Speech ‘Two Perspectives on Monetary Policy’
Talk to a Conference in Honour of Don Sanders
As a long-time fan of Don Sanders, I am delighted to be participating in this tribute to him. As Treasury Secretary, I sat with Don on the Reserve Bank Board, and on the Commonwealth Bank Board, after he was appointed Managing Director of the latter bank in March 1987. As Governor, I have taken a more than passing interest in his commercial banking activities.
Our close contact over recent years has brought home to me just how good a choice Don was to head up the Commonwealth Bank. His cautious but firm hand was precisely what the Commonwealth Bank needed at that point in its history. Whenever you debate an issue with Don you are always uncomfortably aware that he has come to his position after deep deliberation, and will defend it stoutly. His intellectual rigour and integrity, and his effective management style, lay behind the exceptionally solid performance of the Commonwealth Bank under his leadership.
The two perspectives of my talk are not, as the title might possibly suggest, the rules orientated and discretionary approaches to monetary policy. Rather, they are perspectives on monetary policy from the vantage points of someone who has been a practitioner in both the Treasury and the Bank. They are personal observations; I know some of my colleagues – past and current – will not share a number of them.
The title carries an implication that there are two different perspectives. I am acutely aware of certain suggestions that there should be two different perspectives. Some critics who professed to see nothing politically sinister about my own appointment, for example, nonetheless saw something intrinsically wrong in a Secretary to the Treasury being appointed Governor of the Reserve Bank.
That seems a peculiarly Australian notion. Many central bankers around the world have Treasury backgrounds, including the current Governors of central banks in France, the Netherlands, Sweden, Norway and India. The President-elect of the Bundesbank is also from the Finance Ministry. In Japan, the governorship is rotated between the Bank of Japan and the Ministry of Finance. (Curiously, the traffic is all one-way; commercial banking, it seems, is a popular destination with many ex-central bankers.)
Unless you believe in leopards changing their spots, international practice suggests that the perspectives are not intrinsically – or irreconcilably – different.
By virtue of their ex officio positions on the Board of the Reserve Bank, and their Department's legitimate involvement in monetary policy, Treasury Secretaries (like their overseas counterparts) do acquire a familiarity with the objectives and operations of central banks. They also acquire an intimate understanding of wider economic policy and political decision making processes which, potentially, could enhance a central bank's effectiveness: despite occasional suggestions to the contrary, no central bank operates in a political vacuum. That, however, is to anticipate some observations I will come to shortly. The immediate point is simply that appointments such as my own should be judged on the basis of the all-round talents of the individuals concerned, not on the presumption that the perspectives are – or should be – so different as to be irreconcilable.
Few people span the full term of Don Sanders' involvement in monetary matters. My talk will focus mainly on the past decade. This is the period I know best, being privileged to have served as Treasury Secretary for five years, and as Governor for the past three years: in this audience I must capitalise on any comparative advantage I might have! But it does mean passing over some apparently epic contests between Governors and Secretaries. By my time, legendary spectacles of a couple of giants doing battle before the crowd had given way to relatively tame discussions among ‘equals’. Be that as it may, I count myself fortunate to have had first class working relationships with Bob Johnston, Don Sanders and John Phillips when I was in Treasury, and again with Chris Higgins and Tony Cole when I moved to the other side of the table.
Another reason for concentrating on the past decade is that monetary policy instruments – and the associated institutional arrangements – in that period were quite different from earlier periods. Until the early 1980s, monetary policy was exercised through a variety of instruments – such as interest rate ceilings, the setting of bond rates, variations in the Statutory Reserve Deposit Ratio, lending controls, monetary targets, pegged exchange rates – and the Treasurer and Treasury were very much involved in their use. ‘Deregulation’ and other changes have seen these controls abandoned to the point where short term interest rates are now virtually the only monetary policy instrument.
With these changes has come much greater autonomy for the Bank. In particular, decision making has shifted from committees in Canberra to financial markets which are influenced by the Bank's operations. The introduction of bond tenders to fully fund budget deficits in 1982, and the floating of the currency in 1983, were two especially notable changes. Another sign of this shift was that the M3 projection, the watchword for monetary policy changes between 1976 and 1984 and reserved for promulgation in the Treasurer's budget speeches, was dropped in January 1985.
Given such changes, it is only to be expected that the perspectives on monetary policy in the recent period would be very different – from both vantage points – from what they were in earlier periods. But what about differences between the Bank and the Treasury within particular periods, and especially the past decade?
Differences in Perspectives
There are obviously some differences. To begin with, the Bank is established under an Act of Parliament, with a Board of Directors, its own charter, specific procedures for consultation and the resolution of possible disputes with the Treasurer, and so on. This Act affords a degree of clarity and comfort to Governors that is not available to Secretaries.
The legal separateness of the Bank is enhanced by its distance from Canberra. Close physical proximity to the foreign exchange and domestic money markets in which the Bank operates is vital, and I see in the Bank an intimacy with the markets that would be impossible for a Canberra-based Treasury to replicate even if it tried to – which it sensibly never has.
The Bank is closer to the markets but Treasury is closer to Parliament House. It is sometimes inferred from this that Treasury is more susceptible to political pressures, and more preoccupied with day-to-day events to the detriment of longer run trends, than a (‘properly independent’) central bank would be. In my experience, such inferences are incorrect; senior officers of the Treasury, almost to a fault, have always called the shots as they have seen them, whatever the political pressures. And the ‘longer run’, or a ‘medium term framework’, has been an integral part of Treasury thinking for as long as I can remember.
Quite a lot has been written lately about the Bank's ‘independence’. In my experience, the Bank enjoys a high degree of independence and I recall that others in a position to know the facts have rejected all assertions that the Bank has been pressured into taking ‘political’ decisions. It puzzles me that some people seem to equate independence with conflict, consultation with subservience, and (at least until recently) concern for growth with wimpishness.
In the institutional setting that has existed since the mid 1980s, monetary policy is a much more ‘hands on’ experience for the Bank than it could ever be for the Treasury. With that goes a heightened awareness that monetary policy actions and utterances by the Bank are being shadowed more closely by the markets, the media and others than is the case with the Treasury. The need for extra care in this situation is obvious, the more so when it has to be accommodated with an equally compelling need for the Bank to say more about monetary policy issues in the interests of raising public understanding of them.
As important as some of these differences are, they say little about perspectives of monetary policy. Before exploring that issue in the context of the past decade, I want to look briefly at some changes in perspectives on the role of monetary policy over a longer period. These serve to remind us how policies and their intellectual underpinnings change; they suggest also that the main differences in perspectives reflect not so much institutional factors as new sets of circumstances and frameworks of thought.
Monetary Policy and Inflation
At the beginning of the 1960s, when Don was settling back into the Research Department of the Bank after some study at the LSE and a year at the Bank of England, and I had just joined the Public Service, stabilising the cycle was the economic fashion. Inflation was a concern, even in this Keynesian world, and to combat it policy relied heavily on demand management. Fiscal policy was seen by most economists as the preferred instrument for that purpose. Monetary policy was seen as an adjunct to fiscal policy, but doubts remained about its potency; in ‘liquidity traps’, for example, easing monetary policy could be like ‘pushing on a piece of string’. There was no presumption that monetary policy might have a comparative advantage in controlling inflation. Both monetary policy and fiscal policy influenced demand and, then, inflation. Some confidence existed that the cycle could be controlled and that if inflation reared its head, it could be handled by a short sharp shock of the 1960/61 variety.
The view that fiscal policy played the dominant role in affecting economic activity and controlling inflation was reflected by Heinz Arndt at the time. Writing in 1960 on the control of inflation through fiscal policy, Heinz said:
If aggregate spending is excessive, there will be inflation … one way in which governments can control the level of aggregate spending is through budget policy …
But he was careful to note the importance also of other policies:
The efficacy of fiscal policy depends in considerable measure on the simultaneous application of disinflationary monetary and wages policy. Unless monetary restraints are effective, much of the impact of fiscal policy on consumption may be dissipated through increased resort to consumer credit.
The Keynesian framework of policy was a framework for both unemployment and inflation. Changes to the intellectual underpinning of such frameworks usually come not from an abstract rethinking of the theory, but from the arrival of some inconvenient facts which cannot be explained in terms of the old theory. The inconvenient facts which arrived in the early 1970s were the conjunction of high inflation and high unemployment. Inflation was not simply a matter of the economy growing too fast: prices could rise even if there was slack in the economy. This problem was widespread, with inflation rising sharply in major OECD countries from the beginning of the 1970s (see Graph 1).
The emergence of significantly higher inflation and the traumas of the first OPEC oil price shock undermined the demand orientated Keynesian framework and cleared the way for a different strand of theory to emerge – one which had its origins in the classical idea of ‘money neutrality’. It was to revive the role of monetary policy and elevate it from just another instrument for influencing aggregate demand to a more central and specialised role in controlling inflation.
In leading the revival of the old Quantity Theory, Friedman emphasised that money and prices were closely linked, and that increases in the money supply would lead mainly to higher prices, with no long term increase in real activity. This provided a much sharper focus on what caused prices to rise. It had the policy implication that governments can and should control the money supply in order to avoid inflation; in its extreme form, it implied that monetary policy should be put on auto pilot. Australia did not go that far but a form of monetary targeting was introduced in 1976.
A close link with fiscal policy remained, however, because the budget position fed directly into money formation, which was at the heart of the monetary analysis of the time. This and the sale of government bonds to the non-bank public were seen as central to monetary policy. In Statement No. 2 in the 1981/82 Budget papers, Treasury noted:
The Budget is clearly a central policy instrument, both in its own right and in its contribution to the achievement of the objectives of monetary policy ... unless the growth of monetary aggregates can be reduced consistently over a run of years, there can be no real prospect of winding back the rate of inflation.
And a little further on:
Appropriate fiscal and monetary policies are twin imperatives for reducing inflation.
Treasury's less than wholehearted endorsement of monetarism appears to have been shared by the Bank. Responding to the monetarist debate in the Campbell Committee context in mid 1982, Don Sanders said:
Some of you may find anything short of wholehearted embrace of a monetarist rule disappointing; what I see as real complexities should not be read as antipathy to monetarism and all its works. The philosophy is fundamental to efforts to come to grips with inflation. But it can be unwise to rely too heavily on a monetarist policy.
I am pleased that the (Campbell) Committee did not get more closely wedded to these (monetarist) philosophies. I do not think that a monetary rule can tame the authorities …. We practitioners have always felt that fiscal outcomes, as well as money, matter.
By the mid 1980s, even the cautious endorsement which most central banks around the world had given to monetarism – what Charles Goodhart has called ‘pragmatic monetarism’ – was being inconvenienced by the facts. To be useful for policy, a stable relationship between money and prices was a critical pre-requisite; this did not prove to be the case in many countries, including Australia, and our ‘conditional projection’ was abandoned in early 1985. As Treasury noted in Statement No. 2 later that year:
The analytical integrity of simple monetary targeting rests on the existence of a stable relationship between the chosen aggregate and the ultimate policy objectives. That is no longer the case at present in Australia.
The breakdown is illustrated in Graph 2.
Monetary Policy in the 1980s
Others in this audience are better equipped to discuss Treasury and Bank perspectives on monetary policy over the past thirty years. I suspect, however, that while the perspectives changed over time, there was a good deal of common ground between the two institutions at any particular point in time. That is, their views were shaped primarily by changing circumstances and intellectual frameworks, not by their particular vantage points; as the world changed and old views were found wanting, the ‘mainstream’ view changed in both institutions. Certainly, in the period since the mid 1980s, a large degree of unanimity has existed between Canberra and Sydney on the role and objectives of monetary policy.
If there was a difference in perspective in this latter period, it was a preference on the part of the Bank for relatively more of the burden of adjustment to be borne by fiscal policy and, from the Treasury side, a preference for monetary policy to shoulder relatively more of the load. Some of this flavour is contained in Don Sanders' comments to the Campbell Committee which I quoted earlier, and it percolated through some of the advice going forward to the Government from the Bank and Treasury during this period. It is perhaps inherent in every institution to prefer that the more unpalatable medicine be administered by another.
Of greater policy significance around this time was the emergence of a fairly common view that fiscal policy should be seen more as a vehicle for medium term structural change of the public sector than as an instrument for stabilisation. By 1983/84, the PSBR had blown out to 7 per cent of GDP, and that in itself had limited the room to use fiscal policy for counter-cyclical purposes.
The primary objective of the new emphasis was to reduce the structural deficit and the size of the public sector over time, mainly through restraint on government expenditures. It was in keeping with the prevailing international orthodoxy. But one consequence was to push more of the burden of inflation control and counter-cyclical stabilisation onto monetary policy. We all recognise the limitations of macro fine-tuning but policy makers cannot tune all their instruments to just the long run. With the benefit of hindsight, it might be argued that the reluctance to use fiscal policy more vigorously for counter-cyclical purposes in both the upswing and downswing phases of the recent cycle had the effect of over-burdening monetary policy.
In contrast to the ‘stagflation’ of the 1970s, the 1980s saw long periods of lower inflation and stronger growth. Australia shared in this improved but variable performance. Australia, for example, recorded more sustained growth in employment in the 1980s than any other OECD country, but generally higher inflation. I believe there was also much greater attention to ‘social reforms’ in Australia during this period than was embodied in, for example, Reaganomics or Thatcherism. (One's view of the relative merits of these different thrusts depends ultimately, of course, on one's view of the world more generally.)
The 1980s were challenging years for monetary policy. One episode which illustrates this was the easing of monetary policy in 1987, which has been credited by some people with causing the boom at the end of the decade. In hindsight, it can be argued that the easing went too far, and so made some contribution to the boom. But the story is more complicated than that. First, some easing was required: interest rates were very high in 1985/86 (see Graph 3) and by early 1987 those levels could no longer be justified, given that activity had plateaued, inflation had peaked and the earlier weakness of the exchange rate had been reversed.
Secondly, international and domestic conditions outside the influence of monetary policy were such that activity was bound to accelerate. The strong world economy saw Australia's terms of trade rise by about 25 per cent from the low point in 1986 and such rises have always been expansionary in the past (and inflationary, although on this occasion – see Graph 4 – inflation continued to edge down, in contrast to what was occurring in most countries). Domestically, a higher profit share (facilitated by wage restraint), and a more competitive exchange rate, were associated with a very bullish trend in business confidence and investment. In short, monetary policy contributed to but did not cause the boom.
There are, of course, other episodes, the essence of which will be familiar to members of this audience. Critics have argued, for example, that policy should not have been tightened as much as it was in 1988/89; that policy should have been tightened earlier, before the upswing gathered too much momentum; and, after the inevitable recession, that policy should have been eased sooner and faster. I mention these criticisms not to debate their merits: again with the benefit of hindsight, some mistakes were obviously made, although at the time I recall few people arguing that the tightening in 1988/89 had gone too far, and even fewer suggesting that the monetary easings which commenced in January 1990 should have been started earlier or pushed harder. What these and other episodes serve to indicate is that business cycles have a momentum of their own, and that other factors besides monetary policy bear upon booms, recessions, recoveries and inflation.
Inflation and Growth
I have talked mainly about monetary policy in the context of inflation, with growth in activity and employment attracting less attention. This accords broadly with the priorities which, I believe, central banks and Treasuries traditionally assign to these objectives. Indeed, in my observation, those priorities are ingrained in the cultures of both the Reserve Bank and the Treasury (and of their counterparts in most other countries).
There is an important if subtle point here. It is that more attention should be paid to growth, and to sustaining that with low inflation, not that less attention should be paid to inflation. In a world of limited natural constituencies for low inflation, central banks and Treasuries will be obliged to always give a high priority to inflation. But a broader focus (together with better forecasts!) would help to achieve a better balance of the risks inherent in policy adjustments.
I trust those comments can be made without arousing suspicions of being or going soft on inflation. It should be obvious from the performance of monetary policy over recent years that there is no basis for such suspicions. Australia now has one of the lowest rates of inflation in the world. Inflation has come down faster than everyone expected, but it is not just the recession, or a fluke, that has caused it to decline. Policy has been important, and the costs substantial. Price expectations, which are now seen as occupying a central role in the inflationary process, have been cracked; given this, together with continued policy vigilance, there is no reason why the current underlying inflation rate of 2 to 3 per cent cannot be sustained.
In the 1980s, it was common for economists to assign specialised roles to particular policy instruments, and to attach priority rankings to inflation, growth and other objectives. Life would be a lot easier for policy advisers if it was as simple as that! Reality, in my view, requires a broader view than that, including greater acceptance of the following two propositions:
- monetary policy on its own will not deliver sustained low inflation in Australia; and
- growth and jobs, as well as inflation, are important objectives.
Practitioners generally take some comfort from having a full case of policy instruments to draw upon, whether the problem is inflation, growth, jobs or whatever. Experience demonstrates that the accessibility and effectiveness of particular instruments will vary with the circumstances of individual situations, including the phase of the business and budget cycles.
Monetary policy today is more focused on inflation control – that is where its comparative advantage is seen to lie – but that does not mean we have a single objective. Monetary policy cannot ignore activity and the cycle, especially if (as seems likely) fiscal policy must maintain a predominantly medium term horizon for some time to come. We now have a more precise view of the role of fiscal and monetary policy in relation to inflation than was current in the 1960s. Few people today would see fiscal policy as the critical policy instrument in controlling inflation. That said, the thrust of the earlier analysis still seems to be correct: that excess demand can be an important cause of inflation and all macro policy instruments should be used to combat that problem when it occurs.
But controlling inflation at least cost also requires judicious recourse to other available policy instruments. Tariff reductions, better transport, and practically everything that comes under the tag of micro reform has to be included here. So does the Accord, which I believe has been a substantial contributor to the low rate of inflation we now see in Australia: the Accord processes are not perfect but that is the nature of compromise and human affairs generally.
It is a pragmatic approach but I see nothing wrong in following what is most likely to work in the real world. In my view, careful co-ordination of all policy instruments is more likely to deliver sustained lower inflation at acceptable cost in Australia than is, for example, relying on monetary policy to sustain a low inflation target.
As inflation has come down and unemployment has risen, more emphasis has come, understandably, to be placed on growth and jobs. How should the authorities respond? One response is to say that we have never had a single-minded fixation with price stability. That, clearly, has been the situation in recent years. In each of the thirteen announced reductions in cash rates since January 1990, the authorities (meaning the Government and the Bank) have acknowledged the importance of trends in both inflation and activity in those decisions.
It is not as if there is any great mystery about how to bring about low or even zero inflation; the more relevant question is how to achieve a sustainable balance between low inflation and high productivity growth.
How can we put policy content into this? Monetary policy cannot, by itself, awaken the animal spirits which drive business investment. Nor can it, with any precision, keep them in check when they are running fast. What policy can do is to lean against these forces during the cycle, and over the medium term help to set the framework within which businesses will make decisions conducive to growth and employment. Price stability is one important element of this framework, as it facilitates sensible and rational investment decisions.
At the same time, while it might be inconvenient to some, the fact has to be recognised that low inflation is not an end in itself and will not, in itself, be sufficient to ensure growth and bring down unemployment. There are many other elements in the framework, including those which come under the general rubric of micro reform. It is worth remembering that the good pace of growth between 1983 and 1990 followed action to bring the budget deficit and real wages back into kilter; those changes contributed to the business confidence which drove the growth (in its later stages, too fast). It is an old dilemma, but it needs to be resolved afresh in its current setting; we do not want the 1990s to become a decade of low inflation and low growth.