Can Monetary Policy be Made to Work?

INTRODUCTION

THIS booklet has been divided into four sections:-

• A Critique of Monetary Policy during the 1980s;
• How Monetary Policy Should be Operated;
• The Alternative to Central Banks;  and
• New Zealand's Monetary Policy.

As will be evident from what follows, I have adopted widely differing approaches to the broad policy issues I was asked to address, viz:

• Is it theoretically feasible to have a monetary policy, run by a government monopoly, so that a stated economic objective or objectives can be met?
• If the answer to the above is YES, is it feasible in practice for that to happen and, if so, what ground rules are needed?
• Would it be better to leave it to the market to determine the supply and demand for money?

The decision to publish this booklet reflected the widely held feeling that the 1980s had witnessed a major failure of monetary policy.  There is no doubt that the experience of the 1980s has left a great many people in Australia, and I suspect overseas, sceptical of the ability of the public sector to produce "sound money", and, at the least, they are looking for changes to prevent a repetition of recent experiences.  It is hoped that the papers herein will contribute to a substantial improvement in the 1990s.

Many are looking to a solution that would heavily constrain the role of the public sector.  Some would like to eliminate it altogether.  But we should not overlook that there is also a body of opinion that wants to re-regulate.  And that body of opinion is having quite an influence in one area, that of prudential supervision and new entry controls.  This aspect, which is not taken up in detail here, may very well make it more difficult to operate monetary policy, let alone any "free market" in money.

---

MONETARY POLICY

MONETARY POLICY is currently one of Australia's more controversial topics.  The first item in the Coalition's 20-point Fightback! Plan is "A commitment to price stability."  Finally the subject of inflation is getting some debate, and there will be more in the period ahead.  The issue of monetary policy raises several questions:

• What were the mistakes of the 1980s?
• What are the lessons for the 1990s?
• What are the radical alternatives?
• The 1980s revisited.
• What should be done in practice to achieve price stability in Australia?

MISTAKES OF THE 1980S

The 1980s culminated in a burst of asset price inflation and rapidly growing international debt.  More generally, the Australian economy has struggled for the past 20 years.  Current account deficits have been endemic;  international debt has grown far too big;  and there have been recurrent cycles of boom and bust.  Above all, inflation has been deeply entrenched.

It is inappropriate to place all the blame for this sorry state of affairs on our method of organising monetary policy, but monetary policy must take its share of the blame.  One can, I think, make the point that outcomes were better in the 1980s than in the 1970s.  When Peter Jonson left the Reserve Bank in 1988, he produced some "Reflections on Central Banking."  As to monetary policy he concluded that, in the 1980s:  "Monetary policy has been more successful than in the earlier eras examined";  and "Monetary policy would (with the benefit of hindsight) have been even more successful if it had been both steadier in its application and somewhat firmer overall."

This second judgment in particular has been frequently reiterated -- e.g. in the recent critique by Peter Hartley and Michael Porter.

I think there can be little doubt that monetary policy was both too inflationary and too volatile in the 1980s.  In his 1988 critique Jonson argued that "political" bias was not part of the problem.  Rather, he argued that there were persistent errors of judgment about where the economy was headed.  "The most persistent tendency has been to underestimate demand in the Australian economy."  A prophetic statement as far as 1988 and 1989 were concerned;  after that I suppose you can say that the opposite tendency came into play -- perhaps partly as a reaction.  The point, of course, is that underestimating the strength of demand in the economy produces a natural tendency to excessive looseness in monetary policy.

On the question of volatility in monetary policy, Jonson's main explanation was that policy had been excessively influenced by trends in international interest rates and the associated movements in the exchange rate.  His prescription was that "Australian monetary policy should be more self-confident in its application.  As a nation we must be prepared to steer an appropriate course on domestic monetary policy more independently of international forces."

Perhaps the most general mistake of the 1980s was to underestimate the costs of inflation.  There has been much recent debate on this subject and it is now widely agreed that inflation forces up rates of interest, saps competitiveness, reduces incentives to save and invest and ultimately puts at risk a country's financial and economic stability.  Most fundamentally, inflation is theft.  It erodes, in many subtle ways, standards of private and public morality.  Failing to recognise the costs of inflation must be counted as one of the major intellectual failings of the post-war era throughout the world.

As I have indicated, however, one can take criticism of Australian monetary policy in the 1980s too far.  Three points in defence of monetary policy during the 1980s are worth making.

• Although inflation was clearly excessive in the 1980s, it was a good deal less than in the 1970s -- and the trend was clearly down.
• Much of the inflation in the second half of the 1980s was due to the effects of the big currency devaluations of 1985 and 1986.  These devaluations were inevitable given the general problems of the Australian economy.  Firm monetary policy prevented flow-on effects and, in fact, laid the basis for the low inflation which we have now.
• A major criticism of monetary policy at the end of the decade, of course, is that it was tightened excessively.  Obviously, however, if there have been two decades of mistakes in the other direction, a mistake in the direction of being too tough might be seen as part of an inevitable correction.  It could also be observed that the consequences of this "mistake" (if it was a mistake) have included some very desirable changes in other policies and in the attitudes of many Australians.

Notwithstanding this defence, there are many who argue that there are better systems for operating monetary policy than our present system.  I am one of those.

A BETTER SYSTEM

As I thought further about these issues, I came to the view that it was essential to have a central bank much more independent of governmental (as opposed to "political") forces, and with a much simpler charter, than the one we have now.  Jonson's next foray on the subject was in 1990.  The argument included the following:

"Given the wide-ranging influence of monetary policy, its ease of use, the wide-ranging charter of the Reserve Bank and the tradition of close co-operation with government, it is not surprising that in Australia (as in many other countries) monetary policy is constantly seeking an appropriate balance between objectives which often seem to conflict.

"Trying to achieve a lot of things is often a recipe for achieving little.  In particular, it is hard for a central bank to have credibility in fighting inflation if it is also supposed to concern itself with restraining unemployment.

"It cannot be denied that, in Australia, the attempt to achieve too many objectives has meant that the primary aim of monetary policy, the elimination of inflation, has not been achieved.  So it seems to me that policy overall would be more effective if the Reserve Bank were to be given as its main task the achievement and maintenance of price stability.  This could be achieved by simply deleting (b) and (c) from the list of objectives in the existing Act.  Or it could be achieved by a more wide-ranging revision of the Act, including changes to the composition of the Board.

"Price stability could not be achieved overnight, but a reasonable period, say three to five years, could be specified.  For technical reasons, the Bank should also be given discretion as to how to achieve this objective, and also to respond to unexpected events.  The new charter should of course include responsibility for maintaining the stability of the financial system.  It could also list, as subsidiary objectives of monetary policy, the maintenance of full employment and the achievement of economic prosperity.  Yet financial stability, full employment and general prosperity would all be enhanced in the medium term by price stability.  So there is no basic conflict between the various objectives, or indeed between the existing charter and one that makes the achievement of price stability the primary objective of monetary policy.  The subsidiary clauses would remind the Bank, if such a reminder was necessary, that care needs to be exercised in pursuit of the basic objective.

"Such a charter would concentrate the efforts of the Bank in doing what it is best suited to do.  It would also concentrate the minds of those responsible for other policies in a more effective way.

"If price stability were not achieved in the time specified, some sanction should be applied.  Fairness dictates that there be room for discussion before the sanction is applied.  But if the objective were not achieved and if explanations were not acceptable, then the Board of the Bank should presumably be replaced."

All this was regarded as controversial, even a bit eccentric, in 1990.  Now it seems to have become more fashionable.  Certainly I am delighted it has become a major plank in the Coalition's reform package.  I observe with amused detachment the way in which sections of the press are now arguing that the Reserve Bank is indeed already quite independent and should be left alone by any incoming Coalition government.  One could also comment that the Bank itself has moved quite a bit in this direction.  As the starkest and most recent example, Governor Fraser said on the evening of 28 November 1991:  "In a general sense, a strong case exists for a high degree of central bank independence."

In looking ahead, it seems to me that the main issue is whether an independent central bank, with price stability as its sole or major macro-economic objective, is the best system that can be devised.  Certainly this is what I believe, but it is also necessary to examine alternative systems that have been proposed.

ALTERNATIVE RADICAL SYSTEMS

Broadly speaking there are three more or less radical alternatives to the notion of an independent anti-inflationary central bank.  They are:

• The gold standard or some more complicated commodity standard.
• The money supply standard.
• Competitive moneys.

COMMODITY STANDARDS

One proposal is to return to some sort of commodity standard modelled on the old gold standard.  Gold itself can be ruled out (except in the minds of those who own gold mines) since its production is both highly elastic and subject to the fits and starts resulting from major discoveries.  A bundle of valuable commodities could, in principle, underpin the monetary systems of major countries.  In effect this proposal is destined to replace legislative self-restraint in the creation of paper money.  As such, it is at one level unnecessary, and at another would be undermined by clever politicians if they were not bound by law to monetary rules that produce price stability or a central bank committed to the same aim.

It is interesting to note that no private agency has produced a workable version of this system.

THE MONEY SUPPLY STANDARD

This is the notion associated most clearly with the name of Milton Friedman.  He advocated the fixing of money growth to a predetermined rate (e.g. equal to the long term growth of output per head).  Doing this and sticking with it, so the argument runs, will mean that inflation will eventually be eliminated.

This, in principle, is an attractive idea -- especially when one of its corollaries is (or seems to be) that one could replace an expensive central bank with a cheap computer programmed by someone clever like Milton Friedman or Michael Porter.

The flaw in this system is that neither the money supply nor the money base -- nor any other financial aggregate -- has been shown to have had a sufficiently stable relationship with economic output or the rate of inflation.

To be sure, such stability seemed to be established after a lot of painstaking econometric and historical analysis.  It is, in fact, an irony that once the money standard was introduced -- as it was in many Western countries -- relationships on which it was based broke down.  This may, of course, be more than accidental.  It seems to me that any overly simplistic predetermined rule will be picked off by clever people with major incentives to do so.

Certainly I have come to the view that a "money standard" is not the best way to eliminate inflation in the modern world.  That is not to say that eliminating inflation will not require monetary restraint.  Of course it will.  The point, however, is that there is no simple formula for growth in a monetary aggregate which can or should replace the judgment of central banking professionals in achieving price stability at minimum cost.

COMPETITIVE MONEYS

The theory here is attractive.  If the law were changed so that anyone could provide a form of "money", the provider who limited his output in a way that maintained value over time would gain market share.  Ultimately his sound money would dominate the market.

To my mind there are many practical objections to this approach.  Most basically, a stable currency is one of the ultimate public goods.  Provision of currency by a central agency is likely to be the most efficient system, by a large margin.  In any case, current law allows private agencies to issue forms of "money".  There are in fact examples, but none have so far passed the market test.

A further point is that there are already many "competitive moneys".  In the current world of free capital flows, it is legal for Australian citizens, for example, to hold deposits in any of the major currencies.  In the long run, currencies that provide a stable store of value should gain market share in the way envisaged by proponents of competitive moneys.

For the vast majority of Australians, the Australian dollar is still the currency of choice.

A THOUGHT EXPERIMENT:  THE 1980S REVISITED

What would the 1980s in Australia have been like with an independent anti-inflationary central bank?  Such a bank would have raised interest rates more quickly and strongly as the economy rebounded from recession in 1983.  The rising exchange rate which would have accompanied the tightening of monetary policy would have forced the government to have faced, much sooner than it did, the weakness in Australia's external position.  Fiscal policy would have been tightened much earlier and, conceivably, the consumption tax would not have been abandoned in 1985.  In any case, lower inflation would have encouraged higher saving by the private sector.  Tighter monetary policy throughout the first part of the 1980s would also have restrained growth and employment and this pressure would almost certainly have encouraged earlier and more extensive micro-economic reform.

In this alternative scenario, there would have been much less incentive for private banks (and their customers) to indulge in an orgy of lending and borrowing.  A tougher and more independent Reserve Bank would, in any case, have been more inclined to contain any tendency to injudicious lending by the banks.  Banking supervision would have been on the playing field, not "in the pavilion polishing the whistle."  Australia would have entered the 1990s with internal and external debt much lower.  Inflation would have been eliminated and our general economic prosperity would now be much greater.

LOOKING AHEAD

It would be very simple to set the Reserve Bank on a different course.  An incoming Government could tell the Governor that it wanted the Bank to do everything in its power to achieve price stability over the next five years.  It would also tell the Governor that it would change the Reserve Bank Act as soon as possible but that, in the meantime, it considered price stability the major objective for the Bank.

Assuming that this objective had been freely debated in the lead up to the election, the Governor would have no room for doubt that this was the will of the people.  There is plenty of room in the existing charter of the Reserve Bank for the Bank to act more or less as it sees fit in achieving its aims.

I would wish the Bank to start the ball rolling with a major campaign on the costs of inflation and on the need to eliminate it.  Such a campaign has been waged, with conspicuous success, by both the Bank of Canada and the Reserve Bank of New Zealand.  As part of this campaign, the Bank would announce that its main objective now was to achieve price stability over the next five years.

The Governor would then ask his economists for their forecasts for the course of inflation over the next five years.  He would, of course, check these forecasts with outside economists whose judgment he valued.  If the forecasts did not show inflation declining to zero, he might make some suggestions to the Government about changes to policy other than monetary policy.  Certainly if he thought other policy changes were needed he would have a duty to so inform the Government.

None of this would be independent of monetary policy.  If prevailing policy with allowance for other desirable and likely policy changes were not forecast to produce price stability, then monetary policy would need to be tightened.  How much and over what period would be carefully judged at the time.  How much tightening of monetary policy was necessary would also depend on the credibility of the Bank and on the Government's general economic policy.

This process would be continuously repeated.  All of the factors impacting on the rate of inflation would need to be evaluated constantly -- the Bank's famous (or infamous!) check-list would be used heavily.

Forward-looking indicators of inflation would be closely scrutinised -- including measures of inflationary expectations such as the various surveys, long term bond yields and last (but not least!) the economists' forecasts.

Whilst the Governor and the Bank were getting on with this work, the Government would presumably seek to make various legislative changes.  Nowadays there are plenty of examples to choose from and no doubt the Government would choose elements of various models.

The Bundesbank is often held up as the model of an independent anti-inflationary central bank.  But amongst the smaller countries, the new and quite explicit legislation recently adopted by both Canada and New Zealand has many attractive features.

All in all, I believe that the right combination of actions (by the Bank with its monetary policy, by the Government with its general economic program) and words (in "setting the agenda") would enable price stability to be achieved within the five-year period.

Nor do I believe that the short-term cost of this program would be high.  Furthermore, I believe that the long-term benefits would be immense.

---

MONETARY POLICY IN THE 1980s

THIS paper explores two broad propositions which shed some light on how monetary policy influenced economic growth and the pattern of economic activity through the 1980s:

• Officials greatly underestimated the extent to which changes in economic activity are driven by external factors.  Partly as a result, policy changes (especially interest rates) tended to follow economic activity, raising the possibility that policy changes were pro-cyclical through the 1980s;  and
• Financial deregulation had a major impact on the way in which monetary policy influences economic activity, external accounts and inflation.  Monetary policy in the late 1980s was targeted (initially) at reducing excess demand growth (relative to output growth) and the current account, objectives inappropriate for the instrument following financial deregulation.

Recognising changes in how monetary policy works and its limitations in addressing particular policy objectives is essential in answering the question posed in the title of this collection, Can Monetary Policy be Made to Work?

ECONOMIC ACTIVITY, REAL SHOCKS AND MONETARY POLICY

Australia has always faced wide swings in economic activity.  These swings can be shown to follow or coincide closely with external shocks from commodity prices and reduced farm output associated with drought. ⁽¹⁾

If we focus on the 1980s, the impact of the major 1982-83 drought (reflected in changes in Gross Farm Product) coupled with falling commodity prices can be observed on the rate of demand growth of the period (Chart 1).  The 1986 downturn, and the rapid expansion of domestic demand in 1988-89, can be linked to large changes in the terms of trade.  It is also interesting to note that the current recession did not become apparent in the official statistics until after the terms of trade started to decline in early 1990.

Chart 1:  Demand, Farm Product and Terms of Trade

A clearer view of the relationship between demand growth and the terms of trade is shown in Chart 2.  It can be clearly seen that changes in the terms of trade lead changes in domestic demand.

Chart 2:  Terms of Trade and GNE

The major policy instrument which the Government used to influence economic activity during the 1980s was interest rates.  Certainly, high interest rates have been a major factor in exacerbating the current recession.  However, changes in short-term interest rates tend to lag behind changes in the growth of domestic demand (Chart 3).  Through the 1980s the Government was consistently increasing interest rates until well after the peak of economic activity had passed, and reducing rates after growth was rekindled.  This is consistent with the idea of economic cycles being essentially driven by external shocks, and that of economic policy always trying to dampen the swings.  The current cycle is different in that the Government is keeping policy considerably tighter, a turnaround in commodity prices is yet to be seen, and a potentially serious drought is already having a major impact on rural output.

Chart 3:  Demand Growth and Interest Rates

Credit growth has also dropped to unprecedented levels in response to the current recession.  Again, interest rates can be seen to be lagging behind changes in an indicator of economic activity.  Chart 4 also illustrates how interest rates were reduced in 1987, as the turn-around in the Government's fiscal position was thought to have provided an opportunity to reduce interest rates.  Then, after the 1987 share market crash, interest rates were further eased to prevent recession.  It is now history that the worldwide easing of interest rates led to strong growth and higher commodity prices.

Chart 4:  Interest Rates and Credit Growth

MONETARY POLICY AND FINANCIAL DEREGULATION

The second proposition essentially means that the Government (and most economists) did not recognise how financial deregulation had altered the transmission channels of monetary policy.  The combination of free international capital flows, and monetary policy implemented by maintaining a cash rate target, had a major impact on inflation and the current account during the late 1980s.

Monetary policy has, since June 1982, been implemented in Australia largely through open market operations aimed at achieving a target overnight cash rate.  Before that time the Reserve Bank had used a system of "quantitative bank lending guidance." ⁽²⁾  Evidently, the Reserve Bank can maintain a given level of cash rates within a relatively wide range for extended periods.

A key aspect of the financial deregulation seen during the 1980s was the floating of the Australian dollar, and the removal of most foreign exchange controls, from December 1983.  Responding to these developments, financial markets expanded in Australia and became increasingly integrated into international capital markets.  It is likely that international investors were influenced in their portfolio allocations by the short-term rates of interest on offer in Australia, relative to those available elsewhere.

If this is true, there should be a positive relationship between net capital inflow (equivalently, the current account deficit) and short-term interest rates.  This is exactly what appears to have occurred through the late 1980s (Chart 5).  The current account deficit expanded as interest rates were raised from mid 1988, stabilised when rates were at their highest from late 1989 and then declined as rates were eased from January 1990.  The close relationship between that war, international investors pushed further funds into Australia because of Australia's role as a net energy exporter and its geographic isolation from the area of conflict.

An implication of this Mundell-Fleming style of relationship between capital flows and interest rates is that higher interest rates are effective in reducing inflation, but decrease competitiveness.  This is because the exchange rate provides the key mechanism to keep the current and capital accounts in balance.  A further implication is that higher interest rates act to reduce output relative to demand. ⁽³⁾

Chart 5:  Trend Current Account Deficit (Net Capital Inflow) and Interest Rates

The relationship between the import share of domestic demand and the relative price of imports to domestically produced goods illustrates the importance of competitiveness on trade flows (Chart 6).

Chart 6:  Imports and Relative Prices

POLICY OBJECTIVES AND MONETARY POLICY

From mid 1988 the Government increased interest rates.  The rationale for higher interest rates was concern about the current account;  an increase in the growth of demand (Gross National Expenditure) relative to growth in output (Gross Domestic Product);  and inflation.

The following quotes illustrate the balance of objectives set out by the Government during this period:

"The adverse consequences of the exceptional demand growth for inflationary pressures and the external accounts prompted further policy adjustments which were implemented progressively as the strength of demand became clearer.  Monetary policy had a particularly important role to play ..." ⁽⁴⁾

"Monetary policy in the last couple of years had a role as a balancing instrument, and while ever we had GNE outstripping GDP by a factor of two to one obviously it did have a role." ⁽⁵⁾

Concern about high demand growth was reiterated in the 1990/91 Budget Statements.  The growth in private final demand in 1988/89 was described as "unsustainably strong." ⁽⁶⁾

The perceived rapid growth of demand relative to output (eight per cent versus four per cent, measured in 1984/85 prices) during 1988-89 -- shortly after a period of lower interest rates -- appears to be a major factor in the caution officials have demonstrated in easing monetary policy through the current cycle (Chart 7).

Chart 7:  Excess Demand in 1988/89?

However, the supposed excess demand growth during 1988-89 is largely explained by stronger terms of trade, by the resultant investment boom and by an inappropriate policy response.  After we make adjustments for changes in the terms of trade, GDP expanded by nearly 6.5 per cent, only a fifth less than the growth of domestic demand (Chart 8).  In current prices, demand grew only 10 per cent faster than output (export prices had increased by 6.5 per cent and import prices had fallen by 6.4 per cent).  Also, the relative levels of demand and output, measured in current prices, were not in any sense unusual by historical standards. ⁽⁷⁾

Clearly, there was a mismatch between policy objectives and choice of instrument.  Given the clear positive relationship between interest rates and the current account deficit identified above, the gap between demand and output would have been widened rather than narrowed by raising real interest rates.  This should not be surprising, because the difference between GNE and GDP is the goods and services trade deficit;  closely related to the current account deficit, domestic savings less investment and net capital inflow.  In a world of internationally mobile capital, international capital flows have dominated the relationships between these key economic variables.  As noted above, governments generally conduct monetary policy by setting short-term interest rates, and it appears that international investors allocate their funds in response to relative rates of return and other relevant factors.

Chart 8:  Excess of GNE over GDP

To reiterate:  if high interest rates act to increase the current account deficit, then they also must reduce output more than they reduce demand.  In other words, an emphasis on tight monetary policy is inconsistent with the policy objective of reducing domestic demand growth relative to output growth.  Even if excess demand growth was a problem (were officials looking at the wrong table in the national accounts?), the tightening of monetary policy was not the appropriate policy response.

Inflation only belatedly became the major focus of monetary policy.  That high interest rates were counter-productive in dealing with a perceived current account and foreign debt problem did not reduce monetary policy's effectiveness in reducing inflation (Chart 9).  Initially, higher interest rates had the effect of increasing inflation, as higher interest rates were built into cost structures and the mortgage interest charge components of the CPI.

As the recession deepened -- under the joint impact of lower terms of trade and very high real interest rates -- the demand for labour, capital and other assets fell.  It is notable that the impact was very much through the supply side of the economy rather than through consumption.  The earlier chart depicting credit growth and the following charts of private investment and private consumption illustrate this difference.

IMPLICATIONS

The monetary policy experience of the late 1980s reinforces the proposition that it is usually impossible to address more than one objective with a single policy instrument.  We now have lower inflation, but at the cost of lower growth and a larger foreign debt.  It is doubtful that lower inflation will ever give us a sufficient growth dividend to compensate for the lost output, at least at any realistic discount rate.  The lost output, however, is a sunk cost and the issue is how to conduct monetary policy from now, in order to maximise community welfare in the future.

Chart 9:  Inflation and Interest Rates

Chart 10:  Private Capital Expenditure

Chart 11:  Private Consumption Expenditure

Inflation is the obvious target of monetary policy, but the community has broader objectives.  Low inflation might be better seen as a prerequisite for optimum economic growth than as the prime objective of policy.  From this view, would it be rational to have a central bank pursuing low inflation without regard for other, sometimes competing, policy objectives?  There is also the related issue of policy co-ordination.  Recent initiatives in the micro-economic arena (airfare and tariff reductions, rural marketing schemes) have provided considerable downward pressure on prices, and provide substitutes and complements to monetary policy for the control of inflation.  Greater competition can provide both one-off and continuing gains against inflation.  More importantly, these microeconomic reforms generally provide a growth dividend as they act to reduce inflation.  A major challenge for the 1990s must be the co-ordination of micro-economic and macro-economic policy.  Both policies can reduce the demand for factors of production in the short term.

External balance should now be abolished as a policy objective, except where government policy has distorted the external accounts.  If there is a problem, the best solution is to remove the responsible policy-induced distortions.  Monetary policy, in any case, has been shown to be an inappropriate instrument for fixing external account problems.

A major issue is how monetary policy should react to the external shocks which regularly cause wide fluctuations in economic activity.  Current suggestions that low inflation and policy co-ordination will somehow insulate the Australian economy from boom/bust cycles ⁽⁸⁾ are illogical if most of these fluctuations result from external shocks (commodity prices and droughts).  If swings in monetary policy have been pro-cyclical, there may be a strong case for a medium-term approach to policy settings, and thereby for allowing growth to fluctuate under the influence of external shocks.  Long-term average growth rates may well be maximised by allowing the economy to react without monetary intervention aimed at smoothing activity.  One difficulty is defining what a medium-term or non-interventionist monetary policy might look like.

How monetary policy should be operated is the subject of other chapters in this collection.  Yet I cannot resist a few concluding comments.  In particular, what decision rules might characterise a medium-term approach to monetary policy?

One could base these rules on the slope of the yield curve, but recent experience illustrates that bond yields are influenced by a range of factors other than inflationary expectations.  In particular, the expected supply of securities and developments in the markets for alternative investments (which substitute in portfolios for bonds) can have a strong influence on bond yields.

Alternatively, one could target a narrow monetary aggregate such as base money or currency in circulation.  But what magnitude of change in interest rates might be necessary to alter the growth of these aggregates, and how much would resulting changes in the external accounts be tolerated?  While monetary policy is rightly directed to the inflation objective, its impact on other economic variables cannot be ignored.  Also, there are potential problems with fluctuations in the velocity of these aggregates with regulatory and technological change and the delays in the supply and unreliability of nominal activity data.  The recent sharp decline in the velocity of currency in circulation (Chart 12) provides one illustration of these problems.

Chart 12:  Velocity of Currency

CAN MONETARY POLICY BE MADE TO WORK?

WE ARE starting the 1990s against the fairly widespread belief in the community that financial deregulation, and the operation of monetary policy in the financially deregulated environment of the 1980s, were a disaster, leading to the present recession we should not have had.  This belief has, in turn, reinforced questioning of the moves towards deregulation in other areas.  According to one recent commentator:

"The economy has been brought to its knees by financial and economic deregulation, the elimination of tariffs, free trade in agriculture, open slather for imports, privatisation, high interest rates, a taxation system that favours consumption over saving and investment and is an administrative nightmare, and budgetary policies that treat surpluses as triumphs of financial management ... The financial system is in tatters as a result of its own greed and extravagant lending policies ... Monetary policy, with its reliance on high interest rates, has been so perverse as to prevent the adjustments in the exchange rate which would occur if foreign exchange markets were really free to find their own level ... Free-market policies are killing the Australian economy and causing hardship and ruin for millions of Australians." ⁽⁹⁾

These are the considered comments not of a Democrat Senator, I should emphasise, but of a Professor Emeritus at the ANU, Russell Mathews.

Such views seem to derive from three main conclusions about financial deregulation and monetary policy.  First, financial deregulation has been a disaster because it allowed, indeed encouraged, a lowering of lending standards that led to a large increase in borrowing for various speculative and unproductive purposes.  That borrowing, in turn, created an excess demand problem, reflected in an unsustainable current account deficit and high inflation.  In turn, this excess demand and high inflation had to be dealt with by a tightening of policy.

Second, too much weight was, in practice, placed on monetary policy to restrain domestic demand;  ever higher interest rates, and the upward pressure on the exchange rate which accompanied them, had adverse effects on business investment and the traded goods sectors in particular.  The implication is that other policies should have been tightened to a much greater extent, presumably mainly fiscal policy.

Third, the increase in interest rates that resulted was held at "high" levels for too long.  The implication is that we did not "have to have" a recession:  that is, policy did not need to have been tightened so much.

Essentially, the question we want to try to answer is why monetary policy does not appear to have worked during the 1980s -- why we have ended up having a serious recession when the object of policy is supposed to be to prevent such occurrences.  This is not simply a theoretical issue.  We want to know the answer in order to try to prevent a recurrence, not of course in the immediate future but once the economy has started to recover.  It seems to me that the possible explanations can be boiled down to two broad ones.

One explanation is that monetary policy was operated inappropriately during the 1980s, whether because of errors of judgment, inappropriate objectives, or political interference -- or some combination of all three.

The alternative explanation is that the task was always beyond the capacity of monetary policy:  whether because of the effects of financial deregulation;  because the settings of other policies left it with too much weight to carry;  or simply because it is inherent in the present arrangements that political interference will prevent the achievement of "sound money".  Again, the truth may contain elements of all three such potential influences.

I want to outline briefly my own views on this and on what might be done about it from the perspective of a "practical" economist who does not pretend to any deep knowledge of monetary theory.  By way of background, I start by referring to what appears to be the official Reserve Bank explanation as outlined at the June 1991 Conference The Deregulation of Financial Intermediaries -- organised by the Reserve Bank, and attended by 32 selected invitees.  No journalists were invited, but edited proceedings have recently been published, and include a paper by Assistant Governor (Economic), Ian MacFarlane, on "The Lessons for Monetary Policy."

I hope that I do not do MacFarlane or the Reserve Bank an injustice if I say that he apparently absolves monetary policy from any overall blame for the 20 per cent annual growth in credit between the end of 1983 and mid-1989, which he acknowledges to be "credit excesses."  While he admits that it would be "foolish to claim that the setting (of monetary policy) was always right", and he concedes that the easing of monetary policy in 1987 may have been overdone, his conclusion is as follows:

"What could monetary policy have done about it?  In principle, it would have been possible to devise a monetary policy tight enough to have prevented the acceleration in credit.  However, it would have had to have been exceptionally tight, and in my view, tighter than would have been in the overall macro-economic interests of the country.  As it was, monetary policy was characterised by the highest real interest rates in the OECD area, and was tight enough to contribute to a major reduction in the inflation rate.  It would have been very costly for the economy as a whole to have tightened the screw with monetary policy sufficiently to offset all the effects on credit of financial deregulation, in particular the increase in the number and size of providers of credit." ⁽¹⁰⁾

In short, the Reserve Bank's view seems to be that the "credit excesses" could only have been stopped if, to revert to one of its Governor's more colourful recent utterances, Attila the Hun had been in charge of monetary policy.  The Bank is presumably not aware of the fact that The Oxford Classical Dictionary describes Attila as having been "a blustering, arrogant character, a persistent negotiator but not pitiless" -- hardly tough enough or decisive enough, therefore, to run Australian monetary policy!

But, if monetary policy failure was not responsible for the "credit excesses", what was?  In his conclusion, MacFarlane says,

"There can be little doubt that deregulation of financial intermediaries played a major role, but just as important was its timing.  It was unfortunate that deregulation occurred at the very time that inflationary expectations were at their peak, and where conventional business wisdom viewed debt so favourably because of its perceived effectiveness." ⁽¹¹⁾

What this seems to be saying (with the benefit of hindsight) is that, given the economic environment, nothing more could have been done to prevent the credit excesses of the 1980s once financial deregulation was decided upon.  This conclusion is supported by reference to the similar experience of some overseas countries, most notably the US and the UK, where, no doubt coincidentally, they also have Central Banks bearing many resemblances to our own.  In any event, according to MacFarlane, we do not need to be concerned about the experience of the 1980s, because that experience is unlikely to be repeated in the 1990s.  Apparently banks have learnt the lesson and lower inflation will encourage better quality investments.

In assessing the MacFarlane article we have to start by recognising that any Reserve Bank official is necessarily constrained in making comments about the past conduct of monetary policy -- particularly when the Treasurer for the relevant period is on record as stating of his Reserve Bank officials that "they do as I say."  But what I can confirm from personal knowledge is that the conduct of monetary policy under both Liberal-National Party and Labor Party Governments has been influenced by short-term political considerations, such as the timing of both Federal and State elections.  It is also clear that Australia has not been, and is not, alone in that regard.  The precise extent of that influence under the Labor Government is impossible to gauge.  However, if we assess the trend in monetary policy by reference to changes in the relationship between cash rates and the 10-year Treasury bond rate, Chart 1 shows that there was an easing of policy just before the December 1984, July 1987 and March 1990 elections.  Only in the last case did the easing continue after the election.  The easing in the lead up to the July 1987 election is particularly suspect as having been politically influenced.  I note that MacFarlane now acknowledges what I have previously pointed out on more than one occasion (and at the time), namely that the easing before the share market crash in October 1987 was "somewhat greater than it should have been" ⁽¹²⁾ -- a description of the 4.5 percentage points cut in cash rates between January and October 1987 which would have done credit to Attila the negotiator!

Chart 1:  Trends in Monetary Policy

One could argue that these periods before the elections were appropriate times in which to ease monetary policy.  Yet whatever the misuse of monetary policy for short-term electoral purposes, that is only part of the story.  More important was the flawed economic strategy which the Labor Government adopted right from the outset in 1983.  The essence of that strategy was to "manage" the economy so as to maximise economic growth and minimise unemployment, through an expansionary fiscal policy, with appropriate support from other policies.  A key component of the strategy was to have monetary policy playing a secondary "neutral" role, not bearing down on inflation but at the same time not encouraging it to increase.  The then Governor of the Reserve Bank acquiesced in monetary policy simply having the broad aim of providing for an expansion of credit adequate for recovery without fostering inflationary tendencies.  The situation was summed up in Statement No. 2 in the 1983-84 Budget papers (a statement which according to established practice would have been drafted in consultation with the Reserve Bank), as follows:

"The Government's approach to monetary policy is to provide just sufficient monetary growth to finance the prospective increase in nominal output considered desirable and feasible in the circumstances.  That is, sufficient money supply growth will be provided to accommodate the expected increase in real output plus what is considered to be a desirable and feasible increase in prices, with the latter being determined on the basis of the operation of the centralised wage system under conditions consistent with the Prices and Incomes Accord.  The relationship between monetary policy and the Accord is, therefore, a close and important one" ⁽¹³⁾ [emphasis added].

How, then, was inflation to be contained, even reduced?  The answer, of course, was through the Accord -- by which trade-union leaders agreed to deliver wage restraint in return for concessions on the social wage, including tax cuts.  The theory was that, if the expansionary policy led to balance of payments problems, the floating exchange rate would take care of those problems by depreciating.  The arch-exponent of this incomes policy solution, (now) Professor Barry Hughes, was continually on watch in this early period of the Labor Government to ensure that monetary policy did not become "tight" lest it hold back growth.  Once the constraint of the wages break-out was removed by the Accord, Australia was in a position (according to this theorist) where it could step up its annual growth to 4-5 per cent, well above the OECD average, without fear of inflationary or balance-of-payments consequences.

This is not the place to analyse, in detail, the Accord and its claims to have reduced inflation through wage restraint:  it has been done elsewhere. ⁽¹⁴⁾  One point I want to emphasise, however, is that the fall of about 3.5-4.0 percentage points in the average inflation rate between 1982-83 and 1983-84 -- from 11 per cent to 7.0-7.5 per cent -- can hardly be attributed to the Accord.  It had much more to do with the Australian and international recessions and the delayed effects of the sharper downward adjustments in overseas inflation rates. ⁽¹⁵⁾  Moreover, between 1983-84 and 1989-90 there was no significant further reduction in the underlying Australian inflation rate.  It apparently required a further recession to bring inflation down to the current rate of around 3 per cent. ⁽¹⁶⁾  Thus, for the period in which the Accord could claim to have been operative we had an underlying annual inflation rate that averaged around 7 per cent, and that exhibited only a slight downward trend from early 1987 on.  It should also be recalled that the recession-induced (and "wage pause"-induced) fall in Australia's inflation rate from mid-1982, which brought us back towards that of our major OECD trading partners, was then thrown away under the Accord -- so that we did not again touch base with the OECD average until the current recession. ⁽¹⁷⁾

Chart 2:  Inflation

What conclusions can we draw from all this?  The main conclusion is that the holding up of Australia's inflation rate relative to that in our major trading partners reflected the Accord strategy of the "dash for growth" and the effective neutering of monetary policy.  I am not suggesting that the main fault lay with the Reserve Bank;  on the contrary, the main fault lay with the basic strategy adopted by the Government.  Of course the Bank could have challenged that strategy, and I believe more could and should have been done in various ways in that regard. ⁽¹⁸⁾  But, granted the Accord philosophy, any attempt by the Bank to give priority to reducing inflation would have fundamentally challenged the economic philosophy of the Government.  Short of such a challenge, it would have been difficult for the Bank not to accept the rate of inflation which was effectively set when the Industrial Relations Commission largely rubber-stamped the wage increases agreed between the Government and the ACTU. ⁽¹⁹⁾

I agree, therefore, with MacFarlane's basic proposition that the task of preventing the "credit excesses" was beyond the capacity of monetary policy.  However, I differ deeply on the reason.  To my mind, monetary policy was hamstrung not by financial deregulation but by the Accord:  and the general economic strategy that went with it, including the (non-existent) role assigned to monetary policy within that strategy.  This strategy effectively said to borrowers:  "We are going to pursue a policy of high growth and, while we will keep inflation under "control", we will not be pushing to bring prices down."  In circumstances of financial deregulation, and more or less assured access to funds for all and sundry, this constituted a virtual guarantee of a large increase in the demand for loans, particularly when interest costs are tax-deductible.

Now, it is true that at the time nobody foresaw this leading to an actual explosion in borrowing and an asset price boom and bust.  There were a few, however, who opposed some of the financial deregulatory measures on the grounds that other policies, including an anti-inflationary policy, were not in place to handle such a move. ⁽²⁰⁾

Be that as it may, there remains the question of whether monetary policy would have been able to cope with financial deregulation if the Government had had a different economic strategy.  In pointing to the state of inflationary expectations at the time, and to the failure also of some overseas countries to cope with deregulation, MacFarlane seems to be saying that the task would have been beyond monetary policy anyway.  If so, that seems to be a strange attitude to adopt -- a kind of throwing-up of hands (or washing them, as the case may be).  Surely, if the Government had espoused a policy of bringing down inflation and inflationary expectations, instead of accommodating them, the situation would have been markedly different. ⁽²¹⁾  Moreover, the underlying economic circumstances were favourable for such a policy to have been pursued, since Australia's inflation rate had dropped between 1982-83 and 1983-84, and since overseas inflation rates had dropped even more.  The floating of the exchange rate was also relevant, given that this was supposed to provide Australia with the opportunity to run an "independent" monetary policy.  With a more appropriate economic strategy, it would (I consider) have been quite feasible to have reduced inflation much further.  Even if this had involved significant short-term economic costs (which is by no means certain), they would undoubtedly have been as nothing to the economic costs, both short-term and longer-term, which failure to make major reductions in inflation has now imposed on us.  After all, such a policy at the time would have avoided much of the "credit excesses" that occurred, and would thus have bid fair to prevent the severest recession since the 1930s.

The purpose of going back over this history is not to try to score points off either the Government or the Reserve Bank.  Rather, it is to assess and draw out the implications of what happened so that we can try to establish a framework that gives us a better chance of avoiding a repetition of the past.  Of course, we also have to be careful, as Mr MacFarlane reminds us, not to go on "fighting the last war."  However, while I accept that the credit excesses of the 1980s have been a chastening experience for many, and while going into debt is evidently very much out of fashion these days, it is by no means clear to me that "a return to normality, even frugality" will necessarily be sustained through the 1990s.  Inflation is far from beaten, and it would be rather complacent to sit pat and rest upon the hopes which the Reserve Bank now holds out to us.  Is the Bank now telling us that if inflation gathers pace again (even if not quite so frantically as before) we will not see the re-emergence of smart people gearing themselves up (in both senses of that term) to take advantage of it?

Of course, from one perspective that is not happening.  The Reserve Bank is responding by greatly increasing its prudential supervision of banks.  There is a parallel tightening of supervision of other financial institutions and of companies generally.  Yet if there is any "fighting the last war" it may very well be this process of re-regulation.

Indeed, if MacFarlane is correct in suggesting that banks and their customers "will have learnt from the errors of the 1980s", why do we need this massive increase in prudential supervision?  The clear implication is that the over-lending and over-borrowing that occurred in the 1980s was due to excessive competition that could have been prevented by greater regulation and/or supervision.  This is, indeed, the conclusion which has now been reached by the recent Martin Report.  My argument is, rather, that the main cause of the "credit excesses" was an inappropriate economic strategy and the "accommodative" monetary policy that went with it.  It is little short of astonishing that the Martin Report does not even address this issue.

There is a real risk that the move towards greater prudential supervision will reduce competition among financial institutions, and/or create a moral-hazard problem which will make it more difficult to operate monetary policy.  There are already a number of features of the Australian banking industry which restrict competition and/or create a potential moral-hazard problem.  These include:

• The Banks (Shareholding) Act, which effectively prevents a takeover of a bank except by another bank;
• The ban on new foreign entry and the inhibition on foreign bank activities resulting from the requirement that foreign banks be subsidiaries rather than banks;  and
• The existence of government guaranteed banks which control some 30 per cent of the assets of the banking system.

Thus, rather than moving to increase prudential supervision, it would be better to attempt something to improve competition by changing these arrangements.  To give it credit, the Martin Committee has recommended the sale of the government-guaranteed State Banks and the removal of restrictions on foreign bank entry.

Much more important, however, there is a need to review the whole framework within which monetary policy operates.  In this regard, I note the following conclusion of a paper by Messrs MacFarlane and Stevens presented at the October 1989 Reserve Bank Conference:

"... while both the conduct of monetary policy and the way it has its effects on the economy have changed over the past decade, the most important lesson of the 1970s is still valid:  that while monetary policy can and does offset activity in the short run, its ultimate goal should be price stability." ⁽²²⁾

One might simply add that, given the experience of the 1980s, the same lesson applies also to that decade.

However, the object is not simply to pursue price stability for its own sake, important though that is.  Underlying the aim of price stability is also the belief that, the more closely it is achieved, the better the real economy will perform, in terms both of longer-term growth and of reduced fluctuations.  A non-inflationary world is more likely to discourage "speculative" investments and to encourage "productive" investments.  Moreover, we know from recent experience and from history that continually rising rates of inflation tend to lead in the end to a "bust" which destroys confidence and produces a recession.

This is not to suggest that monetary policy should try to eliminate or smooth all fluctuations in the real economy, or even in prices.  The goal should be a medium-term one.  What we should be trying to avoid is a repetition of the sort of asset price boom-bust we have just been through, or the consumer price boom-bust that was experienced in the mid 1970s.  It is through such a preventive role that monetary policy should be used to influence the real level of economic activity.  By creating a monetary environment that encourages stable expectations about prices, the aim is to prevent the loss of growth which results when we have a recession caused by a price boom-bust.

It should be noted that asset price boom-busts are one of the possible causes of recessions, and that these seem to be able to occur even when there is a relative stability in the rate of consumer price inflation.  I wonder whether the characterisation of the 20th century as the century of inflation has tended to overlook the fact that this phenomenon occurred in earlier centuries and was reflected, in turn, in fluctuations in the real economy.  The point is relevant to how we define the goal of price stability.  If asset price boom-busts can cause or at least lead to recessions, this suggests that one should not simply aim to stabilise consumer prices.  There are obvious difficulties here:  but, given the experience of the 1930s and the obvious potential effects for real levels of demand to be affected, there maybe a need for monetary policy to have greater regard to trends in asset prices. ⁽²³⁾

There is also a question as to whether the current account deficit should be a policy target, something which it has now clearly become as a result of past policy failures and the heavy build-up of foreign debt which has resulted from them.  This poses particular difficulties for monetary policy, given the potential for the economy-wide demand-reducing effects of "high" interest rates to be offset by the adverse effects on the trade balance -- effects arising from the higher exchange rate which those high interest rates produce.  Clearly, therefore, the use of monetary policy to attempt to improve the external balance should be minimised and primary reliance should be placed on fiscal and other policies.  There may, nonetheless, be circumstances where the choice is between using monetary policy or having an external crisis which would cause a recession.  I believe that those circumstances arose in mid-1986 and probably again in 1988-89. ⁽²⁴⁾  It is important to note, however, that the need to rely on monetary policy on those occasions would probably never have arisen if there had existed an earlier firm commitment to price stability.

If it is agreed that price stability should be the goal of monetary policy, the question is how best to achieve that goal.

Essentially there are two questions here.  First, whether, if we set our monetary policy the goal of achieving price stability, that goal can be attained and sustained from a technical operational viewpoint.  Second, whether price stability can be attained and sustained as a policy goal, and (if it can) what is the best framework for doing that.

As to technical capability, it appears that financial deregulation and technological changes have made it even more difficult than previously to set with any confidence a target for the growth in a monetary aggregate (or aggregates):  growth that would achieve price stability without risking a recession in economic activity on the way. ⁽²⁵⁾  This difficulty has led to the adoption of interest rate targets as the operational tool of monetary policy, on the ground that manipulation of the yield curve produces a more reliable response in the real economy.  One must question, however, how much confidence there can be in the response of the real economy to the use of the yield curve.  Certainly, it appears that a lot of judgment is still required in varying the settings of monetary policy to try to achieve and sustain price stability.

This is relevant to the more important question of how to attain and sustain price stability as a policy goal.  As recent debate about the trade off between growth and prices indicates, achieving the goal of price stability will not be accepted by the community if that achievement requires large sacrifices in output and employment.  Equally, its maintenance will not be accepted if that requires the acceptance of such sacrifices.

At least part of the answer is that price stability would be achieved over the medium term, thus minimising any short-term effect on the real economy;  once the goal is set and accepted as likely to be sustained, adjustment in expectations should have favourable effects on the actual rate.  This is particularly relevant to the wage-bargaining process.

In his recent comments about the possible adverse effects which setting a 0-2 per cent inflation target would have on the real economy, the Governor of the Reserve Bank appears to have overlooked the potential for favourable expectational effects:  not to mention the fact that -- as comments by the Governor himself have stressed -- the starting point for achieving stability is now much more favourable than it was two years ago. ⁽²⁶⁾  This is not to deny the need, under carefully specified circumstances, to allow discretion to cope with once-for-all shocks to the general price level (e.g. the Gulf War oil price shock).

The most difficult part of the exercise of achieving and sustaining price stability is how to minimise the potential for political interference, or at least ensure that any political intervention is open and subject to public debate.  I say "minimise" because I do not believe that the elected Government can or should be deprived of the right to determine what it believes to be the desirable objective for prices.

The solution to this difficulty seems to me to be for the Government to set the period within which price stability is to be achieved, and to specify in advance what allowances may be made for varying the price stability objective to accommodate shocks.  In short, it would be the Government that decides on the policy priority to be given to price stability.

Accordingly, I see the achievement of price stability as a policy goal as being much less a matter of giving the Reserve Bank "independence" than of securing political acceptance of such a goal as being feasible and desirable.  The political reality is that, at least for the foreseeable future, no government will hand over complete authority to a group of bureaucrats to determine relative priorities in a key area of economic management.  Nor, I believe, should any government do so.

The Coalition parties' proposals to "ensure" the independence of the Reserve Bank and to hold the Bank "publicly accountable for its performance" have attracted some attention.  However, the key part of the Coalition package in the monetary area is its commitment to the medium-term objective of price stability, defined as an inflation rate of 0-2 per cent.  Without that commitment, the Reserve Bank would be put in an almost impossible position, just as it was in the 1980s.  Equally, if that commitment is implemented it will constitute a significant advance in terms of both monetary policy and economic policy generally.

CONCLUSION

The main lesson of the 1980s is that the Government's basic economic strategy was flawed, particularly (but not only) in regard to the failure to give any significant weight to price stability.  That failure, and the associated attempt to push Australia's growth beyond the limits, was the basic underlying reason for the "blow-out" in Australia's external debt and for the assets price boom-bust.  The way in which monetary policy was operated contributed importantly to this situation, although the chief culprit in this regard was the Government, not the Reserve Bank.  Had the Government committed itself to working towards price stability -- or indeed given any role to monetary policy other than to be "accommodating" to the needs of the Accord -- monetary policy should at the very least have been able to prevent the credit "excesses" that occurred.

The recession that we should not have had has created a situation where, for the moment, there is concern that businesses and others are too reluctant to go into debt:  the reverse of the 1980s' situation.  It is remarkable that the Governor of the Reserve Bank now emphasises the difficulties of achieving and sustaining low inflation and low inflationary expectations.  Ironically, this seems to run counter to the thrust of the very recent thesis of the Assistant Governor (Economic), that the change in circumstances means much less need for concern about "fighting the last war" against inflation -- because low inflation has already been achieved, and (he argues) is likely to be maintained.

Be that as it may, I find it difficult to accept that any serious central banker could think inflation to be permanently beaten.  Unless otherwise constrained, Governments will in due course again be tempted to pump up the economy for short-term electoral purposes.  If they do, inflationary pressures will re-emerge.  What would the Reserve Bank do in such circumstances?  Again accept it as a fait accompli?  If so, what reasons will be given for the next round of "credit excesses"?  The deregulation excuse, at least, will no longer be available.

If Governments can commit themselves to achieving price stability, that would provide a basis for the Reserve Bank to point out any inconsistencies between that objective and, say, an expansionary fiscal policy.  Then the Government of the day could take the political responsibility for changing the commitment to price stability, or else suffer the consequences.  That would, to my mind, be a significant improvement on present arrangements.

---

A "BUNDESBANK" OR A "RESERVE CURRENCY BOARD"?

"How does one install and insure stable money in Albania?  In my view, the only sure-fire way is to establish a currency board.  The Hanke-Schuler volume presents a sound blueprint for doing just that."

-- Sir Alan Walters

THIS paper argues that Australia's monetary policy has operated poorly, largely because the Reserve Bank has not been free of political constraint and has been unable to pursue price stability objectives.  While deregulation exposed weaknesses in banking practice, ⁽²⁷⁾ as bankers moved up the competitive learning curve, the real culprit was excess credit and permissive monetary policy.  A more detailed assessment of monetary policy in the 1980s, and particularly under the Hawke/Keating control, is published elsewhere. ⁽²⁸⁾

The 1980s' high inflation, particularly of asset values, interacted with the tax system (which permits full deductability of nominal interest).  It created the excessive borrowing, a policy reversal, and the subsequent sharp downturn, from which we still suffer.  It is crucial to note that it was mismanagement of monetary aggregates and monetary policy which were the problem -- not the liberalising and internationalising of our capital markets (Chart 1).

Chart 1:  Changes in Consumer Prices

The root cause of our debt woes surge in the 1980s was inflation, and the root cause of that was poor conduct of monetary policy.  For example, foreign exchange intervention in the late 1980s was inconsistent with the float:  it created further inflow of capital and so required higher interest rates to "mop up" excess liquidity.  Our monetary aggregates, and through them our interest and inflation rates, were quite out of line with those of our competitors.  We had, in fact, repeated in the late 1980s the monetary mistakes of the early 1970s.  This failure to allow market adjustment of the exchange rate, and Keating-style monetary "they do as I say" activism, meant that our "floating" of the dollar did not enable us to achieve an independently low rate of expansion of domestic credit as was achieved in Germany by the Bundesbank.  The Charts summarise the outcomes for Australia and Germany.

The paper presents two major options:

• to restructure the Board and the Reserve Bank Act to facilitate real independence of the Reserve Bank of Australia, along the lines of the Bundesbank or the new structure of the Reserve Bank of New Zealand;  or
• to convert the Reserve Bank into a currency board along Hong Kong/Singapore lines, removing the scope for discretionary monetary policy, since policy activism has hurt rather than helped the cause of economic stability (Chart 2).

Chart 2:  Domestic Credit

THE AUSTRALIAN MONETARY PROBLEM

The paper "Treasurer Keating's Legacy" assesses the conduct of monetary policy under the stewardship of then Treasurer and now Prime Minister, Paul Keating. ⁽²⁹⁾  What the paper found was that, despite courageous moves to make the financial system competitive and to float the Australian dollar, there was a serious loss of monetary control under the Hawke Government.  Inflation, interest rates and monetary growth all greatly exceeded those of our OECD competitors.  Particularly disappointing is that there has been more, not less, foreign exchange intervention since the Australian dollar was floated.  Chart 3 on movements in foreign exchange holdings makes this clear.

Given, then, that we had an activist policy in money, bond, and foreign exchange markets under Keating;  given that inflation remained high until very recently;  and given that interest rates were forced to excessive levels in a belated attempt to turn off inflation;  given all this, there is no way that monetary performance in Australia can be regarded as satisfactory.  The theoretical and empirical case for discretionary monetary policy has never been strong, and is not helped by an analysis of Australian and RBA performance.  What is worse, the incapacity to unscramble the economy quickly means that all manner of protectionist interventions now hang over Australia, as business tries to undo the damage of monetary mismanagement.

We present these comparisons of Australian and German monetary outcomes since 1960 largely because the debate seems likely to come down to a choice between the following two options:  a Bundesbank-style restructuring of the Reserve Bank, or an Australian Currency Board.

Chart 3:  Australia:  Foreign Exchange

We argue that the major institutions in Australia -- the ACTU, the business lobbies and other key interest groups -- are unlikely to let the Governor of the Reserve Bank act as independently as the President of the Bundesbank has acted (at least until last year).  The political risks, if you like, are so great that Australia would be well advised to go the currency board route.

A BUNDESBANK FOR AUSTRALIA?

Judging both by the evidence on the superior conduct of German monetary policy and by the literature on discretionary monetary policy, there is much to gain from legislation which secures a Bundesbank-style independence of the Reserve Bank;  which removes the Treasurer's representative from the Board;  and which adopts a New Zealand-style price stability objective.  But, for reasons set out below, we would now go further than John Hewson is proposing.  Along lines set out in Hanke and Schuler, we suggest a shift to a currency board of the Singapore and Hong Kong variety.

Chart 2 shows the expansion of domestic credit in Australia and Germany since 1960.  We have chosen domestic credit because that is the measure of domestic financial outcomes on which the IMF tends to focus in comparing the domestic contribution to monetary policy across countries.  When the IMF agrees to support economic reform packages, the required policy adjustments invariably make reference to the behaviour of "domestic credit".  Ian MacFarlane's important review of financial deregulation ⁽³⁰⁾ also focused on credit as the variable which "got away."  And the Australian/German domestic credit graph is a good illustration of just how different the behaviour of domestic credit has been in the two countries since their respective floats.

Before the Whitlam Government, Australian domestic credit grew modestly, and often by less than domestic credit in Germany.  Consistent with this, pre-Whitlam, Australian inflation rates and nominal wage growth were in line with or below those in Germany (Chart 4).  So, whatever lessons there are, history suggests that it is possible under current institutional arrangements for monetary policy to be managed as well as the Germans manage it.  That is, until you allow for the recent conduct of monetary policy, and the underlying reasons for monetary policy getting out of kilter.  Keating-style activism may well be applicable in some areas of government, but not in the conduct of monetary policy.

The gaps in inflationary outcomes between Australia and Germany correlate strongly with the Whitlam and Hawke/Keating periods, as is vividly demonstrated in Charts 1 and 2.  These unsatisfactory outcomes reflect not least the presence of a non-independent central bank -- one which accommodates governments inclined to run large deficits, which reflects wage pressure and which generally does what the Treasurer or government of the day wants.  The most recent "Keating" gap in performance is over a period when our region was prosperous, our markets strong and when our currency was not hooked to any inflating currency (cf. pre-1972).

Chart 4:  Wages Growth

To me, the lesson of monetary history in Australia and around the world is rather simple.  The first lesson is that monetary discipline and price stability are achievable if the central bank is allowed to pursue them.

The second point is that monetary policy cannot achieve much by way of fine tuning.  While monetary policy can certainly turn an economy off, there is no precise understanding of how monetary expansion can turn an economy back on.  1991-92 is a case in point.  We do know that inflationary expectations take a while to form -- witness the very low ex poste real rates of interest in the 1980s associated with what can now be seen as an unanticipated inflation.  Now that the community continues to expect high inflation, our real rates of interest are remaining high.  It will probably take time for inflationary expectations to be pulled back, particularly if we have no reason to be confident about central bank discipline.

This potential volatility of real rates of interest makes a powerful case for a steady hand at the monetary tiller.  It argues for simple price stability objectives, with the Reserve Bank choosing monetary and other rules -- and we would suggest that they look at the monetary base -- for achieving these objectives.

However, while these reforms would help, we fear they will be inadequate.  For example, a major difficulty of any proposed Reserve Bank arrangements for Australia is that the Governor would continue to be subject to political influence, as was the President of the Bundesbank in relation to 1990 currency unification.  The history of Australian unions, of business lobbying and of the political cycle all indicate that it may be best to remove the scope for activist or discretionary monetary policy.  This makes a case for the formation of a currency board along Singapore or Hong Kong lines.  This board would issue domestic money in exchange for its foreign assets holding.

While it will, no doubt, be hard for politicians to give up discretionary monetary policy, the burden of Australian and international evidence is that monetary policy is best left outside the political domain.  The experience of Hong Kong, Singapore and other currency boards should serve as a major inspiration for Australian monetary reform.

THE CASE FOR A CURRENCY BOARD

The Reserve Bank of Australia has been highly susceptible to domestic political influence.  As a result, it has lost considerable credibility.  To have a stable currency, Australia needs to remove monetary policy from political influence.  Australia needs to give its monetary reform credibility, to avoid the dangers of continuing inflation on the one hand and deep recession on the other.  A proven way to do so is to remove from the Reserve Bank of Australia its currency-issuing functions, and to establish a currency board, whose only job will be to issue a convertible currency according to strictly defined rules.  This paper sets out the issues and steps involved in moving to a currency board.

• A currency board is explicitly designed to maintain a fixed exchange rate.
• It would quickly establish a hard domestic currency and instil monetary confidence.  Thus, economic agents would alter their expectations.  The government would have to finance itself exclusively by taxation and borrowing, not by inflation, because a currency board is not an agent of government finance.
• Using a foreign reserve currency for a currency board would not subject Australia's currency system to foreign political influence.  The currency board, and through it the government, would not be giving anything to the government that issues the reserve currency, because the board would hold almost all its assets in interest-earning securities.

WHAT IS A CURRENCY BOARD?

A currency board is an institution that issues notes and coins convertible into a foreign "reserve" currency at a fixed rate and on demand.  It does not accept deposits.  As reserves, a currency board holds high-quality, interest-bearing securities denominated in the reserve currency.  A board's reserves are equal to 100 per cent or slightly more of its notes and coins in circulation, as set by law.  A currency board generates profits (seigniorage) from the difference between the interest earned on the securities that it holds and the expense of maintaining its note and coin circulation.  It remits to its owner (which historically has been the government) all profits beyond what it needs to cover its expenses and to maintain its reserves at the level set by law.  The currency board has no discretion in monetary policy;  market forces alone determine the money supply.

The main characteristics of a currency board are:

• Convertibility:  The currency board system assures that the currency will be convertible at a fixed rate.
• Reserves:  A currency board's reserves are adequate to ensure that, even if all holders of notes and coins want to convert them into the reserve currency, the board will be able to do so.
• Seigniorage:  Unlike securities or most bank deposits, notes and coins do not pay interest.  Hence, notes and coins are like an interest-free loan from the people who hold them to the issuer.  The issuer's profit equals the interest earned on reserves minus the expense of putting the notes and coins into circulation.  Under a currency board system, the domestic currency is as sound as the foreign reserve currency.  The only economic difference between using a domestic currency issued by a board as legal tender, instead of a foreign currency, is that the seigniorage generated by a currency board issue is captured domestically;  whereas, if a foreign currency is used as legal tender, the foreign issuer captures the seigniorage.
• Monetary policy:  By design, a currency board has no discretionary powers.  Its monetary policy is completely automatic, consisting only in exchanging its notes and coins for the foreign reserve currency at a fixed rate.  Since a currency board's role is strictly circumscribed, it is less likely than other monetary systems to suffer political pressures to engage in economically unsound policies.

Over 60 countries have had currency boards during this century.  (New Zealand had a currency board from 1850 to 1856, but Australia never had one.)  Despite currency boards' success, only a few such monetary systems exist today, most notably in Hong Kong and Singapore.  Most other countries that once had currency boards replaced them with central banks.  These changes were made for political, not economic, reasons.  Politicians saw central banking as a way of manipulating the money supply to their own advantage.  Since abandoning the currency board system, many of those countries have experienced inflation and economic stagnation.  Hong Kong and Singapore, on the other hand, have been two of the world's most rapidly growing economies, despite their lack of natural resources.  Moreover, they have realised relatively low rates of inflation.  There are, of course, ways in which the money supply under a currency board can be subject to political manipulation:  for example, by arranging large foreign borrowings against government guarantees.  But we judge these "political risks" to be much less than the risks with discretionary central banking.

HOW A CURRENCY BOARD WORKS

The currency board system relies entirely on market forces to determine the amount of notes and coins that the board supplies, and also to determine the amount of deposits (and other components of the broader money supply) that banks and other financial institutions provide.  The central bank of the reserve-currency country determines the supply of reserves in the whole currency area, including the currency board country.  Competition among commercial banks in the supra-national currency area determines the distribution of the reserves, including how great a proportion of the total becomes the foreign-currency reserve of the currency board country.  The currency board has no role in determining the supply of reserves, because its 100 per cent reserve requirement makes it merely a sort of warehouse for reserves.  Since a board cannot influence the amount of reserves, it cannot influence the total supply of credit.  This stands in contrast to central banks.

In a currency board system -- as under a gold standard, or gold exchange standard -- the amount of credit which banks can create (and hence the total money supply) is limited by their ability to acquire and keep enough reserves to support that amount of credit.  This prevents a currency board system from experiencing inflation as high as the level possible under a floating exchange-rate, central-banking system.

Commercial banks act as middlemen between lenders (depositors) and borrowers (people who spend bank loans).  A bank cannot for long grant more credit to borrowers than depositors wish to let it have.  If a bank grants excessive credit, the borrowers will spend that credit (for instance, by writing cheques), and more funds will flow out of the bank than flow into the bank from cheques written on other banks.  To prevent this sort of mistake resulting in bankruptcy, a bank needs to hold reserves.  The reserves protect it from the consequences of its occasional mistakes.

The ultimate reserves in a currency board system are holdings of the foreign reserve currency.  The only way to acquire new reserves, obviously, is to obtain currency from the reserve-currency country.  In its simplest form this requires running a balance-of-payments surplus.  Simplifying certain assumptions for the sake of clarity, changes in the balance of payments change the total domestic money supply in the same direction.  A balance-of-payments surplus increases the total domestic money supply.  A balance-of-payments deficit, on the other hand, decreases the total domestic money supply.  Later we shall explain how under less simple, more realistic assumptions, investment inflows can enable the domestic money supply to expand even if there is a balance-of-payments deficit.  (Recall that the balance of payments is the value of exports minus the value of imports.  Recall also that the domestic money supply is made up of the currency board's notes and coins in circulation plus commercial bank deposits.)

There are two important points to notice about the adjustment process in a currency board system.  The first is that market forces rather than central bank action set it in motion;  it is completely "automatic", as far as the currency board is concerned.  The second point is that, because the exchange rate is fixed, arbitrage occurs entirely through changes in the quantity of money, interest rates, and the balance of payments, rather than through the exchange rate.

Hong Kong and Singapore have experienced balance-of-payments deficits for decades at a time, yet their domestic money supplies steadily increased because they were attracting large inflows of foreign investment.  Market forces of profit and loss determine and limit money supply expansion in the currency board country.

CENTRAL BANKING

The essential difference between a currency board and a central bank is that a central bank does not work automatically.  Rather, the central bank has discretionary power to influence the supply of money, and is not necessarily guided by considerations of economic profit and loss.  A currency board system is by nature a fixed-exchange-rate monetary system, while central banking is not.  The nature of central banking tends to drive central banking systems away from fixed exchange rates to floating exchange rates.  (If monetary disturbances were largely external, and not internal, then the case for floating rates would be strong.  But in Australia's case the "noise" has been domestic, and a fixed exchange rate exports that instability, and imports external currency stability.)

Central banks typically perform many other functions besides influencing the supply of money.  They regulate commercial banks, serve as lenders of last resort to the banking system, give economic advice to the government, and clear cheques.  However, all these functions are secondary to their role in influencing the money supply.  Only central banks control the supply of reserves in the banking system.

ADVANTAGES OF A CURRENCY BOARD OVER A CENTRAL BANK

The key difference between a currency board system and a central banking system is that a currency board has no power to carry out a discretionary monetary policy.  Two forces "pin down" the board's action:  a fixed exchange rate with a foreign reserve currency;  and a fixed (100 per cent or more) reserve ratio.  The board does not vary the exchange rate, nor does it alter the supply of bank reserves independently of changes in the balance of payments or in other market forces.

A fixed exchange rate eliminates -- without cost -- exchange-rate risk with the reserve currency.  Trade between the currency board country and the reserve-currency country becomes easier than it is under floating rates, because there is no need to allow for a risk premium in goods prices.  People can make more exact price calculations for internationally traded goods.  That enhances economic efficiency by making the lowest-cost producers within the common currency area those with the greatest natural advantages -- not those temporarily benefiting from the distortions to the international price structure that large, sudden exchange rate fluctuations cause.  A fixed exchange rate also enables entrepreneurs to apply to other problems talents that, in a floating rate system, they apply to currency speculation.

Eliminating exchange-rate risk also encourages foreign investment, particularly from the reserve-currency country.  Investors know with certainty what exchange rate they will receive in terms of the reserve currency if they should want to repatriate profits in the future.  By making it easy for them to leave the market, a fixed exchange rate is more likely than a floating rate to encourage them to enter the market.

A fixed exchange rate enables the country that sets the rate to "piggyback" on the reserve country's financial markets.  The currency of such a country is essentially a denomination of the reserve country's currency.  Accordingly, entrepreneurs in the country that sets the fixed rate can take their cues from the highly liquid, well-established markets in the reserve-currency country.  Easy access to large foreign financial markets, with no foreign-exchange risk, raises economic growth.

A fixed exchange rate, if credible, becomes a feature of the economic landscape and ceases to be a subject of political contention.  In particular, it enables the economy to avoid the vicious cycle of inflationary wage and price increases:  increases which cause pressure on the central bank to depreciate the currency as a way of keeping wages internationally competitive.  A currency board stops hyperinflation cold because workers and employers know that if they want to stay in business, wages and prices must be competitive from the start.  Since a currency board always has reserves of at least 100 per cent in assets that it can readily liquidate, it is always able to defend the fixed exchange rate.

It is, of course, possible to have a central bank that offers a fixed exchange rate with a foreign currency.  However, historical experience shows that it is not easy to maintain a fixed exchange rate in a central banking system.

In short, a currency board is an almost foolproof institution.  It cannot act as an independent disturbing element in the economy.  Market forces call the currency board's tune.  In contrast, a central bank has the power to destabilise the economy, and the history of central banks shows that they have often used that power, sometimes intentionally, but at other times by mistake.  A Reserve Bank of Australia, even if run by saints -- as long as they were not all-knowing saints -- would still not work as well as a currency board system.

The experience with discretionary central banking in Australia has been dismal over the last two decades.  While one answer is to try harder, or restructure the Reserve Bank on Bundesbank lines, there are many who see a need to remove the political risk inherent in a central bank which acts for the government of the day.  Currency boards are a tried and true method of avoiding such political risks and ensuring convertibility into a foreign currency at a fixed rate.

---

APPENDIX:  THE MECHANICS OF A CURRENCY BOARD

HOW TO ESTABLISH AN AUSTRALIAN RESERVE CURRENCY BOARD

It is fairly simple to replace a central bank with a currency board.  Central bank functions that do not relate to determination of the supply of money can be delegated to other government departments or to commercial banks themselves.  The central bank's deposit-creating powers can be abolished, its deposit liabilities can be separated from its note and coin liabilities, and then it can be converted into a currency board, issuing only notes and coins.

Here are two step-by-step plans for establishing a currency board.  The first assumes that Australia is replacing the Reserve Bank of Australia with a currency board.  The second assumes that the Reserve Bank will continue to exist, and that the currency board will issue a parallel currency that has equivalent legal tender status with the Reserve Bank currency.  The two currencies will not have a fixed exchange rate, though, unless the State Bank also decides to peg its currency to the same reserve currency that the currency board uses.

If Australia replaces the Reserve Bank with a currency board, the steps are:

• Delegate to other bodies all central banking functions that do not directly concern influencing the supply of money.
• Abolish the Reserve Bank's power to create credit.
• Make sure that existing reserves are appropriate.
• Convert all remaining commercial bank reserves at the Reserve Bank into currency board notes and coins or into foreign currency assets, as the banks prefer.
• Fix an exchange rate.
• Ensure that foreign currency reserves equal 100 percent of note and coin circulation.
• Transfer the Reserve Bank's remaining assets and liabilities to the new currency board and open the board for business.

If an Australian Reserve Currency Board were to come into existence as a parallel issuer alongside the Reserve Bank, the steps are even simpler, to wit:

• Announce a choice of reserve currency and exchange rate.
• Offer a small premium for all foreign hard currency for a short period, during which the board only makes exchanges from foreign currency into its currency.
• Cease offering the premium and begin two-way exchange with the reserve currency only, at the fixed rate.

HOW TO OPERATE AN AUSTRALIAN RESERVE CURRENCY BOARD

A currency board is simple to operate.  Past currency boards have usually had staffs of 10 or fewer people.  They have been able to achieve economies of staff by contracting some clerical and investment functions to outside parties.  This, in brief, is how one can run a currency board.

• Exchange policy:  The currency board's business is to stand ready to exchange its notes and coins on demand at a fixed rate into or from the reserve currency.  Hence, the currency board should not serve as a supplier of reserve currency notes to the public.  It should leave that to the commercial banks.  It should accept and pay in the reserve currency only by check or by electronic funds transfer.
• Clientele:  The public as well as banks should be able to deal directly with the currency board.
• Lower and upper limits to exchanges:  To reduce their handling costs, many currency boards did not exchange sums below a certain minimum.  The minimums prevented most members of the public from doing business with the currency board individually, which reduced the boards' need for clerks to handle transactions.  However, the minimums were low enough not to be a significant barrier to banks that wanted to do business with the currency board, or to private foreign-exchange dealers.  There should be no upper limit to the amount of the reserve currency -- or of the currency's own notes and coins in circulation -- that the currency board accepts for exchange.  No past currency board has ever had an upper limit to exchanges, because that would defeat the full convertibility into and out of the reserve currency that is the purpose of the currency board system.
• Commissions:  Some currency boards have charged commissions of 1/8 per cent to one per cent for every transaction.  If, as we suggest, the board has a minimum for transactions, it should not charge any commission.  Commissions would bring little income to the board.  The social benefits of not having commissions far outweigh the benefits to the board of having commissions.
• Offices:  The currency board should have a main office in Sydney, and perhaps branch offices or agents in other large cities.
• Management:  The currency board should have a small board of directors -- a typical size for past currency boards was five directors -- to oversee the board's managers.  The powers of the board of directors and of the managers will be quite limited;  they will have no influence over monetary policy as central bankers do.  To make the board of directors as independent from political pressures as it can be, directors should have staggered terms.  Furthermore, a majority or even all directors should be appointed by Australian commercial banks or even by foreign commercial banks, not by the Australian government.
• Staff:  The board's staff will perform two functions:  exchanging notes and coins for the reserve currency, and investing its assets in high-grade securities denominated in the reserve currency.
• Reserves -- composition:  The board should hold its reserves in high quality assets denominated in the reserve currency only.  It should not hold assets denominated in local currency, because that would open the way to central banking-type operations.  To prevent the currency board from becoming entangled in the politics of domestic government finance, a board should not be allowed to hold Australian government securities.  Liquid reserves (securities with maturities of two years or less) should be 30 per cent of the total.
• Reserves -- maturities:  It may be desirable to specify in the currency board's charter or by-laws what types of assets it may invest in and what the maximum maturity may be.
• Expenses:  Judging from the experience of past currency boards, expenses, excluding any loan repayments the board may have to make, should average no more than 1 per cent of total assets, and may be as low on average as 0.5 per cent of total assets.
• Profits:  The board's profits will be the difference between the interest it earns on its foreign currency reserve assets and its expenses, including repaying any loans it initially received.  After the board repays any initial loans, it should accumulate a reserve of 10 per cent to protect it against capital losses on securities holdings, as most British colonial boards did.  It should pay all profits into the reserve until the reserve is full, and in the future do likewise should the reserve ever fall below 10 per cent.  All profits beyond that should revert to the board's owner.

HOW TO PROTECT THE CURRENCY BOARD

The reason that currency boards outside Hong Kong and Singapore disappeared was, as mentioned earlier, that they lacked the political independence to prevent them from being changed into central banks.  Suspicion that a new currency board might be reconverted into a central bank would undermine foreign willingness to invest in the country, defeating one of the main advantages of convertibility.  Therefore we now propose ways of preventing new currency boards from suffering the fate that befell most old boards.  My proposals can be summarised as commitment, credibility, and competition.

The board can commit itself to buy and sell foreign exchange at the fixed rate.  A currency board could increase the attractiveness of the Australian dollar through these practices.

The government can improve the board's credibility by insulating it from any possible attempts at government manipulation.  One way to do so would be to have some of the currency board's board of directors be foreign nationals.

The board can be incorporated in a safe-haven country such as Switzerland, and it should be made clear that the board's assets belong to the board itself.

Competition will improve the currency board's incentive to maintain the fixed exchange rate.  Forced-tender laws should be abolished.  People should be able to make contracts in, and to use, any currency that they find mutually agreeable.  In particular, reserve currency notes and coins should be allowed to circulate alongside the currency board's notes and coins.

To subject the currency board itself to direct competition, banks could be allowed to issue circulating notes to compete with the board's notes.

---

THE OPERATION OF MONETARY POLICY
IN A DEREGULATED FINANCIAL SYSTEM

DEREGULATION of the Australian financial system, especially the removal of controls on interest rates and the exchange rate, gave the Reserve Bank of Australia (RBA) unprecedented autonomy in determining the rate of growth of its own balance sheet.  This, in turn, gave the RBA a greater ability to determine monetary conditions in Australia than it had possessed for decades.  The rate of growth of the central bank's balance sheet is the prime determinant of monetary conditions.

This paper discusses two aspects of the operation of monetary policy since deregulation:  the first is the mechanics of its operation (in other words, how monetary policy works);  the second is the mode of operation chosen by the monetary authorities (in other words, how monetary policy is used).  In each case, the discussion considers the way things are at present and proceeds to outline possible reforms which might improve the operation of monetary policy.

THE MECHANICS OF MONETARY POLICY ⁽³¹⁾

Prior to deregulation of the Australian financial system in the early 1980s, monetary policy was conducted primarily through direct controls imposed on banks.  The abolition of those controls obliged the RBA to administer monetary policy through alternative means.  The new arrangements emphasised market-oriented controls and were founded upon the RBA's monopoly position in the official short-term money market (or "cash market").  Under these arrangements, the central bank deals in the cash market in order to set the interest rate on overnight loans of cash (the so-called "cash rate").  By setting the cash rate, the RBA is able to exploit financial market linkages in order to influence the level and pattern of interest rates on securities of longer maturity, both official and non-official.

To understand how the RBA manipulates the cash rate requires an appreciation of the factors governing the demand for and supply of cash.  This, in turn, requires a detailed understanding of the institutional arrangements under which the RBA exchanges securities with the banks and with the authorised short-term money market dealers as part of its daily market operations.

To begin, we need a definition of "cash".  Legally, cash is any asset the RBA deems to be cash.  In practice, the RBA bestows cash status exclusively on its own liabilities -- principally currency and balances held by banks and authorised dealers in exchange settlement accounts. ⁽³²⁾

At the beginning of each business day, banks find their exchange settlement balances altered because of a wide range of transactions that have occurred over the preceding few days.  If the government has been spending more than it has received, banks as a group tend to accumulate surplus balances as cheques written by the government clear.  If the RBA has been particularly active in selling securities, banks find themselves with declining balances in their exchange settlement accounts.  In this respect, exchange settlement accounts resemble current accounts operated by corporations to provide working capital.

Unlike most working capital accounts, however, exchange settlement accounts cannot remain in deficit.  If, as a result of transactions cleared overnight, a bank loses more cash from its account than its closing balance on the previous day, it must restore its account to balance before the close of business;  it must not carry a negative balance in its exchange settlement account overnight.  This restoration can only be achieved by depositing cash with the RBA.  One bank in deficit can clear its position by "buying" excess exchange settlement balances from another bank.  When the banks as a group find their exchange settlement accounts in deficit, they must seek out an alternative source of cash.

In principle, commercial banks could hold large positive balances in their exchange settlement accounts to provide some insurance against having to search for alternative sources of cash.  In practice, banks in Australia do not hold significant exchange settlement balances.  The RBA pays no interest on these accounts, and it would be poor business practice for banks to carry large positive balances from day to day.  Banks with positive balances typically seek to withdraw their funds and invest them elsewhere in order to earn interest.

Besides exchange settlement account balances, the primary sources of cash include:  currency held by the banks and by the non-bank public, the rediscount facility offered by the RBA, certain RBA transactions and the authorised short-term money market.  The RBA's control of the various sources of exchange settlement funds constitutes its monopoly in the cash market, and confers upon it the power to manipulate the cash rate.

While fluctuations in the stock of currency have a large impact on the banks' exchange settlement account balances, they are not normally regarded as a ready source of cash with which to settle debts to the RBA.  To obtain currency from the non-bank public, banks would need to raise interest rates enough to attract currency deposits.  Experience suggests that the public's demand for currency is insufficiently interest-elastic to provide banks with a reliable response to what could be an expensive exercise.

Under the rediscount facility, the RBA undertakes to repurchase Treasury Notes from the market at an agreed rate of discount (the "rediscount rate").  This facility is available to all holders of Treasury Notes, and can be exercised at the holder's discretion.  Significantly, the RBA agrees to give the market immediate value (i.e. cash) for rediscounted securities.  In the case of a bank or an authorised dealer, this immediate value is credited to its exchange settlement account and can be used to settle debts on the same day.

While rediscountable securities are not themselves cash, they are "cashable", in that they can be exchanged for cash relatively quickly.  Since the RBA is effectively making cash available perfectly elastically through the rediscount facility, it sets the rediscount rate at a penalty to the market.  The rediscount facility is essentially a "safety valve", to relieve market pressure when other sources of cash become too expensive.

The arrangement under which the Reserve Bank grants immediate value to rediscounted Treasury Notes does not encompass all RBA transactions.  RBA purchases and sales of foreign exchange do not normally affect the exchange settlement accounts of banks until two days after the transaction (although "cash" transactions are possible).  Similarly, transactions in government securities between the RBA and commercial banks are cleared one day after the transaction.  Thus banks cannot use transactions in foreign exchange or government securities (other than rediscounts) to clear debts with each other or with the RBA.

The remaining alternative source of cash is the official short-term money market.  The authorised short-term money-market dealers act as a buffer between the RBA and the financial system, as well as making a market in Commonwealth Government securities.  As part of this buffering process, the RBA treats the dealers in a unique way:  it confers "same day" value on all their transactions.  Thus, when a dealer deposits a bank cheque into its exchange settlement account, the RBA credits it for immediate value (i.e. turns it into cash) even though the cheque is not cleared until the following day.  This convention plays an important role in the mechanics of monetary policy.  It is essentially a means via which the RBA "lends" cash to the dealers against the security of a bank cheque.

Since the dealers also hold exchange settlement accounts, banks can clear their accounts with the RBA by drawing cheques on callable deposits they hold with the dealers.  Unlike exchange settlement accounts held with the RBA, callable deposits with authorised dealers pay interest at a market-determined rate.  In fact, this rate of interest, which applies to deposits on 24-hour call with authorised dealers, is the "cash" rate.  Since it represents the opportunity cost to a bank with positive exchange settlement balances to invest overnight, it will also be the minimum rate demanded by such a bank in an interbank loan of exchange settlement funds.  The existence of interest-bearing callable deposits that are immediately convertible into cash explains why banks do not hold surplus liquid balances in non-interest-bearing exchange settlement accounts with the RBA,

When the banks as a group find themselves with excess exchange settlement balances, they simply deposit them with the dealers.  When they find themselves with deficit balances, they run down their deposits with the dealers.  This transfers the balancing process and the need for cash from the banks to the dealers.  As is the case with banks, the dealers can generate cash by rediscounting Treasury Notes.  The dealers have two additional means of generating cash which are not available to banks, however:  selling securities or repurchase agreements ⁽³³⁾ directly to the RBA;  and raising deposits from the non-bank public. ⁽³⁴⁾  The latter route exploits the convention under which the RBA confers "same-day" value on all transactions of authorised dealers.

The cash rate is determined in the official short-term money market as the dealers strive to clear negative balances in their exchange settlement accounts resulting from banks' withdrawal of funds from their callable deposits with dealers.  Any net demand for cash which remains after discretionary purchases of securities or "repos" by the RBA is cleared by the dealers raising deposits from the non-bank public, or rediscounting Treasury Notes, or both.  This process is summarised in Figures 1 and 2.

The net demand for cash by the authorised dealers is shown by the kinked schedule DEFG in both figures.  This demand consists of two components.  The first -- given by the horizontal distance OD -- represents the gross demand for (supply of) cash arising from the aggregate deficit (surplus) in dealers' exchange settlement accounts, minus the cash supplied (withdrawn) by the RBA through discretionary purchases (sales) of securities or "repos".  The second component -- given by the horizontal distance EF -- represents the demand for cash required to settle purchases of Treasury Notes still outstanding from the most recent tender. ⁽³⁵⁾

Figure 1:  An Opening Deficit for the Cash Market

The supply of cash is dictated by the supply of non-bank deposits to dealers.  This is shown by the schedule AA in Figures 1 and 2 to be an increasing function of the cash rate.  Once the cash rate rises above the rediscount rate on Treasury Notes, however, the supply of cash becomes highly elastic, provided the dealers hold an adequate stock of Treasury Notes (or are able to purchase them in the market).

The equilibrium cash rate in both figures is i*.  As can be seen from Figures 1 and 2, the impact of the rediscount facility is to put an effective ceiling on the daily cash rate at irediscount, whereas the unsettled Treasury Notes tend to put something of a floor under the rate at itender.

In conducting monetary policy, the RBA has several degrees of freedom.  Consistent sales of securities will ensure that the opening position of the market is consistently on the excess demand side (as in Figure 1).  By offering less accommodation, the RBA pushes the demand schedule further to the right, forcing dealers to bid more actively for funds from the non-banks.  Finally, by raising the rediscount rate, the RBA raises the ceiling on the cash rate.  Experience has shown the RBA has a high degree of success in manipulating the cash rate.  It is nonetheless true that, on each day, the demand for cash is met by supply.  Monetary policy operates not by denying supply but by manipulating the price at which supply is made available (Chart 2).

A sustained change in cash rates is transmitted in to the financial system in two ways.  First, since cash itself yields no interest, a sustained increase in the cash rate will induce banks and individuals to economise on their stock holdings of cash (as distinct from their flow demands for cash to settle their daily deficits).  This will have a contractionary effect on both spending and credit.

Secondly, and more importantly, since overnight loans provide the underlying finance for holdings of Treasury Notes and bank bills, sustained changes in the cash rate feed quickly into other short-term interest rates;  and, through them, into the financial system more generally.  This transmission process is accelerated when the RBA makes its intentions known to the market.

Control of the cash rate gives the RBA partial control of the yield curve and, through this, some influence over the general level of economic activity.

Figure 2:  An Opening Surplus for the Cash Market

REFORM OF OPERATING PROCEDURES

While current operating procedures for monetary policy are effective, they are also cumbersome.  In particular, there seems no good reason to retain the authorised dealers as separate institutions with special access to RBA liquidity support through the repurchase agreement mechanism.  One of the major functions of the dealers was to make a market in Commonwealth Government securities (CGS) in return for privileged access to central bank liquidity support and special status in the cash market.  The market for CGS is well established and does not need a group of specialist institutions to keep it alive.  Indeed, if recent trends in public finance continue (after the interlude of the current cyclical downturn), there may be no need for a market in CGS as the Commonwealth's demands for loanable funds steadily diminish.

Furthermore, alternative banking arrangements for the Commonwealth government might well see the Commonwealth no longer banking with the RBA. ⁽³⁶⁾  Under these circumstances, the Commonwealth's borrowing program would be conducted by private banks:  and, again, the role of the authorised dealers would be vestigial.

The dealers are certainly not necessary for the operation of monetary policy through the cash market.  While the analysis above pointed to the key role played by dealers' borrowing from the non-bank public in the transmission of monetary policy, this role could be filled equally well by banks.  An alternative set of institutional arrangements which dispensed with the authorised dealers would contain the following elements:

• authorised banks alone would hold exchange settlement accounts with the RBA;  there would be no authorised money-market dealers;
• banks would be obliged to hold reserves with the RBA in special accounts (like non-callable deposit or NCD accounts) on which interest would be paid (although this is not necessary);
• banks could hold "excess reserves" with the RBA at its discretion and these would bear interest at a rate determined by the RBA; ⁽³⁷⁾
• banks could overdraw their exchange-settlement accounts and pay the appropriate overdraft rate to the RBA.

Such a system is similar to that of the Federal Reserve System in the USA.  It exploits the central bank's monopoly in the cash market, as the existing system in Australia does, but would provide the RBA with more direct and immediate control over the cash rate.  Under such a system, the RBA would set the level or rate of growth of "non-borrowed reserves".  In other words, it would determine the base level or growth rate of its own liabilities which are available to be held in the form of currency and deposit balances with the RBA (whether exchange settlement account balances, NCD balances or balances in interest-bearing deposits).  Like the present system, this determines the base supply of cash which is available to banks and the general public for settling debt.  Again, like the present system, the level or rate of growth of this supply will determine the extent to which the banking system must borrow cash to meet the demands for cash which the general public places on it from time to time.

Unlike the present system, however, "borrowed reserves" would not be supplied by the roundabout method of forcing money-market dealers to borrow short-term deposits from non-banks:  deposits which are then converted into cash by the RBA.  The system is much more straightforward than the present procedure, and is analogous to the overdraft mechanism which most private banks operate.  Banks that need additional cash to clear their settlement obligations would simply borrow it from the RBA, via an overdraft on their exchange settlement accounts.  The RBA would set the interest rate on the overdraft, and this rate would rise commensurately with the amount borrowed.

The need to borrow cash on overdraft may be the result of temporary shortages which a particular bank experiences, and in that case the cash supply would be used just like a revolving credit line.  More often, however, the borrowing need would be the result of deliberate central bank policy.  Under the proposed alternative system, the RBA would tighten monetary policy by reducing the rate of growth of non-borrowed reserves, i.e., the base rate of growth of its balance sheet.  As cash becomes scarcer, banks find themselves running larger and larger overdrafts with the central bank through their exchange settlement accounts, and paying commensurately higher interest rates.  The higher rate of interest which they must pay for cash raises their cost of funds, and it forces them to charge higher rates of interest on loans and advances.  As bank loan rates rise, rates on alternative sources of loan funds rise as well.  So do deposit rates, both at banks and at non-banks.  In this way, the higher cash rate is transmitted throughout the financial system as rates on loans and deposits rise pari passu.

Rising interest rates eventually lower the demand for cash as the transactions demand falls and banks' balance sheets begin to grow more slowly, implying a reduced demand for cash reserves.  The end result is the same under the alternative system as under the present arrangements.  The main difference is that the RBA knows in advance how the cash rate will respond, as the unsatisfied demand for non-borrowed reserves spills over into the market for borrowed reserves.  In effect, the RBA fixes the supply curve for borrowed reserves, rather than (as now) relying upon a market-determined supply curve for non-bank deposits to dealers.  Given that the RBA would have no less influence over the demand for cash under the alternative system than it has under the present system, the addition of control on the supply side would make determining the cash rate a much more predictable task.  In particular, there would no longer be any need for the Treasury Note rediscount facility to remain, just in case of an unpredicted fall in the elasticity of supply of non-bank deposits to dealers.

The key to the operation of both systems is that the central bank can force banks to accommodate shortfalls in their reserves of cash by borrowing cash at interest.  Under present arrangements, cash is borrowed indirectly via the dealers to whom the RBA extends credit;  under the alternative system, cash would be borrowed directly from the RBA on overdraft at interest rates set by it.  In both cases, the RBA can determine the extent of borrowing which takes place, by selectively reducing the supply of non-borrowed reserves, or else by refusing to expand it in the face of demand pressure.  In one case, the resulting cash rate which clears the cash market is determined by market forces and may be too high or too low, adding to uncertainty and instability;  in the other, the rate is set by the RBA, which can raise or lower it administratively to achieve its desired impact on financial activity.  Moreover, the range of rates which the cash market can achieve is known in advance.

THE USE OF MONETARY POLICY

The analysis depicted in Figures 1 and 2 explains how the RBA is able to influence the level of the cash interest rate on a given day.  By repeating a particular operating strategy over a sequence of days, weeks or even months, the central bank can keep the cash rate high or low.  In so doing, it can influence the level of interest rates on financial instruments of longer-term maturity.  But there is a limit to the time over which the RBA can keep the cash rate at the level of its choosing.  By the same token, there is an upper limit on the maturity of securities whose rates will respond to central bank cash market operations;  for instance, throughout the recent episode of tight monetary policy in Australia, during which the cash rate rose above 18 per cent per annum, the yield on 10-year Commonwealth bonds remained around 10-11 percent per annum.  More recently, the yield on long-term bonds has begun to fall, as community expectations of inflation have been adjusted downwards.

The point is that, over the longer term, the RBA has less control over the price of cash than it has over the supply of cash.  On a daily basis, the demand for cash is stable enough (as a result of the various restrictions placed on banks) for the RBA to rely on a very close relationship between its manipulation of the supply of cash and the price of cash (i.e., the cash interest rate) which emerges in the money market.  This is not so over longer periods.  In the longer term, the demand for cash will respond to changes in the level of economic activity, the rate of inflation and expectations of these two variables.  Furthermore, over the very long term, there are likely to be structural changes in the financial system which affect the demand for cash, including the development of substitute forms of payment such as EFTPOS.

In the long term, the demand for cash cannot be expected to remain stable;  it will respond to the monetary conditions which the central bank establishes through the manipulation of its balance sheet.  Moreover, the induced changes in the demand for cash will tend to counteract the stimulatory or restrictive effects of changes in the supply of cash.  Another way to make the same point is to say that the long-run impact of monetary policy is neutral, i.e., non-existent.  Figure 3 illustrates the argument.

Figure 3

Figure 3 depicts the results of a sustained monetary tightening, i.e., reduction in the supply of cash.  In the short to medium term, such a reduction is effective in raising the cash interest rate from r' to r".  This is consistent with the intra-day analysis depicted in Figure 1.  But how long does the interest rate rise last?  The longer the cash rate remains high, the more likely will the future level of nominal economic activity begin to fall, both because real economic activity slows and because the price level begins to fall or to rise more slowly.  As expectations of future falls in nominal economic activity become widespread, and as those falls actually occur, the demand for cash to service the lower level of nominal transactions and to supply reserves for bank balance sheets (which are growing more slowly or contracting) begins to fall.  This is represented by the downward shift of the cash demand curve from DD to D'D' in Figure 3.

The falling demand for cash induces a reduction in the level of the cash interest rate.  Eventually, the cash rate falls to its original level (r' in Figure 3) consistent with the new lower level of nominal economic activity, a lower price level and the original supply of cash in real terms.  The balance sheet of the central bank is smaller in this new equilibrium but all real variables are unchanged:  i.e., the long-run impact of monetary tightening is neutral.  The same conclusion is reached in the case of monetary easing.

The point is that, in the long term, the only variable which the central bank will be sure to affect is the size of its own balance sheet:  in other words, the supply of cash to the financial system.  Its impact on the nominal cash interest rate will be transitory.  In particular, the central bank will not be able to maintain a particular level of the cash interest rate in the long term without repeated bursts of monetary contraction or expansion.  Interest rate targeting in the long term is ineffective and in the short term is probably pro-cyclical.

THE TROUBLE WITH INTEREST RATE TARGETS ...

Since deregulation in Australia, especially the floating of the $A exchange rate, the RBA has conducted monetary policy in a way which essentially targets the cash interest rate.  As explained above, the RBA has the power to manipulate the cash rate in the short run and can use this power to influence interest rates on securities of longer-term maturity.  There are two problems with using monetary policy to achieve a particular level of the cash rate over time.  The first is, as noted already, that the impact of changes in the supply of cash wears off, requiring a further round of tightening/loosening to maintain the effect.  The second is that, in making the subsequent changes in cash supply to maintain the effect on the level of the cash rate, one imparts a pro-cyclical bias to monetary policy.

Consider the example of monetary expansion.  An increase in the supply of cash, effected by leaving the cash market in surplus for a sustained period, leads to a lowering of the cash rate and other short- to medium-term interest rates.  Eventually, however (and this may be a matter of months rather than years), the demand for cash will rise:  as easier monetary conditions make people come to anticipate higher levels of nominal income and higher prices in future.  The higher demand for cash will tend to raise the cash rate towards its original level, requiring a further round of monetary expansion.

If the central bank is determined to keep interests rates down, it will be induced continually to expand its balance sheet and inflate the system.  Moreover, the monetary stimulus it delivers to keep rates down occurs in response to the effects of its earlier stimulus;  in other words, it reinforces the original stimulus or is pro-cyclical.  The same holds for attempts to keep the cash rate up:  as the cash rate drifts downwards with the induced reduction in the demand for cash, the central bank is propelled by its interest rate target to deliver a further round of tightening to raise the cash rate at the very time when the economy is already turning down.

The problem here is that the central bank's undisputed power over the cash rate in the short term leads it into a false sense of security concerning its ability to set the level of nominal interest rates over the long term.  Faced with the inevitable fading of its influence over that longer term, the central bank is induced to a further round of intervention.  The result, if it is trying to keep interest rates down, is chronic inflation;  and chronic deflation, if it is trying to keep interest rates up.  In either case, monetary policy targeted on the level of the cash rate will tend to produce cycles in interest rates which reinforce or amplify the underlying cycle in nominal economic activity. ⁽³⁸⁾

TARGET THE CASH BASE INSTEAD

Unlike the cash rate, which the central bank can set in the short run but not the long run, the size of its own balance sheet is a variable which the central bank can fix in both the short and the long run at any level of its choosing.  In order to set the cash rate, the central bank must influence both the demand for and the supply of cash.  While its influence over the demand for cash in the short run is complete, this dissipates with time as the market responds to central bank initiatives.  On the other hand, the central bank's balance sheet is entirely within its control over any period of time (assuming, of course, that the exchange rate and interest rates remain market-determined).

Rather than manipulate its balance sheet to achieve a particular level of the cash rate -- which, as explained above, leads inevitably to pro-cyclical intervention -- the RBA should set the rate of growth of its own balance sheet at a level roughly commensurate with the anticipated long-term growth of the demand for real cash balances.  Such a steadily expanding real supply of cash provides the solid base upon which economic activity can expand in a non-inflationary manner, and at a rate which is non artificially perturbed by pro-cyclical intervention.

The RBA does not have to guess correctly the long-term trend growth of real cash demand.  If it chooses too fast a rate at which to expand the cash supply, nominal cash demand will grow somewhat more quickly over time than real cash demand does, as the economy responds to a higher long-term rate of inflation.  Provided that the rate at which the cash supply expands is fixed and widely-known, the real economy will accommodate to the stable monetary environment.  Similarly, if the rate chosen by the central bank is too slow, there will be steady deflation accompanied by a growth of nominal cash demand which is slower than the growth of real cash demand.  Again, however, the real economy will accommodate to the stable monetary environment.

Proposals to target the supply of cash (or the "cash base" as it is formally known) are not new.  They are invariably rejected by central banks as naive, however.  This reflects in part a natural revulsion by an institution whose charter is explicitly interventionist, against proposals which would strip it of its interventionist power.  It also reflects a concern that setting the system to automatic pilot runs the grave danger of steering the ship onto rocks;  the central bank does not trust the automatic pilot, and cannot resist keeping its hands on the wheel.

One concern often raised in this context is that fixing the cash supply onto some long-term trajectory would leave the central bank powerless to intervene in the event of a sudden, unexpectedly surge in the demand for cash (say, as a result of a widespread run on financial institutions).  The bank would then be forced to stand idly by while the cash rate rose to astronomic heights and large numbers of financial institutions were forced into insolvency.  The spectre of such a scenario is often raised to strike terror into the hearts of would-be cash base targeters, and to help them see the reason of allowing the central bank some discretion in setting the cash rate.

The spectre is nothing but an illusion, however.  Surges in the demand for cash would be accommodated automatically, as described above, via the device of borrowing cash on overdraft from the central bank.  The supply of cash would be expanded temporarily to overcome a short-lived crisis in financial markets, without upsetting the long-run stability of monetary policy.  Besides, the level of the cash rate in such a circumstance would be predetermined by the central bank as part of its already announced overdraft policy.  The cash rate would fall, along with the cash base, as the demand for cash resumed its long-run trend after the crisis passed.  Similarly, a sudden fall in the demand for cash would be accommodated via repayments of borrowed reserves, thus contracting the central bank's balance sheet temporarily.

The stock of borrowed reserves of cash acts as the equivalent of "surge tank" in a water reticulation system, which absorbs temporary shocks and maintains pressure at a constant level.  The target rate of growth of the stock of non-borrowed reserves would be chosen so as to allow the stock of borrowed reserves to be big enough to serve effectively as a surge-inhibitor.

Deregulation has strengthened the hand of the RBA by shoring up its monopoly in the cash market.  As a result, the power of monetary policy is considerably enhanced.  Nobody who has lived through the last two years in Australia could be in any doubt about this!  With such a powerful weapon at its disposal, it is important that the central bank use it responsibly and effectively.

---

THE COLLAPSE OF CENTRAL BANKING

"But whatever may be our verdict as to the comparative outcome of the two systems in terms of stability it is unlikely that the choice can ever again become a practical one.  To the vast majority of people government interference in matters of banking has become so much an integral part of the accepted institutions that to suggest its abandonment is to invite ridicule."

-- Vera Smith, 1935

VERA SMITH gained her PhD under Friedrich Hayek at the London School of Economics in the early 1930s.  Her dissertation was published under the title The Rationale of Central Banking and the Free Banking Alternative, and it has recently been republished by Liberty Press.  The above passage is taken from her final chapter.

The first of the two systems to which she refers is the government monopoly system that has prevailed in Australia since 1910.  This system comprises a central bank which has either a total or residual monopoly over the issuance of bank notes.  Apart from this monopoly, the central bank will typically have control over the bulk of gold reserves for the entire banking system.  It will buy government debt, it will intervene on foreign exchange markets, and it will attempt to affect economic activity by setting interest rates, usually at the short end of the yield curve.  The Reserve Bank is now setting interest rates by operating on overnight cash rates.

Secondly, Vera Smith refers to the free banking system, the most famous example of which is the Scottish system which operated from the middle of the 18th century until 1845.  Australia also had a free banking system which was effectively shut down by the passage of the 1910 Commonwealth Bank Notes Act.  Regrettably there has not been the same interest in our own free banking history as in other free banking systems that have operated around the world.

The issue of central banking, i.e. government monopoly, versus free banking is now back on the agenda.  Vera Smith's lament of 57 years ago is no longer accurate.  The fundamental reason behind this change in intellectual fashion is the universal failure of central banks to provide sound money.  Some central banks, of course, have been more successful than others.  The Bundesbank has long been the great symbol of central bank success, and most contemporary political solutions for solving inflation and other symptoms of unsound money are proposals for instituting Bundesbanks in the local capital.  The Opposition is promising monetary reform on a Bundesbank policy, and Peter Reith has indicated that his first step will be to dispense with the services of Bernie Fraser, presumably because he does not speak German and does not look like Herr Poehl.  The NZ Reserve Bank is the contemporary antipodean representative of the Bundesbank.  It is worth noting that the record of the Bundesbank is only good by contemporary standards.  A currency half-life of 23 years, for the period 1975-89, is good by current Australian standards but not good when compared with the 19th-century gold standard experience.

As well as the inflation problem, which is now widely recognised, there is also a growing understanding of the powerfully destabilising effects of attempts by governments to use the monopoly powers of the central banks to achieve desired political objectives.  To quote the immortal words of Paul Keating:  "They do what I say.  You can depend upon it.  That you can put your money on." ⁽³⁹⁾

By "they" of course, he meant the Reserve Bank of Australia.  And in his momentous Placido Domingo speech, of nearly 12 months ago, he was even more frank.  "Leadership is about getting people into your pocket.  I have Treasury in my pocket, the Reserve Bank in my pocket."

THE PROBLEM OF THE RBA

To summarise the RBA record over the last 20 years, it is responsible for a currency with a half-life of less than 10 years.  In recent times we have had totally uncompetitive real interest rates, and a roller-coaster yield curve.  Exchange rates are now principally determined by relative real interest rates, subject to perceptions of political risk, and so Australia's exchange rate has been held substantially above the value which would result if export-import balances were the deciding factor.  This exchange rate situation has contributed, substantially, to our balance of payments difficulties, and has tightly squeezed those industries which sell predominantly in overseas markets.

Given this record it is astonishing, in my view, to continue to see a great deal of argument concerning the level of interest rates which the RBA should be setting at any given moment.  By and large the conservatives argue for higher interest rates, their opponents for lower interest rates.  The problem is that no-one knows what interest rates should be in order to achieve, say, a smooth landing, or 0-2 per cent inflation, or any other economic objective which might currently appeal.  One of the interesting things about this debate is that the moral issues involved -- massive transfers of income and capital from one community group to another, as a consequence of bureaucratic discretion -- are never discussed.

It is the Reserve Bank and its political masters, sui generis, which constitute the problem:  and they cannot, of their own volition, overcome themselves.  Because of the political framework in which these institutions operate, interest rate settings now have a positive feedback mechanism built into them, as described earlier.  There is no escape from this dilemma other than through major institutional change, and the acknowledgment, by political leaders, of the destabilising consequences of political risk.

Generally speaking, the failure of economic planning, price-fixing by government fiat, and the undesirability of rent-seeking, on moral as well as efficiency grounds, is now very widely acknowledged.  This is true with one vital exception.  Markets and contracts are now widely deemed to be greatly superior to planning, price-fixing, rent-seeking and regulation, in all areas of the economy except banking and monetary supply.

There is now a campaign under way to turn the label "economic rationalist" into a deadly insult.  An economic rationalist, as I understand the term, is one who argues that prosperity is a good thing, and that prosperity is most rapidly achieved by allowing markets (as opposed to government regulation or manipulation) to set prices, and by allowing people to trade with each other, both within and across national borders, on a basis of mutual advantage.  A key element in this mutual trade is agreement by the parties themselves, and not by government regulators, concerning the rents to be charged for the use of time, labour, capital and property, or concerning prices for commodities and products.  Rents, according to economic rationalists (if that term means anything at all), should, like other prices, be market-determined.

TABLE 1

Country	Currency Half-life (yrs) 1975-82	Country	Currency Half-life (yrs) 1983-89	Country	Currency Half-life (yrs) 1975-89
Switzerland	21.0	Singapore	76.9	Singapore	28.0
Singapore	19.3	Netherlands	59.2	Switzerland	24.0
Germany	17.0	Japan	57.1	Germany	22.7
Malaysia	15.1	Germany	45.6	Malaysia	20.3
India	13.9	Malaysia	42.3	Japan	18.4
Austria	13.7	Switzerland	33.9	Netherlands	18.3
Japan	12.3	Austria	28.5	Austria	17.1
Netherlands	12.1	Thailand	27.4	Thailand	12.5
Hong Kong	9.5	USA	22.2	USA	12.0
USA	9.3	Korea	21.2	India	10.9
Thailand	9.1	Canada	17.9	Canada	10.5
Canada	8.4	France	16.4	Hong Kong	10.0
Pakistan	8.0	UK	15.9	France	9.4
Sweden	7.9	Pakistan	13.3	Pakistan	9.3
Australia	7.4	Sweden	12.8	Sweden	9.1
France	7.4	Hong Kong	12.1	Australia	8.2
Philippines	7.2	Australia	10.8	UK	7.7
South Africa	6.5	Italy	10.3	Korea	7.2
UK	5.8	Indonesia	10.3	Indonesia	6.9
Indonesia	5.8	India	9.8	Italy	6.3
New Zealand	5.6	New Zealand	8.4	New Zealand	6.2
Italy	5.1	Philippines	5.9	Philippines	6.1
Korea	4.9	South Africa	5.8	South Africa	5.8

It is a fascinating thing that none of the leading economic rationalists in the media, in the Treasury, in academia and (with two exceptions) in the think-tanks has ever deviated from the position that economic irrationalism, using that term as the opposite of my definition above, should apply to money supply and banking.

Price controls, particularly rent controls, are rightly regarded by economic rationalists as totally destructive of a productive economy.  Yet government control of the rent paid for the use of capital -- i.e. interest rates, perhaps the most important price in the entire economy -- is taken for granted as being desirable, if not essential, by these luminaries.  Furthermore they have heated debates as to what that price, set by the RBA, should be.

MILTON FRIEDMAN AND MONETARY THEORY

Why should this vital part of economic life still be regarded as unique, and as requiring the continuance of statutorily guaranteed government monopoly and price control?  The answer, in two words, is Milton Friedman.  Friedman belongs to the quantity school, which traces its antecedents back to David Hume, and which includes Sir Robert Peel who as British Prime Minister brought down the 1844 Banking Act.  This Act confirmed the monopoly privileges of the Bank of England but at the same time turned the Bank into a currency board based on gold.

Chart 1:  Monetary Base Annual Growth Rate

Friedman proselytised over many years for a quantity monetary rule, and such a rule, of course, requires a central bank for enforcement.  At least it was assumed that a central bank could impose the sort of monetary rule which Friedman advocated.  Such was his prestige amongst market economists, and such was the prestige of the Chicago School generally, that critics of the quantity theory (who also tended to be free-banking advocates) were frozen out, not through any malevolent design by Friedman, but because his prestige and powers of argument were so great.

For example, in 1960 Friedman wrote:

"This analysis, then, leads to the conclusion that some external limit must be placed on the volume of currency in order to maintain its value.  Competition does not provide an effective limit, since the value of the promise to pay, if the currency is to remain fiduciary [a vital caveat -- RJW], must be kept higher than the cost of producing additional units.  The production of a fiduciary currency is, as it were, a technical monopoly, and hence, there is no such presumption in favour of the market as there is when competition is feasible." ⁽⁴⁰⁾

Friedman's dominance in monetary theory suffered a major blow in the mid 1980s when the substantial inflation he had been prophesying in 1983 (a forecast based on a blow-out in key monetary aggregates at the time) failed, absolutely, to materialise.  If the quantity theory on which his predictions had been based was no longer a useful or accurate way of looking at monetary theory then, in due course, the argument for central banks began to vanish.

I should note that Milton Friedman wrote the following for the dust jacket on the new edition of Vera Smith's classic.

"Vera Smith's book remains an authoritative and accurate summary of the theoretical arguments for and against free banking, a subject that has recently received renewed attention."

While not exactly a glowing tribute it is nonetheless a significant acknowledgment.

Friedman's latest position on the monetary problem, one picked up in the Hartley-Porter monograph, is that the monetary base should be frozen.  Under this rule the Federal Reserve does nothing other than replace worn-out banknotes and rolls over existing Certificates of Deposit.  What would happen if the very great proportion of the US banknote issue -- variously estimated at between 50 per cent and 80 per cent -- which is held overseas, came back to the US?  Presumably that would result in a significant inflation.

I should also note that quantity theory has been helpful when considering situations of extreme inflation.  The German experience of 1922, the American Civil War experience of 1863-4, the Latin American experience most of the time, are situations when quantity theory does help to understand the processes at work.  Contrariwise the experience of Britain, under the Bank of England currency board system of the later 19th century, cannot be understood within the Humean framework.

Adam Smith, of course, was deeply influenced by the Scottish banking experience in which he grew up and his criticism of Hume, subtly expressed, was as follows:

"A paper currency which (because of doubtful convertibility) falls below the value of gold and silver coin, does not thereby sink the value of those metals, or occasion equal quantities of them to exchange for a smaller quantity of goods of any other kind.  The proportion between the value of gold and silver and that of goods of any other kind, depends in all cases, not upon the nature and quantity of any particular paper money which may be current in any particular country, but on the richness or poverty of the mines, which happen at any particular time to supply the great market of the commercial world with those metals."

There we have perhaps the earliest statement of what we can call the supply-side theory of money.

The defenders of central banking theory, which is, ultimately, the defence of Humean quantity theory, are now in deep trouble.  Charles Goodhart, the most distinguished advocate for the central banks, is now reduced to three defences:  the risk of bank runs, the need for depositor protection, and the irresponsibility of banks with respect to their portfolios.

AUSTRALIA'S FINANCIAL DISASTERS

In Australia we have had a number of financial disasters in recent years, and critics of economic rationalism have attributed these scandals to financial deregulation.  Three in particular that I wish to consider are Pyramid, Tricontinental, and the Bank of South Australia.  Would these disasters have taken place under a free-banking regime?

The cases of Tricontinental and the State Bank of South Australia (SBSA) are fundamentally similar.  Both banks enjoyed the backing, either express or implied, of the Governments of Victoria and South Australia respectively.  Both the SBSA and the State Bank of Victoria (SBV) had been under the prudential supervision of the RBA, albeit on a voluntary basis rather than statutory requirement, since the early 1980s.  Both banks borrowed at high interest rates from lenders who were quite prepared to lend to institutions which they regarded, in the event quite justifiably, as being beyond bankruptcy.  Here we have an outstanding example of moral hazard.  The liabilities of these government-backed banks were high-interest borrowings.  Their assets were high-risk CBD real estate, whiz-bang, high-tech, no-income firms, etc., etc., and now the taxpayer is collecting the tab.  The moral here is plain.  Governments should not be in the banking industry.  They provide extreme situations of moral hazard and political risk.  The wonder is that both the SBV and the SBSA lasted as long as they did.  The reason is that whilst they confined their lending mainly to domestic real estate, only major recessions could cause them any difficulty.

What about the Commonwealth Bank?  Why has that institution not suffered from moral hazard?  For two reasons, I suggest.  First, the moral hazard at the Commonwealth level was concentrated in the Reserve Bank.  Second, the Commonwealth Bank increasingly saw itself as competing with the big three private banks and modelled itself on them.

Pyramid is a more interesting example.  It obtained a substantial share of the deposit market in Geelong, and subsequently beyond Geelong, by offering significantly higher interest rates than its competitors.  Its asset portfolio included domestic housing but also substantial chunks of what turned out to be grossly overpriced CBD real estate.  In a climate of an expanding monetary supply in the mid 1980s and of increasing asset values, both in the stock market and in the real-estate market, this policy proved highly successful, and Pyramid increased its balance sheet very rapidly during this period.  After October 1987 the balance sheet began to look much less robust.

The climate in which Pyramid expanded was that of the classic inflationary boom fuelled by substantial injections of new base money during the period 1983-1986.

The fate of Pyramid was similar to the fate of the building societies and land banks of the early 1890s.  But instead of enjoying the stimulus of locally printed money, the Victorian Government was (along with municipalities) during the late 1880s borrowing heavily in London, raising 18 million pounds over the three years 1888-90.  This borrowing sustained asset prices in Victoria despite a significant fall in export prices, and the 1880s boom was maintained a little longer.  The London money ceased to flow when Baring Bros crashed in 1890.

In the absence during the mid 1980s of a central bank which was creating very substantial quantities of high-powered money, the whole commercial climate would of course have been different and the rise and rise of Pyramid could not have occurred.  Of course, under a free banking regime it is possible for banks to go belly-up.  In the Scottish free banking period the oft-quoted example is the Ayr Bank, which collapsed in 1772 after an over-issue of notes.  In Australia the banking crash of Easter 1893 saw 13 major banks close their doors, and saw them use the Victorian legislation of 1892 to reconstruct.  The 1893 crash provided the political base for demagogues such as King O'Malley to whip up sentiment against the banks, and to introduce the Australia Bank Notes Act of 1910 and its successors.

I have had the opportunity to study Kevin Dowd's chapter on the Australian free banking system, and the 1893 crash in particular, which will be published soon in an omnibus volume on free banking around the world.  In my view he makes a convincing case that Government folly was, as so often is the case, a major element in the 1890s' crash.  I quote his conclusion.

"The bottom line, then, is that the bank failures were caused primarily by a combination of 'real' factors and misguided government intervention, and the 'failures' were not what they might appear to be anyway.  It is ironic that a crisis in which inept government intervention played such a major part should have become so widely regarded as a failure of free banking."

If I have a criticism of the Dowd thesis it is that he has not given sufficient attention to the asset-appreciation consequences of two decades of massive government borrowing (particularly in Victoria) for gold-plated railways which were never, ever, going to yield a commercial return.  There have been a number of books written about the 1880s and 1890s in Victoria:  David Cannon's The Land Boomers, for example.  None of them, to my knowledge, has examined the 1890s' crash in the light of two decades of government profligacy and rent-seeking, both in terms of where the next railway line was to be built, and in terms of tariffs.  There is a great book, I think, waiting to be written.

There do seem to me to be some lessons which we can learn from the 1890s and Pyramid experiences.  One lesson is that there may well be certain rules which should apply to deposit-creating institutions, such as banks and building societies, which do not apply to normal corporations.  For example, I suspect that all deposit-creating institutions should be required to issue their own notes, and that the privilege of limited liability on the part of shareholders should not apply to the note issue.  Or at least that shareholders should be liable to twice their shareholding, or some such ratio.  Such provisions would certainly concentrate the minds of shareholders and their directors wonderfully.  Alternatively one might establish a moral-hazard situation in which depositors assumed that such arrangements would enable them to assume that all deposit-taking and note-issuing institutions were soundly managed.  This is an area of analysis which merits further debate.

A FREE BANKING REGIME

The essence of a free banking regime is:

• No central bank, no government monopoly issue of bank-notes, no capacity to monetise government debt, no lender of last resort.
• Private banking firms compete for market share not only of deposits, but for share of the public holding of bank notes.  The law concerning deposit creating and note issuing companies might, with advantage, be different from ordinary company law.  But it should not restrict freedom of entry, either from domestic or from foreign entrants, and it should be designed to minimise the problems of moral hazard.
• Competition means banks will have to promise to redeem their notes with outside money of some kind, e.g. Amex traveller's cheques, US dollars, reputable government debt, gold.
• Government's role in a free banking system is defining the unit of account, and upholding the rule of law.

Unless the unit of account, i.e. the Australian dollar, is defined in terms of some major outside currency such as the US dollar, then a free banking regime must have a floating exchange rate.  Contrariwise, a currency board is necessarily tied to a foreign currency and a fixed exchange rate or, as with Singapore, a CPI-adjusted exchange rate.  Interest rates are then, likewise, tied through arbitrage to the metropolitan interest rates.

The Australian dollar could be defined, by statute, in terms of the US dollar.  It could be defined, by statute, in terms of a CPI-adjusted US dollar.  It could be defined in terms of a quantity of gold, or in terms of a basket of commodities.  David Glasner and George A. Selgin are now arguing for a dollar defined in terms of labour productivity.  This would require the establishment of a productivity index.

Without getting deeply involved at this stage in the definition of the unit of account, i.e. of the Australian dollar, let us look at how a free banking regime would work in Australia.  It would be very similar to the way it worked before 1910.  The unit of account was then defined in terms of gold.  One pound was defined as 113 grains of fine gold.  Banks issued their own notes which were redeemable on demand with gold sovereigns, and they maintained substantial reserves of gold for this purpose.  A clearing-house system was established in Melbourne in 1867, under which the banks exchanged notes with each other at the end of the trading day.  In Sydney bilateral clearing was the norm until 1895.  Any excess would have to be redeemed with outside money, almost always in gold.  The rapid reflux mechanism provided by this clearing system made note over-issue more or less impossible.

Let us assume for convenience that the Australian dollar was defined by statute, in 1993, in terms of an inflation adjusted US dollar.  Thus the $A would initially be set equal, say, to US 70 cents, and the conversion rate would be automatically adjusted quarterly in accordance with the US CPI statistics.  Whilst the banks would not be required to redeem their notes in US dollars, I believe competition would require them to do so.  Circumstances might change and gold might be more attractive to customers.  Banks might also compete by offering redemption packages involving a time delay.  But because of the statutory definition of the $A, the redemption rate would be automatically adjusted and so bank reserves of US dollars for redemption purposes would be adjusted in the light of US CPI movements, and views of the likelihood of demand for redemption.

The existing gold holdings of the RBA should be returned to the private banks which were forced to yield them up first during the Great War, and subsequently during the Depression.  There should be no prudential supervision, except that provided by the private sector such as Moodys or S&P. Entry into the banking industry should not be subject to political discretion;  any firm which wishes to establish itself as a reliable issuer of deposits and notes should be free to do so, provided it can meet the costs of entry, and is prepared to accept the law relating to deposit-creating institutions.  The current 10 per cent shareholding limit for banks is a major source of moral hazard, since it renders banks immune from takeovers except by other banks.

Interest rates would be determined by the market, and the major factor affecting these rates would be the definition of the unit of account as set out by statute.  If the $A were defined as a fixed number of US cents, then our interest rates would mirror US interest rates.  If our dollar were defined in terms of an inflation-adjusted $US, then our interest rates would be down several percentage points on the US rate.  And going the final step, if our dollar were defined in terms of a productivity-indexed $US, then our interest rates would come down yet again.  These comments are only valid if the monetary regime which established all these desirable things was perceived by the markets to be immune from political risk.

The study of contemporary monetary institutions is fundamentally a study of political risk.  Governments have done well from their monopoly of money supply and their ability to monetise their debts through inflation.  Why, then, should an Australian government give up its monopoly over the note issue and the creation of high-powered money;  give up the power and patronage which the levers of monetary policy bestow upon a Federal Treasurer?

LESSONS FOR AUSTRALIA

Bernie Fraser is, in my view, quite right in arguing that a central bank can never be free of political influence.  The central bank exists because of its statutory monopoly position.  That statutory monopoly is always at the discretion of the Parliament.  The Bundesbank has been overridden recently on two major policy issues.  How can it be otherwise?  Prime Ministers, Chancellors and Presidents are elected.  Governors are appointed.  That is the reality and Bernie Fraser is, above all, a realist.  Peter Reith cannot escape from that reality either.

So far, of course, the Opposition is not thinking along these lines.  There are, however, major political gains waiting for a government which provides its citizens with the world's best money.  Furthermore, unlike labour market reform, there are no significant vested interests which stand to lose rents as a consequence of an end to the present regime.  There is of course, the RBA itself and its 800 or so employees, together with the money market dealers who make a living from the current monetary instability.  I would not have thought they constituted a major interest group.  The banks themselves would not be too keen on the end of their protection from takeover, but politically they have taken a fair bit of stick lately, and are not in a position to protest at being asked to do what they did as a matter of course in the last century.

Currently Singapore, with its inflation-adjusted, $US-based currency board, holds the Number One position on the price stability ladder.

If Australia were to move to a Singapore-type currency board, there is no doubt that it would be a significant improvement on what we now have.  We would still have, nevertheless, a government monopoly of the note issue, and government appointments to the currency board.  A currency board ties up, unnecessarily, outside money as reserves.  That is not really a great price to pay for the improvements that would follow.  It is the element of government monopoly which is fundamentally the source of political risk for a currency board, and already in Hong Kong there are moves to break down the rules in favour of flexibility and discretion.

The present Opposition policy does not address the central issue of political risk at all, and I suspect that shadow ministers believe that as soon as they occupy the government benches the problem of political risk will vanish.  If that is so then we have a lot of education to carry out.  But a government which does understand the costs of political risk, and accordingly confines its role to defining the unit of account, and is able to entrench legislation which persuades the markets that that alone will be its role, will reap enormous economic gains for the country and, justifiably, substantial political kudos for itself.

This is the great political prize.  If New Zealand had adopted a currency board back in 1985, then much of the pain which accompanied the Roger Douglas reforms would have been avoided.  It is worth noting that in the NZ constitutional system, which is that of a unicameral house and no written constitution -- Westminster without the House of Lords, in effect -- political risk is deeply embedded in the structure of society.  A simple majority in the one chamber parliament can do anything it wants.  This was the basis of Muldoon's power and it enabled him to do great damage.

CONCLUSION

• Statutorily based monopoly privilege bestowed upon a central bank brings with it unavoidable political risk.  That risk is built into long term interest rates, and exchange rate expectations.  Because of political risk, interest rate manipulation by a central bank is necessarily accompanied by positive feedback effects which lead to economic instability.  Because it is both theoretically and practically impossible to use interest rate control to obtain a desired economic outcome (assuming agreement on a desired outcome) we have a "random walk" monetary régime, in which the incentives for the central bankers are directed towards surprising the markets.
• This instability can only be remedied through institutional change.  A halfway house is replacement of the central bank with a currency board.  There is ample evidence which shows that such a move would be a major improvement on our present situation.  It necessitates a fixed exchange rate, although it could be a CPI-adjusted rate.
• The full reform package is a return to free banking, with the government's role confined to that of defining the dollar, and entrenching that definition in statute.  How best to define the $A and how best to entrench that definition is a subject requiring a full day's seminar.  Perhaps I could handle that topic next year.

---

SHOULD AUSTRALIA HAVE A CURRENCY BOARD?

CURRENCY BOARDS were ubiquitous in the British colonial regimes of Africa, Asia, the Caribbean and the Middle East, where over seventy once operated.  Boards were also employed in other contexts, for example the free city of Danzig and the government of North Russia used them.  Today, currency boards survive in such small countries as Singapore, Brunei and Hong Kong.

The main characteristic of the currency board system is that the board stands ready to exchange domestic currency for the foreign reserve currency at a specified and fixed rate.  All boards maintained that fixed-rate convertibility, even in the most trying times.  That unblemished record obtained because currency boards are required to hold realisable financial assets in the reserve currency at least equal to the value of the domestic currency outstanding.  Hence in the currency board system there can be no fiduciary issue.  The backing to the currency must be at least 100 per cent.  Although in principle it is the currency board that is required to convert on demand all offers of domestic or reserve currency, in practice, where there is a banking system, however elementary, it is the banks that have carried out most of the exchange business.  The buying and selling rates for both currencies have a sufficient spread so that the costs of exchanges are covered.  The convertibility of currencies and the 100 per cent reserve-currency backing requirement in the currency board system do not extend to bank deposits or any other financial assets.  If a person has a bank deposit and wished to use the currency board to convert it to foreign currency then the deposit must be first converted into domestic currency and then presented to the currency board.

The disciplines of convertibility and the avoidance of deficit financing were characteristic of much of 19th century Britain.  The principle of the currency board was enshrined in the provisions of the (British) Bank Charter Act of 1844.  The Issue Department of the Bank of England was to act like a currency board.  (The Banking Department of the Bank of England, however, was not required to act as a currency board because it was not subject to reserve requirements).  It is not surprising that this principle was considered proper for the newly acquired colonies, and that the first currency board was established in Mauritius only five years after the Bank Charter Act of 1844.  At first, settlers and officials used the notes and coin of the imperial power as a normal extension of imperial trade.  The metropolitan currency and coin, since it was widely accepted and considered "as good as gold", served as a stable means of exchange and as a store of value in those largely inflation-free days of colonial occupation.

There were disadvantages of circulating the metropolitan currency, for example sterling notes and gold sovereigns, in the colonies.  First, there was a high risk of destruction or loss.  Second, real resources were locked into the circulating media and produced no return.  Any loss of currency notes would be to the benefit of the issuer (e.g. the Bank of England) and the colony would correspondingly lose the real value of the notes.  The institution of a currency board enabled the colony to avoid such loss.  The Bank of England note could be stored in the currency boards vaults and local currency issued to the same value.  Thus the accidental loss of a domestic note would not diminish the net assets of the colony.  In addition, the currency board would find it efficient to replace worn notes from its stock without having its assets tied up in sending battered Bank of England notes back to London for reissue.

In practice the currency board did not need to hold all its reserves in Bank of England notes.  It could buy interest-bearing sterling financial assets of suitable liquidity.  Provided these assets could be converted at sufficiently short notice without significant loss into bank notes (or provided the currency board could borrow notes on such security), the principles of convertibility, 100 per cent reserves and no fiduciary issue were satisfied.  This more sophisticated currency board system could be used to earn at least some of the profits of seignorage for the benefit of the colony.

In most colonies currency boards were monopoly note issuers.  However, parallel competing currencies did circulate in some cases:  a few colonies continued to circulate foreign notes and coins;  in the British Caribbean colonies, currency boards and banks both issued notes;  and the North Russian currency board's notes competed with notes issued by other Russian governments.  The non-colonial countries of Liberia and Panama, however, have used the United States dollar as a circulating medium.  In the case of Liberia it was argued that there were doubts whether the people would have confidence in a currency board supervised by the government of Liberia.  In particular, there were fears -- alas not groundless -- that the monetary system would be used improperly to finance government spending.

THE NORTH RUSSIAN EXAMPLE

The currency board in North Russia has hitherto gone unnoticed, and merits special attention because archival evidence has recently brought to light the fact that the National Emission Caisse (North Russia) was the creation of John Maynard Keynes.  In the summer of 1918, British and other Allied forces became entangled in the Russian civil war.  The Allies supported the White (anti-Bolshevik) government of North Russia.  Currencies issued by the many parties to the civil war were circulating in North Russia;  they were all depreciating and distrusted by the local population.  To facilitate Allied payment for local goods and services, Keynes, then a Treasury official with responsibility for war finance, designed a currency board scheme in August of 1918.  The North Russian government accepted Keynes' scheme and established the National Emission Caisse in Archangel on 28 November 1918.

The Caisse issued notes for one Pound to 500 Rubles.  It exchanged its rubles for pounds sterling at a fixed rate of 40 rubles per one Pound by issuing cheques on banks abroad (mainly in London).  The Caisse's note issue was backed with a sterling reserve equal to 75 per cent of the issue, provided through a loan from Britain to the North Russian government.  It was kept in the form of an interest-bearing deposit at the Bank of England.  The remaining 25 per cent of the Caisse's note issue was covered by bonds issued by the North Russian government.  This exception to currency board orthodoxy was most probably based on Keynes' experience with the Indian sterling exchange standard, which he described in Indian Currency and Finance (1913).

To provide the Caisse's sterling reserves, the British government bought 100 million ruble of notes.  The notes entered circulation at Archangel and Murmansk by British military payments to the local populace.  The Caisse's note circulation, excluding notes held in reserve by the British military, went from zero in November 1918 to a peak of 52.7 million rubles in July 1919.

In the spring of 1919, the Allies decided to withdraw their troops from North Russia.  By 27 September, the withdrawal was complete.  The Caisse officially closed to the public in Archangel on 4 October 1919.  It continued to redeem notes collected by the North Russian government and the State Bank of Northern Russia until 15 October.  The Caisse then moved to London.  Its main business there was redemption of the 55 million ruble credit that British military held.  About 13.5 million rubles remained in the hands of the public.  British troops returning from North Russia held a small amount of rubles, but most rubles were still in Russia.

North Russia fell to the Bolsheviks in February 1920, making worthless the North Russian government bonds that the Caisse held.  The Emission Caisse remained open in London until 30 April 1920.  After that date, note redemption ceased.  Most of the 13.5 million rubles in the hands of the public never were redeemed, inflicting a loss on their holders.  Britain, therefore, lost about 15.5 million rubles (378,500 Pounds), the difference between the Caisse's worthless North Russian bonds and the notes that were never redeemed.

The North Russian currency issue scheme was on the whole quite successful.  The currency was always convertible and never deviated from its fixed exchange rate with sterling.  In contrast to the nearly 2,000 different types of fiat currencies issued by other Russian governments at the time, it was a reliable store of value.  In consequence, the Caisse's rubles tended to drive the others out of circulation in North Russia.  With the North Russian currency, Allied forces were able to buy and sell goods almost as easily as if they had been at home on manoeuvres.

During its short life, the Caisse lost money because of the high cost of printing notes quickly and because its North Russian government bonds became worthless.  As matters turned out, the North Russian bonds became a British "gift", because Britain made good on the losses the Caisse incurred from North Russian government bonds.  The Caisse could have avoided that loss by keeping 100 per cent reserves in external assets, as most other currency boards did.  As colonies became independent states in the 1950s and 1960s they generally eschewed the currency board system and formed Central Banks to manage their currencies, ostensibly for "development" purposes.  The central banks, as distinct from the currency boards that they replaced, required the commercial banks to hold reserves as deposits at the Central Bank.  And the government could create money and finance government deficits by borrowing from the newly created Central Bank.  Some countries, such as Singapore, continued to operate a modified currency board system.  And Hong Kong, after experimenting with an unpegged currency from 1974, returned to a currency board system based on the United States dollar in October 1983.

The financial experience of countries which departed from currency board systems has not been auspicious.  Increasing inflation, generated largely by deficit financing through Central Bank credit and note issue, has been characteristic of most of the two or three decades since independence.  The objective of promoting growth and development has not been generally achieved;  indeed, in Africa the experience and the forecast is one of degeneration.  It is difficult to avoid the conclusion that the financial instability brought in train by the abrogation of the currency board system has played a considerable role in this process.  Until the collapse of communism in Eastern Europe and the Soviet Union, it appeared that, whatever the arguments in favour of the currency board system, there was little interest in reinstating that system.  Today, however, many formerly communist nations, in their search for stable, convertible currencies, are considering the currency board system.

CRITICISMS OF CURRENCY BOARDS

One of the main criticisms of the currency board system was that it provides a stultifying monetary constraint on development and inhibits growth.  But this criterion ignores changes in the currency-deposit ratio of the public.  The modern process of financial innovation economises on cash.  The use of the cheque rather than cash is the predominant financial trend in all countries.  For any given quantity of currency and other bank reserves the choice of the public for a larger ratio of deposits to currency has provided the main impetus for an expansion of the money supply (M1) in currency board systems.  The stability and confidence generated by the currency board system undoubtedly much encouraged the use of deposits.  It is perhaps ironic that the countries that have retained their currency board arrangements, Singapore and Hong Kong, have been the highest growth economies in the oil-importing Third World.  Their money supply has expanded partly through current balance surpluses and capital imports, but mainly through the increased use of deposits associated with the financial stability of the currency board system.  Both Singapore and Hong Kong inflated during the late 1960s and 1970s at roughly the same relatively low rates as the currencies on which they were based.  Thus they avoided the excessive inflation that affected many Third World countries which had adopted more "advanced" systems of central bank finance.

Another important criticism was thought to be not only the preclusion of counter-cyclical policy, but also the promotion of an actually pro-cyclical policy.  Many currency board colonies produced export crops which were sold in markets with widely fluctuating prices.  The collapse of commodity prices in a world recession generally gave rise to a large deficit on the current balance and so induced the currency board to contract the money supply.  The isomorphic case of a boom in export prices was thought to be less onerous.  Under a liberal trade regime the increase in foreign exchange receipts from exports could be offset in part by an expansion of imports which would damp down the inflationary pressure generated by the rise in export prices.  Although no monetary manipulation can turn a one-commodity export economy into a nicely diversified recession-proof system, there was no reason, apart from extraneous regulations, why the authorities as well as firms and persons could not hold or transfer foreign assets in any form and any currency is the sine qua non of the financial system.  And, for many decades, the government of Hong Kong has held a large portfolio of foreign financial assets which can be used to expand or contract the money supply and hence influence the currency board issue.

Although the currency board system did not have all the virtues or faults which were attributed to it, it did have some singular advantages.  To some extent it depoliticised the monetary system and insulated the public purse from plundering politicians.  There was no resort to the printing press to reward political allies or ruin one's opponents.  It gave a real credibility to the fixed exchange rate so that people willingly held both currency and deposits knowing that they would maintain their value.  Similarly it precluded the possibility that the exchange rate would be used to attempt to solve political and social problems.  These constraints, once thought to be vices, are now widely regarded as virtues.  The evident failure of trying to promote growth or equality by inflationary finance may create a new respect for currency boards.  The return of Hong Kong in 1983 to a full currency board system based on the US dollar was in response to political uncertainties as well as the realisation that such financial stability was sorely needed.

HONG KONG

The story of Hong Kong's Currency Board is an interesting one.  It begins in London in September 1983, when John Greenwood of G.T. said that the Hong Kong dollar was in great trouble financially and politically.  Over the past week the Hong Kong dollar had gone down 40 per cent.  It was agreed that it would be best for Hong Kong to return to a Currency Board, which it had been off since 1972.

Sir Alan Walters was in Washington at the time, and attending the IMF/World Bank meetings.  He tried to see Nigel Lawson (who had just been appointed Chancellor of the Exchequer) and Robin Leigh-Pemberton, the Governor of the Bank of England.  But he couldn't get hold of them.  So he waited till 2:30 in the morning and phoned Mrs Thatcher.  He said that they had to bring back the Currency Board system.  They went on arguing about this for, I think, about three-quarters of an hour.  She said, "I'm convinced of your case.  I'm coming to Washington on Tuesday and we'll have a meeting to decide the issue."

They sent off immediately two civil servants to Hong Kong and met on the Tuesday in the Washington Embassy.  The decision was then made to return to a Currency Board.  He'd argued for a mixed bag of currencies on which to fix but he conceded the argument that the US dollar was a very good, indeed the best, base.

De facto that Currency Board was in operation before the end of the week.  De jure it was in operation by the middle of the following week.

Even the journals who were most critical of Thatcher policies were constrained six months after the event to say, in the words of The Economist, "This was a masterstroke."  The Financial Times said it was an "unalloyed success."  The goods flowed back into the shops, and the capital flowed back.  Instead of people packing their little bags with greenbacks, and taking them out, the flow was reversed.  It was a remarkable event.  Subsequently, I think everyone would agree that the Currency Board has been the ideal solution for Hong Kong.

Nigel Lawson said to him afterwards, "How can you promote this -- fixing the Hong Kong dollar when you're absolutely against pegging sterling?"  Very easily.  There are two viable systems.  One that institutionally fixes the currency absolutely to a commodity or another currency.  This is a truly fixed exchange rate.  The other is a free or floating exchange rate.  In the middle is a mishmash of pegged rates.  There's a kinship between a fixed exchange rate and a completely free-floating exchange rate.  They are much closer together, in principle and in operational matters, than anything in between, such as a pegged currency or one involving discretion.  In both these extreme cases, the cases of fixed rates and of floating rates, you deal very easily with capital flows.  But in the middle you are hammered by capital flows --just as they are now being hammered to death in the European Monetary System.

Both fixed and floating systems, of course, are basically trying to eliminate discretionary power and political control of the system.  The middle ground is where you get all the politicisation of monetary policy and exchange-rate policy.

It is important to know what the Currency Board system does and doesn't do.  And it's important not to blame it for the many things it has been blamed for.  Currency Boards have been blamed for colonial exploitation and all sorts of other things of that kind:  on the contrary, they are nothing to do with colonial exploitation at all.  In fact, I've had many countries say "We'd like to be exploited like Hong Kong, please exploit us first if you would!"  It is also nothing to do with the idea that somehow it's going to inhibit development-type monetary policies, which is one of the main reasons why Currency Boards were withdrawn from the many colonial regimes in which they existed around the world.

It allows enormous flexibility.  It allows, for instance, Hong Kong to be, I suppose, still third among the major financial centres of the world.  It's remarkable, isn't it?  With complete freedom of capital movement, in and out, and so on.  Yet it has a Currency Board system.  I'd go further -- I'd say because it has the Currency Board system it has this immense freedom of capital flow.  I think it is remarkable how well the monetary system of Hong Kong has worked.  It has been tampered with, it's not as pure a Currency Board as I would wish -- but it is still substantially a Currency Board.

The Hong Kong Currency Board has served Hong Kong over these most difficult periods of negotiating over 1997, and you remember that this has been operating since October 1983, involving the political wash back and forth.  What sort of Hong Kong would it be without it?  The Board has worked well and, of course, allowed a very considerable repatriation of funds.  Basically, this is what Sir Alan Walters wanted to do, with John Greenwood.  They felt it was a moral obligation to the people of Hong Kong, to enable them to sell whatever they had in Hong Kong, go and swap their Hong Kong dollars for greenbacks, put them in a case and fly out of Hong Kong.  They wanted the freedom for a flight of capital.  That was the moral thing to do.  It was a moral issue as well as an economic one -- but most things that are economically right in my view are also morally right.  He emphasised this to Mrs Thatcher, and she said, "Yes, we must, we've done them down in so many ways, we must give them this."  It's important to bear the moral dimension of this in mind.

There is a lot of misapprehension involved here.  This is particularly so in the debate about monetary unions, which is going on in Europe to a great extent.  You will find it if you read The Financial Times, The Economist, and I think many other journals (but they're supposed to be the best -- they're called the quality press!).  Samuel Brittan, for example, says "If we fix it to the Deutschemark, then we'll have the same inflation as the Deutschemark, as Germany."  This is, of course, rubbish.  And the reason is this:  you're not fixing inflation.  All you're fixing (this is under free conditions -- no tariffs, etc., none of the nasties, and assuming the fraction of transport costs to price does not vary) is the prices of tradeable goods.  Now that is not to say that inflation will be the same, at all.  You might say, ultimately, inflation will be the same in the long run.  And I say, yes, in the long run, we are all dead.  But you can have quite big differences.  This is important to note.

The interesting thing is that in the Dutch draft at Maastricht -- I think we've probably not seen it around here, but we ought to pay more attention than we do because it's going to affect us as much as those poor souls in Europe -- it says that none of the member countries will have an inflation rate which deviates by more than one-and-a-half per cent from the average inflation rate.  Well, let's go through the arithmetic and see what happens in practice.  As I said, it is nothing to do with the rate of inflation per se.  What it does affect is the price of tradeable goods.  The inflation rate of tradeable goods may be very different from the inflation rate of non-tradeable goods, as you may have noticed in our recent house boom here.  Houses are not tradeable;  haircuts are not tradeable.  Steel is tradeable, and all those things.  Let me illustrate this by reference to Japan.  Japan is an interesting example.  It had a fixed exchange rate with the dollar from 1947 through to 1972 of 360.  (Incidentally, why did they choose 360?  The most distinguished Japanese Finance Minister told me.  Well, it was very simple.  The sign for Yen in Japan is a circle, and there are 360 degrees in a circle, and that's how there were 360 Yen in a dollar!)

Chart 1:  The Law of One Price -- Japan

Japan made enormous productivity gains in tradeable goods but virtually none in non-tradeable goods.  The prices of tradeable goods (given no tariffs and all those sorts of things which interfere, but not substantially, with those prices) had to remain the same as those in the United States.  So the inflation rate of tradeable goods, roughly speaking, in the United States and in Japan had to be the same.  The Japanese inflation rate was rather more than twice the inflation rate of Japanese tradeable goods.

(Of course, if productivity in tradeable goods is considerably less than your partners and the inflation rate of your partners is 3 per cent, you may have to have a deflation.  Because you've got to keep the price of tradeable goods the same -- that's the lynchpin of the system.)

Since the Currency Board was fixed at the end of 1983, Hong Kong's CPI has risen about twice as much as the US CPI but tradeable goods' prices again increased at the same rate (Chart 2).

The next point is, if you go onto a Currency Board, it is enormously important to marry the appropriate wife.  (Don't get me wrong -- divorces happen, but they can be expensive.)  There are a number of lessons -- a number of happy outcomes and a number of unhappy outcomes.  Probably one of the unhappiest outcomes was that of Chile.  In 1979, it did not actually adopt a Currency Board, but it was heading that way.  It fixed the peso at 39 to the US dollar.  This was about the worst time to fix to the dollar.  From 1979, the US had a tremendous monetary squeeze.  The dollar soared until February 1985.

Because Chile was fixed to the dollar, it was caught with a massive real appreciation of the peso.  And that appreciation of the peso caused difficulties.  First, with high rates of return, an enormous influx of capital, and it was marked in the books at a fixed rate of 39.  And then its export markets were devastated by this rise in the dollar.  Eventually, in 1982, Chile came formally off the peg.  For two, or nearly three, years after that there was a big mess.  Banks were bankrupt, and those which had previously been privatised had to be re-nationalised.

Chart 2:  The Law of One Price -- Hong Kong

After Chile freed the peso the economy began to recover.  But with all this debt denominated in US dollars, banks were broke, and a large fraction of the economy was bankrupt.  What they did, in fact, was to restructure virtually all debt.  Income declined steeply.  It was a very painful process.

A CURRENCY BOARD FOR AUSTRALIA?

I am not entirely certain that a Currency Board would be a good thing for Australia.  I think it would.  But I can review some of the advantages and disadvantages.

Suppose that Australia pegged to the Yen?  Obviously, you are going to have some difficulty if Japan pursues, let's say, a monetary policy which delivers an inflation rate of three per cent overall.  In traded goods, it's going to have a decline in prices.  This is because productivity in traded goods (manufactures) in Japan increases much more rapidly than productivity in non-traded goods.  If you had a deviation between your productivity gains in tradeable goods and the Japanese productivity gains -- a deviation of, say, a fairly modest four per cent, which is probably historically about right -- then you would have grave difficulty.  You would have to have a decline in the overall level of prices in Australia, if the Japanese were to have a three per cent inflation rate.  This, by the way, is true under any sort of fixed exchange-rate system, whether it is a monetary board, a currency board or what have you.  A fixed exchange rate merely equalises the level of tradeable goods prices in both countries.  The relative rates of equilibrium are determined by the relative productivity gains in tradeable versus non-tradeable goods and services.

MONETARY BASE CONTROL

I think the alternative of a free-floating exchange rate, which has been suggested, is the monetary base control.  In Britain, at the Prime Minister's office, they tried to introduce a monetary base control in Britain in 1980-81.  Sir Alan Walters had been for some time quite convinced that this would be a much better system of monetary control than giving the Bank of England very considerable discretion.  Monetary base control basically says, "Over the long haul we know that something like a three per cent expansion in Britain's monetary base is associated with about two per cent inflation.  So let's have a three per cent expansion in the monetary base."  Now, the simple way to do this is to say, "Well, this requires the injection, say, of 100 million pounds a year."  With 250 working days in the year, divide 100 million pounds by 250, and dump that many reserves every morning.  Let interest rates be what the market makes them.  Interest rates set themselves;  they are depoliticised.

Now, you must have an override.  Basically the override is needed when you have a liquidity crisis and people are demanding money.  Then the only thing to do is stuff greenbacks down their throats.  People only want greenbacks when they think they can't get them.  When they think they can get them they don't want them.

Sir Walters was convinced that technically everything could be settled right.  There are, of course, velocity problems but I don't think they matter very much.

As far as I know, you cannot produce evidence of any substantial, sustained inflation except when you have this increase in the rate of growth in the monetary base.  There are arguments about causality there, but they can be dealt with in the usual way.

So Sir Walters was convinced monetary base control was right.  I think if he had put his weight behind it, Mrs Thatcher would have gone along with it.  But he decided not to.  He was a chicken, and the reason is that, to put that into operation, it had to be done by the existing staff of the Bank of England.  And I'm perfectly certain they would have had great difficulty in implementing monetary base control.  If you've said it wouldn't work and you've got an opportunity to make sure it wouldn't work ... Well, they're human.  Also, they had the political cycles to consider.  At that stage they had about two-and-a-half years before the election -- just in time to bring down on themselves all the kerfuffle you could imagine, if anything went dramatically wrong.

Sir Walters apologises even now that he did chicken out.  But I still think in retrospect it was the right thing to do.  It was far more important to get another nine years of Mrs Thatcher than to get monetary base control, even if it had come off.

The irony of it is that Britain is going by the road which is absolutely the thickest with thorns, boulders and hazards, towards a fixed exchange rate system and towards Deutschemark dominance;  and, indeed, to what I absolutely believe is not only a politically unified Europe, but a centralised, corporatist, protectionist Europe.  It will not be a Currency Board.  It will be a system dominated by what is called a Central Bank of Europe:  which in turn will be dominated, in my view, by French socialist civil servants.

What French socialist civil servants have been demanding for a long time now from the Bundesbank is symmetry.  They said, "Il est necessaire que nous gagnons la symmetry."  What the French mean by symmetry (it's not in the dictionary but it should be) -- is, "the French will control German monetary policy."  And to cut a long story short, I think they will gain it.  Now this is very different from the depoliticised institutional system of the Currency Board.  It is in fact, in my view, politicising, bureaucratising, all the other nasty words you can think of, on a grand scale, the monetary system.  And I am sure it's going to end in a disaster.

---

THE NEW ZEALAND EXPERIENCE

GETTING monetary policy to work has been one of the great unresolved tasks of economic management.  After vast amounts of research, enough to fill any good-sized library, one is forced to conclude that the only solutions that hold out solid prospects involve much more than tinkering with monetary policy techniques.  Significant institutional reforms are required, together with the all-important forbearance of society that is necessary to see the institutional reforms take deep root.

"WORK" IN RELATION TO WHAT?

In thinking about the topic at hand, it is important to understand what one is trying to achieve with monetary policy.  When we ask "Can monetary policy be made to work?", what do we have in mind as the objective?  Though a number of possibilities exist, there are really only two serious contenders:  the business cycle and price stability.  But let me deal briefly first with two possible contenders that are often in peoples' minds, yet that are not sensible active or direct objectives of monetary policy.

LONG-TERM GROWTH?

Nobody these days expects monetary policy actively to be able to lift the long-run potential growth rate of an economy.  It is possible, indeed common, for bad monetary policy to damage the actual growth rate.  So I would rule long-term growth out of contention as something which monetary policy can successfully target in any active way.  But a better long-term growth rate outcome is likely to be the by-product of monetary policy successfully targeted elsewhere.

EXTERNAL BALANCE?

A second possible objective of monetary policy is the current account.  Can monetary policy be made to work in respect of external balance?  I know this has, from time to time, been a subject of great but often inconclusive debate in Australia.  It has been less the subject of debate in New Zealand.  But if the debate were to arise in New Zealand, my view on it would parallel my view on long-term growth as an objective for monetary policy.

Monetary policy in general does not seem to have lasting effects on real variables, including in this the real exchange rate.  If monetary policy does not sustainably influence the real exchange rate, it is hard to see how it sustainably influences the external accounts.

On the one hand, to the extent that monetary policy is moving to dampen excessive demand, monetary policy tightness probably helps deal with a current account blow-out.  On the other hand, easy monetary policy can temporarily depress the real exchange rate by driving down the nominal exchange rate.  But both of these are cyclical and not permanent.

Thinking longer-term, some would argue that the process of nominal exchange rate adjustment is not a smooth function of differential inflation rates.  I would agree that, during New Zealand's high inflation history, exchange rate adjustment often lagged.  The result was a tendency to real exchange-rate over-valuation, and incipient balance of payments problems.

Does this work in reverse, and in a floating exchange-rate environment?  I don't know, but I suspect that holding down the inflation rate below world inflation rates could to some extent cause a tendency towards real exchange rate under-valuation.  Certainly monetary policy directed to keeping inflation low, even though often accompanied by periods of nominal exchange rate appreciation, has not hurt the external position of Japan or West Germany.

It is worth keeping in mind the distinction between something as the active target of monetary policy, as opposed to a potentially useful by-product of monetary policy successfully targeted elsewhere.

THE BUSINESS CYCLE?

Turning now to one of the prime contenders for the objective of monetary policy, can we make monetary policy work in relation to the business cycle?  Can we reliably adjust monetary policy settings in a way that actually dampens the amplitude of the cycle?  Much monetary policy-making world-wide appears to be based on the belief that we indeed can.  However, I would have to argue, and argue strongly, that our forecasting ability is not that good.  Nor is our knowledge of the power and speed of the effects of monetary policy.  Let me illustrate this point with a rather sobering reflection on New Zealand's own recent experience.

Forecasters in New Zealand, as elsewhere in the world, generally miss the mark -- but by no more than the statisticians miss the mark as far as history is concerned.  In other words, even in normal times forecasting is an imprecise business, focusing on variables such as GDP that are measured with wide margins of error.

Once every so often there occurs a change in growth which is large enough to be outside the bounds of measurement error.  Forecasters almost always miss predicting such events.  In New Zealand they have missed big swings in the 1970s and again in the 1980s.

And it looks like they have missed another one in the 1990s.  As recently as four months ago they were forecasting that the 1991-92 March year would see 1.5 per cent growth.  While the Reserve Bank has yet to complete its latest forecast round, most private sector forecasters are now projecting a fall in GDP of the same magnitude.  If they can't even predict a turning point of that magnitude, how can they ever hope to offset the cycle?

INFLATION?

Now the big one -- inflation.  Can monetary policy be made to work in respect of inflation?  Let me acknowledge at the outset that there is plenty of evidence around that monetary management has not worked very well with respect to the objective of price stability -- though monetary policy is equally clearly a potent influence on inflation.  Anyone who takes a moment to examine a graph of long-term inflation rates can see that the 20th century, particularly since the move away from the gold standard, is the century of inflation.  Not coincidentally, the 20th century is the era of the central bank, or at least of the central bank as an active manager of monetary affairs.

If central bankers have made such a mess of the additional responsibilities assigned to them, when the external check of the gold standard was removed, ought we not think of a way of doing without them, or instead, restoring a system of external checks?  The answer depends on what we see as the source of the problem.  One possibility is limited technical capability.  That is certainly a relevant issue in relation to the management of the business cycle, as I have already discussed.  But I'm not sure that it is all that much of a problem in relation to getting inflation down, and keeping it down.  More fundamentally worrying, as a potential source of the 20th century disease, is the case that has been developed in the literature over recent years -- that there is a basic inconsistency between the political demands placed on the monetary managers and the standards of monetary management required to safeguard the value of the currency.

The argument is familiar.  At a simple level, it is that the nature of the incentives facing the politicians means that a high premium is placed on output growth in the short term.  Although the presence of inflation might be detrimental to re-election prospects, reducing inflation typically brings only economic and social costs within the timed horizon of the electoral cycle;  the benefits come well thereafter.  This rather reduces the prospect of politicians choosing the inflation reduction route.  The incentives facing central bankers deserve mention here also.  Very often the central bankers are directed to pursue national welfare, loosely defined.  That very broadly stated objective permits central bankers excessive scope to get waylaid by unrealistic faith in their own technical competence to dampen the business cycle at the same time as controlling inflation.  It also allows central bankers scope to give freer rein to their hearts than to their heads.  Clearly, nobody in public service wants always to be seen as depressing income growth, and it is very easy to be tempted by apparently obvious -- but ultimately illusory -- opportunities to avoid doing so.

All these things tend to produce a bias towards higher inflation, without higher inflation being an active objective of monetary policy.  A growing understanding of this fact is one of the major contributions made by economic theorists in recent years.

OPTIONS FOR MONETARY POLICY TO WORK IN RELATION TO INFLATION

So we have come some way in understanding the nature of the problem of making monetary policy work in relation to the thing that monetary policy can sustainably affect -- inflation.  Is there a solution?

Many people have advocated the use of operational rules that prevent monetary policy from being turned towards the ultimately futile, and dangerous (for price stability, that is) pursuit of management of the business cycle.  The advocates' faith in simple rules has been shattered by the rapid realisation that the world is not a simple place.  Trying to operate by way of simple rules in these circumstances would not lead to the best of outcomes.

It is tempting, therefore, to attempt to devise rules which match the complexity of the real world.  Some of these complex, "state-contingent" rule systems even seem to be able to perform moderately well when the researcher is endowed with the benefit of hindsight.  But, without that benefit, complicated rule systems that probably cannot be explained -- let alone operated -- are clearly non-starters.

One alternative approach is to abrogate responsibility to a reputable outside monetary authority, such as the Bundesbank (despite the fact that even the Bundesbank has been far from perfect as a guardian of price stability).  While this is a possibility for a single country, it simply shifts the location of the underlying problem, when viewed from a world perspective.  Besides, there are probably large trade-offs involved, particularly where frequent or substantial real exchange rate adjustment is likely to be a feature of the economic scene.

Going back to a gold standard, or inventing a new commodity standard, has some theoretical attractiveness.  But it would be a radical change in the technology of monetary policy (and related financial institutions).  The commodity or "gold"-standard approaches are similar in many respects to the idea of pegging the currency to that of a low-inflation country.  In both cases, the choices that the politicians would have to make in order to implement the new monetary arrangements are not fundamentally different from the choices they would have to make to eschew the inflation option under current arrangements.  The only advantage of an "external" standard might be that the radical nature of the change might itself lock the system in place in a more credible way than is possible under current arrangements.

So none of the alternatives is overwhelming.  All of the fundamental solutions to the bad incentives problem involve politicians and central bankers forgoing the capacity to misuse monetary policy in pursuit of short term political or social conscience objectives.  But in few cases can one imagine that we'd be daring enough to change the institutional arrangements for monetary management in a way that would permanently cut off the option of making adjustments.  Given our lack of knowledge about the workability, or the sense, of locking in a new and radical approach (or even an old and radical approach), it would seem undesirable to go quite that far.

THE NEW ZEALAND APPROACH

Are we, then, completely stuck for an answer?  I don't think so.  At various times and in various circumstances, politicians and societies have the capacity to think long-term.  In New Zealand's case, that has been demonstrated during the last decade.  As is common, the incentive to think long-term came from a long period of difficulty with economic policy and performance, brought to a head by a crisis.

In such circumstances, politicians can distance themselves from the capability to micro-manage certain elements of economic policy.  The trick is to find a way which is not too radical (in the sense of not overturning completely the existing institutional arrangements and in the sense of remaining within the broad constitutional parameters developed over the years), but which cements in place as much as possible a new approach that will last in the face of a reversion to short term political decision-making.  Moreover, the approach must firmly direct the operation of monetary policy towards the objective that is relevant -- price stability.

The new arrangements in New Zealand go some considerable way towards satisfying these design criteria, though there remain inevitable weaknesses.  Let me now spend the rest of my time describing these institutional arrangements.

The Reserve Bank of New Zealand currently operates under a piece of legislation called the Reserve Bank of New Zealand Act 1989, which is quite unusual in several respects.  First of all, it was passed with almost unanimous support in the House -- and that was as unusual in their Parliament as, I imagine, it would be in ours.  To be fair, I guess, it should be noted for the sake of completeness that Sir Robert Muldoon was in hospital at the time!

The second point to note is the Act specifies that monetary policy will be directed towards the single objective of achieving and maintaining stability in the general level of prices.  Price stability is not defined in the Act, but it is explicitly clear that price stability is the only objective for monetary policy.

Thirdly, the Act requires a written agreement between the Minister of Finance and the Governor on what price stability actually means.  In other words, what index is used, what numerical range is consistent with price stability, when it must be accomplished, and so on.  That agreement must be made public -- it is a public document, tabled in the House within so many days of being signed.  The policy targets agreement that have been signed by the Governor requires the central bank to achieve a movement in the CPI of 0-2 per cent by the end of calendar 1993:  although it explicitly excludes certain exogenous shocks, like an increase in the goods and services tax, a pronounced shift in the terms of trade, or the price impact of an outbreak of foot-and-mouth disease.

Fourthly, there is total autonomy for the central bank regarding the levers of monetary policy, subject only to achieving that nominated objective.

Fifthly, the Act requires a high degree of accountability.  The central bank must publish, at least every six months, a comprehensive statement of their progress towards the published objective:  the indicators they are watching, how they think they are advancing, and so on.  What they've in fact done is to nominate not simply, of course, the end target (which was already public), but also the interim inflation objectives prior to 1993, so that markets get a reasonable feel for what they think they are doing with monetary policy.

I mentioned the agreement between the Minister of Finance and the Governor.  I emphasise that point because to ensure effective accountability, the Act explicitly provides that any Governor can be dismissed for failure to achieve the nominated inflation target.  And to that end, the Board of the Bank has explicitly no authority at all to make decisions on monetary policy.  The board is explicitly an advisory board, a monitoring board, which can recommend to the Minister that the Governor be fired for not managing the Bank effectively or not achieving the inflation objective.  But the board does not make decisions, and that is principally so that the Minister would know who to fire if the objective was not met.  I must say that constitutionally I find that very elegant -- whether the Governor finds it so elegant in due course remains to be seen.

The next point to make is that politicians have actually made it difficult for themselves to abandon or emasculate the price stability objective, by making any reversions very public.  They have the right under the Act to change the policy targets agreement, but they have to do it in a totally public way.  What they cannot do is to say to the public, "We believe in price stability", but say to the Bank, "For Pete's sake get the mortgage rate down before the election".  That option has been closed off for them.  Even the casual suggestion that they might modify the Act or the agreement produces an immediate and adverse effect on bond yields.

The last point about the Act itself is that it clearly quantified and time-stamped targets which played a very important part in building credibility, both with the policy targets agreement and with their self-imposed indicative inflation ranges.

Looking at the effect on credibility, it seems to me that there has been a very marked improvement in the credibility of both the Bank specifically, and monetary policy more generally, over the last couple of years.  One piece of evidence for that is the newspaper article from Ken Douglas, the head of the New Zealand Council of Trade Unions.  In that article, published late in 1990, he described the Governor, in very unflattering terms indeed, as a megalomaniac, totally obsessed with price stability.  He didn't approve of the Act, totally opposed it and advocated its change.  But unfortunately, he said as long as the Act remained, the union movement had no alternative but to gear its wage settlements to that reality, otherwise unemployment would clearly go through the roof.

The fact that bond rates in New Zealand have fallen from about 13 per cent for the six-year bond 12 months ago to around 8.5 per cent now is also a partial measure of the very substantial improvement in monetary policy credibility over the last year.

Interestingly, just before last year's election, an opinion poll was conducted on whether the incoming government, whichever party it was, should stick to 0-2 per cent as its inflation target.  Given the fact that they had high and fairly rapidly rising unemployment, I would have expected only a small minority of New Zealanders would have advocated sticking with 0-2 per cent.  But to my surprise, most of those who had an opinion favoured sticking with 0-2 per cent as the goal of monetary policy, which was certainly very encouraging.

Finally, I think it's fair to say that the system remains open to political reversal, and as such it is not a permanent rearrangement of the institutional structure for monetary management.  The openness for political reversal in part reflects a view on the appropriate constitutional balance to be struck in societies like theirs.  Charles Goodhart made quite a considerable contribution to the thinking behind the New Zealand Act.  He argued two propositions:  (1), that independent central banks achieve price stability at a lower real social cost than do the ones which are directly controlled day-to-day by governments;  (2), that in a democratic society, it is very unlikely that governments will totally refrain from any involvement in monetary policy at all.  The New Zealand structure allows the Government to define what it means by price stability, and to make the decision of how fast monetary policy should achieve that objective.  But having done that, the Government then leaves the implementation to the central bank.

In the absence of even more radical changes than they've instituted, which are unlikely, what would best limit the chance of reversal is time -- time for the price stability objective of an independent central bank to become embedded in the implicit contract with society.

---

FURTHER READING

"ANALYST", "Currency and Banking in Jamaica."  Social and Economic Studies, Vol. 1, 1953, pp. 41-53.

BASU, S.K., A Review of Current Banking Theory and Practice.  Calcutta:  Macmillan, 1971.

BATTELINO, R., and MacFARLANE, I.J., "Open Market Operations:  Some International Comparisons", Economic Papers, Vol. 7 No. 1, March 1988, pp. 65-81.

BIRNBAUM, Eugene A., "The Cost of a Foreign Exchange Standard or of the Use of a Foreign Currency as the Circulating Medium", International Monetary Fund Staff Papers, Vol. 5, No. 3, February 1957, pp. 477-91.

BLOWERS, G.A., and MCLEOD, A.N., "Currency Unification in Libya", International Monetary Fund Staff Papers, Vol. 2, No. 3, November 1952, pp. 439-67.

CAPIE, Forrest, and WOOD, Geoffrey E. (eds.), Unregulated Banking:  Chaos or Order?, St Martin's Press, New York, 1991.

CARMICHAEL, J., and HARPER, Ian R., Implementing Monetary Policy in a Deregulated Financial System:  A Model of the Short-Term Money Market in Australia, 1991, mimeo.

CHALMERS, Robert C., A History of Currency in the British Colonies.  London:  Spottiswode and Eyre for HMSO, 1893.

CLAUSON, G.L.M., "The British Colonial Currency System", Economic Journal, Vol. 54, No. 213, April 1944, pp. 1-25.

COLLYNS, Charles, Alternatives to the Central Bank in the Developing World:  Occasional Paper No. 20.  International Monetary Fund, Washington, 1983.

DEMSETZ, Harold, "Why Regulate Utilities?", Journal of Law and Economics, Vol. 11, No. 1, April 1968.  pp. 55-65.

DOTSEY, M., "Open Market Operations in Australia:  A US Perspective", The Economic Record, September 1991, pp. 243-256.

DOWD, Kevin (ed.), The Experience of Free Banking, Routledge, London, forthcoming.

DOWD, Kevin, The State and The Monetary System, Philip Allan, Oxford, 1989.

DRAKE, P.J. (ed.) Money and Banking in Malaya and Singapore, Malayan Publications, Singapore, 1966.

EDO, Michael E., "Currency and Banking Legislation in the Arabian Peninsula", International Monetary Fund Staff Papers, Vol. 22, no. 2, July 1975, pp. 510-38.

GLASSNER, David, Free Banking and Monetary Reform, Cambridge University Press, Cambridge, 1989.

GREAVES, Ida C., Colonial Monetary Conditions, HMSO, London, 1953.

GREENWOOD, John G., "How to Rescue the HK$:  Three Practical Proposals", Asian Monetary Monitor (Hong Kong), Vol. 7, No. 5, September-October 1983, pp. 11-39.

GREENWOOD, John G., "The Stabilisation of the Hong Kong Dollar", Asian Monetary Monitor (Hong Kong), Vol. 7, No. 6, November-December 1983, pp. 9-37.

GREENWOOD, John G., "Time to Blow the Whistle", Asian Monetary Monitor (Hong Kong), Vol. 5, no. 4.  July-August 1981, pp. 15-33.

GREENWOOD, John G., "Why the HK$/US$ Linked Rate System Should Not Be Changed", Asian Monetary Monitor (Hong Kong), Vol. 8, No. 6, November-December 1984.

GREENWOOD, John G., and GRESSEL, Daniel L., "How to Tighten Up the Linked Rate Mechanism", Asian Monetary Monitor (Hong Kong), Vol. 12, No. 1, January-February 1988, pp. 2-13.

GRENVILLE, S., "The Operation of Monetary Policy", The Australian Economic Review, February 1990, pp. 6-17.

GRESSEL, Daniel, Soviet Macroeconomic Imbalances During Perestroika, unpublished manuscript, GT Capital Management, San Francisco, 1989.

HANKE, Steve H. and SCHULER, Kurt, "Keynes' Russian currency board", in Steve H. Hanke and Sir Alan Walters (eds.), Capital Markets and Development, Institute for Contemporary Studies Press San Francisco, 1991.

HANKE, Steve H. and SCHULER, Kurt, Currency Boards and Economic Development, forthcoming.

HANKE, Steve H., and SCHULER, Kurt, "Keynes and Currency Reform:  Some Lessons for Eastern Europe", Journal of Economic Growth, Vol. 4, no. 2, 1990, pp. 10-16.

HANKE, Steve H., and WALTERS, Sir Alan.  "Reform Begins with a Currency Board", Financial Times, February 21, 1990.

HARPER, Ian R., "Monetary Policy in a Deregulated Financial System", The Australian Economic Review, April, 1988, pp. 58-63.

HARPER, Ian R., and PEARCE, J.J., "Implementing Monetary Policy in an Era of Budget Surpluses", The Australian Economic Review, February 1990, pp. 53-65.

HAZLEWOOD, Arthur, "The Economics of Colonial Monetary Arrangements", Social and Economic Studies, Vol. 3, Nos. 3-4, December 1954, pp. 291-315.

HETZEL, Robert, "Free Enterprise and Central Banking in Formerly Communist Countries", Federal Reserve Bank of Richmond Economic Review, July 1990.

INTERNATIONAL MONETARY FUND, International Financial Statistics Yearbook 1990.  International Monetary Fund, Washington, 1990.

KING, Frank H.H., 1957, Money in British East Asia, HMSO, London, 1957.

KRATZ, Joachim W., "The East African Currency Board", International Monetary Fund Staff Papers, Vol. 13, no. 2, July 1966, pp. 229-55.

KYDLAND, Finn E., and PRESCOTT, Edward C., "Rules Rather Than Discretion:  The Inconsistency of Optimal Plans", Journal of Political Economy, Vol. 85, No. 3, June 1977, pp. 473-92.

LOYNES, J.B., The West African Currency Board 1912-1962, West African Currency Board, London, 1962.

McTAGGART, D., and ROGERS, C., "Monetary Policy and the Terms of Trade:  A Case for Monetary Base Control in Australia?", The Australian Economic Review, February 1990, pp. 38-49.

NELSON, William Evan, The Imperial Administration of Currency and British Banking in the Straits Settlements, 1867-1908, Ph.D. dissertation, Duke University, 1984.

NEVIN, Edward, Capital Funds in Underdeveloped Countries, St Martin's Press, New York, 1961.

NEWLYN, W.T. and ROWAN, D.C., Money and Banking in British Colonial Africa, Oxford University Press, London, 1954.

OW, Chwee Huay, The Currency Board Monetary System -- The Case of Singapore and Hong Kong, Ph.D. dissertation, The Johns Hopkins University, 1985.

SELGIN, George A., "A Free Banking Approach to Reforming Hong Kong's Monetary System", Asian Monetary Monitor (Hong Kong), Vol. 12, No. 1, January-February 1988, pp. 14-24.

SELGIN, George A., The Theory of Free Banking:  Money Supply under Competitive Note Issue, Cato Institute, Rowman & Littlefield, New Jersey, 1988.

SHANNON, H.A., "Evolution of the Colonial Sterling Exchange Standard", International Monetary Fund Staff Papers, Vol. 1, No. 3, April 1951, pp. 334-354.

SHANNON, H.A., "The Modern Colonial Sterling Exchange Standard", International Monetary Fund Staff Papers, Vol. 2, No. 2, April 1952, pp. 318-362.

SMITH, Vera C., The Rationale of Central Banking and the Free Banking Alternative, P.S. King & Son, Westminster, 1936.  Republished Liberty Press, Indianapolis, 1990, with a foreword by Leland Yeager.

WALTERS, Sir Alan, "Currency Boards", in John Eatwell, Murray Milgate and Peter Newman (eds.), The New Palgrave:  A Dictionary of Economics, Vol. 1, Cambridge University Press, Cambridge, 1988, pp. 740-2.

WHITE, Laurence H., Free Banking in Britain:  Theory, Experience, and Debate, 1800-1845, Cambridge University Press, Cambridge, 1984.

---

ENDNOTES

1.  WHITE, G.M., "Economic Recovery, Drought and Commodity Prices", Australian Financial Outlook 91/10, Chase AMP Bank, Sydney, October 1991.

2.  BATTELLINO, R., and McMILLAN, N., "Changes in the Behaviour of Banks and Their Implications for Financial Aggregates", Research Discussion Paper, RDP 8904, Reserve Bank of Australia, Sydney, July 1989, p.26.

3.  BEWLEY, R.A., and WHITE, G.M., "Do High Interest Rates Improve or Worsen the Current Account?", Economic Papers, 9:4, December 1990, pp. 19-33.

4.  Budget Paper No. 1, Statement No. 2, Commonwealth of Australia, Canberra, August 1989, p. 2.8.

5.  KEATING, P.J., Press Conference Transcript, Canberra, August 1990.

6.  Commonwealth of Australia, op. cit., p. 2.2.

7.  WHITE, G.M. "Was Demand Growth Unsustainable in 1988/89?", Economic Papers, 10:3, September 1991, pp. 81-88.

8.  Budget Statement No. 2, Canberra, Commonwealth of Australia, August 1991, p. 2.33.

9.  MATHEWS, R., "Free-market policies have been disastrous", Canberra Times, 14-15 November 1991.

10.  MacFARLANE, I.J., "The Lessons For Monetary Policy", Proceedings of a Conference on The Deregulation of Financial Intermediaries, Reserve Bank, Sydney, 1991.

11.  MacFARLANE, I.J., op. cit.

12.  MacFARLANE, I.J., op. cit.

13.  COMMONWEALTH OF AUSTRALIA, Budget Paper No. 1, Statement No. 2., 1984.

14.  MOORE, J.D., "Industrial Relations and The Failure of The Accord:  What Should Be Done", Australian Bulletin of Labour, Vol. 15 No. 3, June 1989.

15.  Between mid-1980 and mid-1983 the average CPI rate in Australia's major OECD trading partners fell from around 12 per cent per annum to around 3.5 per cent per annum.  Australia's CPI rate fell from a peak of around 12 per cent per annum in mid-1982 to a bottom of around 5 per cent per annum at the end of 1984 (assisted toward the end by the so-called "Medicare fiddle" on the CPI calculations).

16.  Interestingly, proponents of the Accord have been strangely silent about the causes of the most recent reduction in inflation!

17.  There is a similar pattern with the increase in Australia's nominal unit labour costs, which actually fell below the OECD average in 1984.

18.  One of my concerns in the early period of deregulation was that the increase in broader credit aggregates following deregulation indicated that monetary policy needed to be considerably tighter.  This was strongly opposed by the Reserve Bank, on the ground that the increase in broader credit aggregates was largely a function of re-intermediation -- in order words, credit that had previously been provided from outside the financial system was now being provided from inside the system.  Indeed, through most of the period from the early 1980s up to early 1987, the Bank was often able to find a reason why policy should not be tightened -- or, at least, "not yet."

19.  Indexation was the initial basis for such increases, the theory being that productivity growth would feed through into lower prices.  The initial justification for moving away from full indexation was that there needed to be discounting for the fall in the terms of trade and the depreciation in the dollar.  Later on, recognition of the need to emphasise productivity led to the abandonment of any formal link with price movement.  As time progressed, differences also developed between the Government and the ACTU, in the cases for wage increases which they put to the Industrial Relations Commission.

20.  The most notorious deregulatory move opposed by the Treasury was the move in December 1983 to remove all exchange controls on capital flows between Australia and overseas (other than those imposed for reasons of inward foreign investment policy, most of which still exist) and to "float" the $A.  That decision was taken without the benefit of any submission to senior Ministers on possible broader policy implications.

21.  Indeed, MacFarlane might be said to have implicitly acknowledged that in his earlier joint article with Glenn Stevens entitled "Overview:  Monetary Policy and the Economy" in Proceedings of a Conference on Studies in Money and Credit, Reserve Bank, Sydney, October 1989.  The conclusion of that article was that the ultimate goal of monetary policy is price stability (p. 8).

22.  MacFARLANE, I.J. and STEVENS, G., "Overview:  Monetary Policy and the Economy", Proceedings of a Conference on Studies in Money and Credit, Reserve Bank, Sydney, October 1989.

23.  Clearly, it would be ridiculous to try to stabilise share prices, for example.  However, if prices of all major types of assets are increasing strongly together, that would be a warning signal.  A senior official of the Bank of England recently told me that the Bank is now paying close attention to trends in house prices.

24.  It maybe argued that, as the use of monetary policy in 1988-89 did produce a recession, it might have been better to have let the external crisis occur anyway.  However, as noted, the way in which monetary policy was operated in that period was inept.

25.  See papers in Studies in Money and Credit, Reserve Bank, Sydney, October 1989.

26.  It is interesting also that the Governor could do no more than play his "hunch" that setting a 0-2 per cent inflation rate two years ago would have deepened the recession.  This is scarcely surprising given that he said this without even specifying the time period set for its achievement!

27.  MacFARLANE, I.J., "The Lessons for Monetary Policy", in MacFarlane, Ian (ed.), The Deregulation of Financial Intermediaries, Proceedings of a Conference, Reserve Bank of Australia, Sydney, 21 June 1991.

28.  HARTLEY, Peter R., and PORTER, Michael G., "Treasurer Keating's Legacy", Tasman Institute, Occasional Paper B10, July 1991.

29.  HARTLEY, Peter R., and PORTER, Michael G., op. cit.

30.  MacFARLANE, I.J., op. cit.

31.  This section draws heavily on Carmichael and Harper, op. cit.

32.  Exchange settlement accounts are current accounts held by banks and authorised dealers with the RBA in order to effect final settlement of debt among banks themselves, between banks and dealers or between banks/dealers and the RBA.

33.  A repurchase agreement or "repo" is effectively a secured loan in which title to the collateral is actually transferred.  When the RBA "buys" a repurchase agreement, it buys a Treasury Note from a dealer for immediate value and simultaneously agrees to sell the Note back to the dealer at a future date at an agreed price.

34.  The deposits must come from the non-bank sector because any deposits raised from the banks would immediately put the banks' exchange settlement accounts back into deficit, thereby restarting the entire process.

35.  Each Wednesday, the RBA offers Treasury notes for auction;  it offers Commonwealth Bonds at less regular intervals.  Successful bidders (mostly banks and dealers) have one week in which to settle their lots.  The decision to settle is heavily influenced by the overnight money-market rate, since this rate represents the funding cost of taking the securities up before the end of the week.  The only settlement transactions that affect the cash market on the day are settlements by dealers (all other Treasury Note settlements affect banks' exchange settlement accounts the following day).  Thus, once the cash rate falls below the yield on securities won by the dealers at the last tender, the dealers will demand cash to settle.  This amount varies from day to day according to the stock of unsettled securities available to dealers.  This is shown as the rightward kink EF in the demand schedules in Figures 1 and 2.

36.  See Harper and Pearce, op. cit., for discussion of this issue.

37.  The RBA currently allows banks to hold excess reserves with itself in the form of interest-bearing deposits which it offers to banks from time to time.

38.  This point is made to great effect in McTaggart and Rogers, op. cit.

39.  KEATING, P.J., Transcript of Press Conference, 16 February 1989.

40.  FRIEDMAN, Milton, A Program For Monetary Stability, Fordham University Press, New York, 1960.