to achieve real outcomes. Because of this uncertainty, there is a risk that tight monetary policy might be sustained for too long (a hard landing) or, alternatively, not sustained long enough, requiring further tightening of monetary policy at a later date. Now there is debate about the nature, as well as the lag structure, of the transmission mechanism. There has been limited empirical research undertaken to determine the transmission mechanism in the current financial environment. Milbourne highlights the breakdown in the transmission mechanism by observing, first, the recent failure of high real interest rates to reduce expenditure in 1988/89 and, secondly, sustained high levels of the current account deficit. Clearly, the need for more research in this area should be emphasised.

Increased importance of expectations

The removal of financial regulations has made possible a true market outcome in financial markets. In the past, financial regulation meant that many future outcomes could be reasonably well anticipated by extrapolating past policy responses. Under deregulation, forward-looking expectations that anticipate future outcomes in financial markets have become more important. A starker contrast can now be drawn between outcomes in the financial market and outcomes in the labour market, still constrained by institutionalised rigidities.

The importance of forward-looking expectations has recently been emphasised in a study using the Murphy model of the Australian economy.^[2] The forward-looking behaviour in the Murphy model is embodied in the assumption of rational (or model consistent) expectations in financial markets so that agents' expectations for future asset prices are assumed to coincide with the model's forecasts. In the model, rational expectations appear in three places in the financial sector. First, the spot exchange rate is determined assuming that expectations for future values of the exchange rate are rational. Second, the determination of the long-term bond rate assumes expectations for future values of the short-term interest rate are rational. Third, the long-term expected inflation rate is based on rational expectations for future inflation rates. The long-term expected inflation rate is used in the model to adjust the long-term bond rate to obtain a long-term real interest rate which in turn is used in the calculation of the market values of real assets.

Using this model, it is possible to examine the consequences of reducing the annual rate of monetary growth by one percentage point. Two experiments have been considered. These are distinguished only by the proposition that, in one experiment, the reduction is expected to be permanent while in the second experiment the reduction is temporary, expected to last only for a year. The table shows that, at the end of one year, the temporary change has less than half the effect on the consumer price index that a permanent change would have. The effects on gross domestic product are also significantly different.

The point for emphasis is that all forward-looking expectations in the Murphy model occur in financial markets. Since the only difference in the two experiments examined above occur in expectations about what will happen in the future, the significantly different outcomes must be transmitted solely through the financial markets. In particular, by allowing unconstrained outcomes for asset prices, removal of financial regulations can mean that the expected future policy stance will have a much more significant impact on policy outcomes than before.

There appears to be some debate at this conference about the proximate causes of inflation in the short run, reflected in Carmichael's paper. In a world with forward-looking behaviour in financial markets, monetary policy will have some effect on short-run prices provided monetary policy can affect the long-run rate of inflation. Transmission of monetary policy to long-run expectations and thence to the short run, via the expectations mechanism, is the mechanism emphasised in these Murphy model simulations.

Loss of instruments

Milbourne highlights three sets of regulations which the RBA used in the operation of monetary policy during the 1970s. The first set of regulations involved interest rate restrictions on deposits and loans of trading and savings banks. The second involved the use of quantitative lending guidelines for banks. Thirdly, the RBA used portfolio restrictions to vary the liquidity position of banks. He also highlights another consideration in the operation of monetary policy: Australia operated on a crawling-peg exchange rate – effectively very similar to a fixed exchange rate.

With the removal of financial regulations, the RBA is left with only one instrument, the cash rate. For any government that desires to use macro-economic policy intervention to ensure internal balance and, more contentiously, distributional outcomes, there has been a significant loss of macro-economic policy instruments.

In addition, the removal of financial regulation means that the agents who bear the ultimate costs of policy changes are different. A clear example is the incidence of costs associated with high interest rates for home loans. In the (regulated) past, high interest rates led to a rationing of housing finance, with the heaviest cost falling on those who were trying to enter the housing market. Recently, high interest rate costs in the housing sector have fallen with greatest force on those who had borrowed heavily to purchase a house just prior to the tightening of monetary policy. In the future, financial innovations, such as five-year fixed-interest-rate loans will be able to minimise the effects on past home buyers. However, if tight monetary policy is to be effective, costs must be borne somewhere in the community.

The loss of policy instruments must mean that monetary policy becomes an increasingly blunt instrument, with the monetary authority increasingly unable to control the ultimate incidence of costs associated with its policies. For a government eager to achieve re-election, there must be a temptation to reregulate to protect certain groups in the electorate.

There is also likely to be pressure to try to achieve too much with just one policy instrument. This can lead to increasing uncertainty about the future stance of monetary policy. With the increased importance of expectations in the deregulated environment, this uncertainty means that particular outcomes, such as reduced inflation, will be harder to achieve.

(b) Monetary Targeting

Several times, Milbourne implies that monetary policy has been successful through the 1980s because the rate of inflation has been reduced. The figure below (reproduced from Carmichael's paper at the conference) shows that such a reduction has occurred but Australia has been less successful than the OECD average. There has been a particular lack of success since the adoption of the check-list approach in January 1985.

One of the strong themes running through the paper is the rejection of any form of monetary targeting. In particular, money base targeting is singled out for strong criticism. This rejection seems to be partially based on the false assessment, noted above, that recent monetary policy has been successful in fighting inflation.

There are two other arguments that underlie Milbourne's assessment that monetary targeting is inappropriate. First, he uses the Poole^[3] analysis to suggest that “if most shocks during the 1980s were on the financial side, the interest rate should have been the intermediate target, not the money stock” (p. 259). Secondly, fixed monetary growth rules are argued to be inappropriate because they fail to take account of shifts in velocity and seasonal factors.

The first argument, using the Poole analysis, is only appropriate if monetary targeting is directed towards a short-run stabilisation objective. Yet, elsewhere, Milbourne acknowledges that the current approach to monetary policy was pro-cyclical with respect to the terms of trade during the 1985/86 period and that monetary policy can be destabilising: “[i]f monetary policy takes 12 months before its effects are felt, yet disturbances rectify themselves within 6 months, then monetary policy might stimulate the economy at precisely the wrong time, thus de-stabilising it” (p. 254). These examples illustrate that it may be undesirable to use monetary policy for stabilisation purposes.

There are a range of reasons, apart from stabilisation purposes, why monetary targeting might be usefully employed. Milbourne lists four reasons why pre-announced rules for monetary growth might be appropriate. These reasons are:

1. pre-announced rules provide a firm anchor for expectations of inflation;
2. fixed rules give clear benchmarks by which the performance of monetary authorities can be judged;
3. fixed rules reduce the incentive for time-inconsistent behaviour on the part of policy makers; and
4. pre-announced rules impose discipline on negotiators of wage bargains.

Commentators do not commonly suggest that monetary targets are best for stabilisation purposes. However, monetary targets can be supported on other grounds.

The second criticism – that monetary targets fail to take account of shifts in velocity and seasonal factors – involves a rigid interpretation of policy. It is possible to adopt a contingent approach to monetary targets that allows for revisions in well-defined circumstances. Clearly, seasonal factors could easily be incorporated into such an approach; so, too, could sustained changes in velocity.

My view is that the use of the monetary instrument for stabilisation purposes has created considerable uncertainty about longer-term outcomes. In turn, this has made policy objectives with a longer time horizon much harder to achieve. Ultimately, a choice will have to be made between using monetary policy for stabilisation purposes or for longer-term outcomes. If it is determined that a longer-term policy objective, such as the reduction of inflation, takes priority, then monetary targeting may be the best approach.

(c) Concluding Remarks

Milbourne concludes his paper by asserting that a reduction in inflation should be the key objective of monetary policy in the 1990s. The distortions introduced to the economy by high levels of inflation are clearly undesirable. I am sure that all here would agree that, in an ideal world, inflation should be eliminated or, at least, substantially reduced. But this is not an ideal world. There can be substantial costs to reducing inflation. Also, the Australian Government must face election at most at three-year intervals. Substantial reduction in inflation over the 1990s may prove more difficult than many of us anticipate.

Footnotes

Australian National University [*]

Norton, W.E. (editor), Conference in Applied Economic Research, Reserve Bank of Australia, December 1979. [1]

Stemp, P.J. and Murphy, C.W., “Monetary Policy in Australia: The Conflict between Short-term and Medium-term Objectives”, Discussion Paper No. 237, Centre for Economics Policy Research, Australian National University, July 1990. Also forthcoming in the Australian Economic Review. [2]

Poole, W. (1970), “Optimal Choice of Monetary Policy Instruments in a Simple Stochastic Macro Model”, Quarterly Journal of Economics, 84, pp 197–216. [3]

3. General Discussion

There was some discussion of the pros and cons of monetary targeting, with the proponents tending to focus on the money base. Money-base targeting was advocated because the central bank should be able to control the quantity of its own liabilities; this would make the money base preferable to other monetary aggregates. Other participants pointed out that the central bank could use its position to set either the quantity or the price of its liabilities and there are reasons why it may wish to do one or the other.

There was also a fair amount of disagreement on whether the relationship between the money base and nominal income or inflation was stable. Another separate issue was whether the relationship would retain its present degree of stability (or instability) if pressure was put on it; would Goodhart's Law operate if it was used as a supply relationship?

The point was made that if movements in money supply are correlated with movements in the cycle of economic activity, then this implies that monetary policy has been pro-cyclical. Against this, other interpretations of such a correlation were:

• money supply is endogenous and so the cycles in the economy are feeding back into money supply;
• changes in money supply are not a good measure of changes in monetary policy. In the Australian case, changes in the overnight cash rate or the degree of inversity of the yield curve are better measures;
• in the absence of exact information as to whether a shock is financial or real, the appropriate response may be a middle course, smoothing both interest rates and base money. This could cause ‘pro-cyclical’ correlation.

An alternative way of looking at the issue would be to follow, episode by episode, the various movements in the economy and monetary policy over the decade. If this were done, the only recent period where it could be maintained that monetary policy was pro-cyclical was in 1987.

A number of participants agreed that the appropriate monetary policy response depends on the nature of the shock. Poole established a couple of decades ago that the optimal response is to concentrate on the interest rate if the shock is financial, and on the money supply if there is a real shock. While the source of the shock cannot be known with certainty at the time it occurs, the central bank is not completely ignorant and so there is a case for something other than a rigid rule.

Several participants objected to looking at monetary policy in isolation from other policies. For example, if it takes a long time to change fiscal policy in the desired direction, monetary policy could be adjusted in the interim to offset the sluggishness in fiscal policy. A variant on this was that in the Accord arrangements in 1986, nominal wages were set with a view to real wages falling. If monetary policy had been extremely tight, inflation may have been lower than wage growth and the desired adjustment in real wages would not have occurred. Other participants disagreed strongly with the view that monetary policy should adjust on an ad hoc basis in order to offset weaknesses or delays in adjusting other policies.

Several participants supported Milbourne's view that the Reserve Bank should not make monthly justifications of why it adjusted (or did not adjust) monetary policy. They felt the Reserve Bank should adopt a more medium-term perspective and only comment if a change in policy occurred.

One participant congratulated the author on not having devoted any space to the subject of central bank independence. He felt that until economists were confident that they knew how the transmission mechanism worked and were able to obtain stable relationships, they should not have the confidence to suggest that monetary policy be handled as a technical matter outside the political process.

From an international perspective, the view was put that although there is much less confidence in setting intermediate targets for monetary policy than was the case a decade ago, there is more agreement on what the ultimate aim should be. All central banks now gave primacy, if not exclusivity, to reducing inflation as the goal of monetary policy.