RBA Annual Conference – 1989 Monetary Targeting: The International Experience Victor Argy,^[*] Anthony Brennan^[**] and Glenn Stevens^[**]

1. Introduction

A number of countries adopted monetary targeting from around 1975.^[1] By the late 1970s, all of the seven largest Western economies – the U.S., Japan, West Germany, the U.K., France, Italy and Canada – and two smaller economies – Australia and Switzerland – had monetary targets of one form or another in place.^[2] This represented the culmination of a shift towards quantities as the focus of monetary policy, with correspondingly less emphasis on financial prices as ends in themselves, which began in many cases in the 1960s or early 1970s.

But monetary targets were frequently missed. In addition, by the mid 1980s, two of these countries had explicitly dropped monetary targets; others were placing less influence on targets as the centrepiece of monetary policy, and most had changed, at least once, the aggregate chosen as the subject of the target.

This paper aims to review this experience. Previous treatments of the international experience of monetary targeting include Foot (1981), Johnson (1983), OECD (1979), Sumner (1980), Isard and Rojas-Suarez (1985). Some time has passed since most of these papers were written, however, and with that passage of time have come some new developments that may have a bearing on the assessment of targeting.

The second section of the paper discusses the rationale for targeting. Section 3 looks at the experience of nine countries in turn. To explore in depth all the aspects for each country would be a large undertaking, and could not be confined within the available space. The aim is rather to give the reader some of the economic background and a feel for the issues during the period in question. Section 4 attempts to distil the lessons from the various experiences, in a common framework.

2. The Adoption of Monetary Targeting

The conceptual case for monetary targeting has its roots in the calls for a monetary rule – low, stable growth of the money stock – made by Friedman (1960). In his conceptual framework, there is a reasonably close relationship between money and nominal national income over the medium term. The analytical foundation of this is the existence of a stable demand function for real money balances in terms of real income (given interest rates). For Friedman, monetary policy – thought of as supply-side variations in the nominal stock of money – could have large short-term effects on the economy, but in the medium term, such disturbances would show up only in prices. In addition, the short-term relationship between money and the economy was not well understood, and therefore not dependable for stabilisation purposes. These factors suggested a non-discretionary monetary rule as an appropriate focus for monetary policy.

Throughout the 1970s, the question of the appropriate rule for monetary policy was a topic widely discussed in the macroeconomics literature. In a very widely-quoted paper, Poole (1970) concluded that, given an objective of stabilising output (not prices – they were fixed in Poole's analysis), an interest rate target was superior to a money target for a money demand shock, and that a money target was superior to an interest rate target for a shock to the real sector of the economy. He also proposed that, in the face of shocks originating from both sources (the more realistic situation), an optimal policy is a compromise which allows the money supply to respond positively to the interest rate but not to the extent that interest rate changes are fully offset. (For more recent analyses of such a rule, see Benavie and Froyen (1983) and Dotsey and King (1983).)

Other influential papers in the 1970s went further than Poole, by explicitly considering expectations and price variability. From this literature came several propositions which lent support to a monetary rule. In new classical models, anticipated (or systematic) monetary policy had no effects, and unanticipated monetary changes had temporary but not persistent effects.

Later, and in a similar vein, it was argued that discretion would continually tempt monetary authorities into over-expansionary monetary policy, thus ultimately creating a bias in favour of “excessive” inflation (see Argy (1988a)).

The alternative to a money-stock rule, a nominal interest rate rule, was not thought to be a useful approach to monetary policy. In the event of a positive shock to output, for example, or an upward shift in inflationary expectations, attempting to offset the natural tendency for interest rates to rise would mean easier monetary policy, which would exacerbate the problem. In formal models with rational expectations, the price level has no determinate solution under an interest rate rule. (For a treatment of these questions, see Isard and Rojas-Suarez (1985).)

All of these arguments were doubtless influential. Probably more influential than conceptual treatments was recent economic history itself, which had demonstrated in fairly devastating fashion that rapid growth in prices and rapid growth in the money stock went together. Certainly, at the time of adoption of targeting in the mid 1970s, the recent experience of all countries concerned had been one of relatively high inflation, accompanied by rapid and variable growth in monetary aggregates. It became widely accepted that a necessary, if not sufficient, condition for lower inflation was lower monetary growth.

Against that background, the strategy of monetary targeting as proposed by its supporters had three principal operational features.

First, monetary aggregates should be targeted to grow at rates which would be conducive, over the medium term, to low rates of inflation. In most countries, this would involve lower rates of monetary growth than had been experienced in the early 1970s, though in some countries money growth had already been reduced by the time targets were adopted.

Second, monetary growth was not only to achieve an “end-point” target, but was supposed to move along a reasonably stable path towards that end point. This stability in itself was an objective of targeting, and would require more stability in rates of growth of the monetary aggregates than had been experienced in the first half of the 1970s.

Third, the targets were to be publicly announced in advance. The objective here was one of favourably affecting expectations by making a credible commitment to policies which would contribute to a reduction in inflation. This was seen as reducing the costs in terms of foregone output, under some circumstances, of lowering inflation.

While the experience of rapid inflation in the first half of the 1970s had heightened perceptions of the need for firmer monetary policies, and the dangers of nominal interest rates as guides, and had thus shifted opinion in favour of monetary targeting, there were other factors which were also important.

The first of these was the move away from the Bretton-Woods system of fixed exchange rates to a system of generalised floating among the major countries. This, conceptually at least, allowed domestic monetary policy to pursue domestic objectives and as such had long been advocated by monetary targeting's supporters. At the same time, of course, it removed one of the candidates for the role of anchor to the system. Some comfort could perhaps be taken from a policy such as monetary targeting which explicitly sought to tie down another nominal magnitude.

The second was that in some countries, large and growing public sector deficits created tensions in monetary management. A monetary target could be a way of ensuring that monetary policy was not too “accommodative”. This was particularly so in countries which used broadly-defined targets, where the “counterparts” type of analysis showed clearly the connection between public deficits and monetary growth. (For a discussion of the U.K., for example, see Goodhart (1989); for France, Raymond (1983a).) In such circumstances, too, there may have been a desire for a “heat shield” for policy-makers (a phrase coined by Blinder (1987)), in the event that painful decisions had to be made on interest rates.

But while monetary targeting was adopted by all of the big countries, in a number of the smaller countries other approaches to monetary policy were favoured.

Two reasons are sometimes offered for why smaller economies have shied away from targeting. The first is that many of them are and have been on fixed exchange rates, so they have not had the capacity to control the money stock. This argument, in part at least, explains why countries like Belgium, Austria, Ireland, Denmark have not announced money growth targets. It may also explain why Switzerland, which had a flexible rate regime, felt able to adopt targets. On the other hand, EMS members such as France and Italy, with only limited exchange rate flexibility against European countries, did target. Australia, on a crawling peg regime after 1976 (to 1983), chose to adopt a form of money growth targeting. The argument is also one of degree; as we shall see, countries which did have monetary targets ignored them at some times in favour of exchange rate considerations.

The second reason is that many smaller economies have the perception that their demand for money is relatively unstable, thus rendering money growth targeting inappropriate. The judgment may be correct (indeed their relative openness may argue for this) but, as it happens, there is little empirical support available for the proposition.

3. Country Experiences

This section looks briefly at the experiences of nine industrial countries which adopted monetary targets or projections. The order is alphabetical.

(a) Australia

The first half of the 1970s saw a marked acceleration in monetary growth in Australia. The world commodity price boom in 1972/73 provided a very large lift in the country's terms of trade, pushing the current account temporarily into surplus. At the same time, international markets saw Australian dollar assets as underpriced and there was considerable capital inflow. With a fixed exchange rate, these factors provided a substantial increase in domestic liquidity and monetary growth, as well as in incomes and expenditures.

By 1974, inflation had picked up substantially. A sharp tightening in financial conditions in 1974 (a product of a marked deterioration in the current account, a large penalty on foreign borrowing through a high reserve requirement on foreign funds, and the fixed exchange rate) produced a very sharp slowing in monetary growth. But this proved to be short-lived and monetary growth quickly accelerated again. In the year to December 1975, M3, long the preferred measure of money in Australia, rose by 22 per cent. Thus monetary growth had not only been fast on average, but also highly volatile.

Against this background, the newly-elected government took several steps to tighten monetary policy, and announced in March 1976 an expectation that M3 growth in the six months to June 1976 would be in the range 11–13 per cent (at a seasonally-adjusted annual rate). In August that year, the Treasurer announced at the time of the Commonwealth budget a projection for growth in M3 over the twelve months to the following June, a practice which was maintained, with only slight variations, in subsequent years.

No specific reason was given for preferring M3, which includes currency and all bank deposits, over the narrower M1. But the emergence of large budget deficits in the mid 1970s, and the tendency for large flows (in either direction) across the foreign exchanges, led naturally to a “formation table” approach to monetary analysis, with M3 at the “bottom line” of the consolidated balance sheets of the official and private banking sectors.

In the process of implementing monetary policy, the Reserve Bank of Australia relied at that time on a number of direct controls. A variable reserve requirement could influence the split between free and frozen reserves of the banks. The Reserve Bank had the power to enforce quantitative lending restrictions on the banks, and to set the interest rates banks could pay on deposits and charge for loans.

The first two years of the targeting regime saw success in stabilising the growth of M3, and reducing it to around 8 per cent by 1977/78, in line with the projections. Inflation also came down in this period.

In the following year, 1978/79, growth was to have fallen further under the projection, but instead rose to almost 12 per cent. Growth in M3 exceeded the target for the next two years as well. A combination of factors contributed to this: an improving balance of payments situation added to monetary growth; inflexibility in yields for long-dated government paper meant the authorities could not always place sufficient debt in private non-bank hands to neutralise the effects of budgetary and official transactions. In addition, there was increasing difficulty in influencing bank lending. The prevalence of overdraft-type lending in Australia, with controlled interest rates, meant that even if banks kept the increase in new overdraft limits within the guidelines set by the Reserve Bank, a rise in usage of already existing limits could increase bank loans outstanding by a large margin. Indeed, a tightening of general credit market conditions could increase, in the short term, the demand for bank finance.

Inflation picked up as well in this period, with the strength of the economy and the rise in international oil prices. By 1981/82, Australia faced high inflation, an explosion of labour costs (further adding to inflation), and a worsening current account deficit, against the backdrop of a pronounced international recession.

Monetary policy had been moving in the direction of tightening for some time. In the first half of 1982, short-term interest rates reached levels not seen even in the very tight conditions of 1974; in real terms the level was much higher than anything seen before. Monetary growth slowed in 1981/82, more or less achieving the projected range for the year.

The economy slowed more abruptly, recording in 1982/83 its biggest fall in real GDP since the 1930s. Monetary growth in that year, while close to the projected range, did not slow much at all, and certainly not as much as might have been expected given the severity of the recession.

By 1983/84, the economy was recovering strongly from the recession, under the stimulation of very expansionary fiscal policy and declining interest rates. The monetary projection for that year was unchanged from 1982/83; this implied a large increase in M3 velocity, which had reached an unusually low point during the recession.

There was some considerable success in slowing inflation in this period. The rate of increase of the CPI came down from around 12 per cent in mid 1982 to around 5 per cent by the end of 1984. Monetary policy could claim direct responsibility for only some of this slowing, however, and monetary projections in particular even less: the projections were, if anything, slightly overshot, with M3 growth of over 11 per cent. Two other important factors were the wages freeze of 1982 and the weakness of the economy.

Throughout 1983, the conduct of monetary policy had been complicated by large changes of sentiment in the foreign exchange market. A large capital outflow in March, associated with the change of government, precipitated a downward adjustment of the Australian dollar. But by the end of 1983, speculative capital inflow had proved so difficult to contain that the currency was floated.

This was arguably the most important step in a series of deregulating moves. By the late 1970s, it had been recognised that the regulatory regime had hindered the development of the banking system, but had not really been effective in rationing finance provided to the private sector, since a large non-bank financial sector, largely free of the regulations, had grown up. The response to concerns over efficiency in the financial system was to deregulate. In the last years of the 1970s, and through the first half of the 1980s, the Australian financial system was deregulated more quickly and more completely than in almost any other country. As well as the floating of the currency, interest rates were freed up, and quantitative lending restrictions on banks lifted. These changes offered for the first time to the Reserve Bank of Australia the ability to exercise better control over the money stock (see Macfarlane (1984)).

But other changes which took place around the same time made things more difficult. The freeing up of the banks to set their rates at competitive market levels made for profound changes in the behaviour of monetary aggregates. From passively accepting deposits and managing the assets side of their balance sheets, banks became active liability managers. The response of monetary aggregates to changes in interest rates altered, since the authorities could not alter the opportunity cost of holding “money” simply by allowing market interest rates to rise or fall. The greater competitiveness of banks vis-a-vis the non-bank financial intermediaries, heightened by foreign participation in the banking system from 1985, led to a large change in market shares (reversing a long-term trend). This pointed to a likely rise in M3 (which was restricted to bank liabilities), which it would not have been sensible to resist with monetary policy. In addition to this, some non-bank intermediaries took up a number of new banking licenses, further adding to M3. As a result, conventional empirical relationships between M3, GDP and interest rates did not generally handle the 1980s well.^[3]

By the end of 1984, growth in M3 was running above the projection for 1984/85, and set to rise further. (The outcome for the year to June 1985 was 17½ per cent, well above the projected 8–10 per cent.)

In the light of the above factors, the Government took the view that the rapid growth in M3 did not in itself call for a tightening of monetary policy. The M3 projection was therefore abandoned in early 1985, and has not been reinstated or replaced by any other target. In its place has stood an openly-declared approach of assessing the appropriate stance of policy on the basis of a range of indicators.

Monetary growth since the abandonment of the target has tended to be high, and variable, partly as a result of the sorts of factors noted above. At the same time, inflation in Australia has been lower on average than it was during the targeting period, even though held up by the effects of a large depreciation of the currency in 1985 and 1986.

(b) Canada

Rapid monetary growth in Canada in the early 1970s had been associated with a large increase in inflation, which reached 15 per cent in 1974. In 1975, inflation had fallen to around 10 per cent, but had become stuck there. At the same time, monetary growth had risen again.

Late in 1975, the Bank of Canada announced a target for growth of M1 in 1976 of between 10 and 15 per cent, citing the need to contain monetary growth in the medium term to reduce inflation.

The policy of reducing inflation was avowedly gradualist in nature. Consistent with this, Canada had not undergone a contraction of output in the 1974–75 recession to the same degree that most other major economies had. The basis for setting the target range in 1976 (and subsequent years) was, quite explicitly, an allowance for reasonable growth in real output, with some downward pressure on inflation.

M1 was preferred on the basis of empirical evidence that it was well-related to GNE and responsive to short-term interest rates. The Bank of Canada clearly saw short-term interest rates as the way in which M1 would be influenced by monetary policy. Indeed, interest rates generally were seen as the channel through which monetary policy affected the economy. There was less thought of money as an intermediate step in some transmission mechanism than as an indicator to help in the setting of monetary policy. Freedman (1981, 1983), for example, described the monetary targeting process as a feedback rule for interest-rate setting.

The Bank also interpreted the monetary growth rate in the context of broader economic developments. Thus in 1978, monetary policy tightened, not to contain the growth of M1, which had come in at the bottom of the target range in 1977, but to moderate the fall in the Canadian dollar against the U.S. dollar. With the close trading relationship between the U.S. and Canada, interest rates in the U.S. and the Canadian – U.S. dollar exchange rate have tended to weigh heavily in the setting of Canadian monetary policy.

In successive years, the target for M1 growth was gradually reduced, mostly by one percentage point each year. A band, four percentage points wide in most years, was preferred over a point target.

The Canadian authorities were, by the standards of most other countries, remarkably successful in bringing down the rate of growth of M1 in line with the targets. From 1976 to 1980, an outcome within the target band was achieved five times; the other occasion was an undershoot by 0.1 of a percentage point. Although the target band was quite wide, more often than not the outcome was in the lower half of the band.

Early on in the targeting period, inflation declined. But at the end of the 1970s, Canada, with most countries, saw a resurgence of inflation; growth of the GNE deflator rose to around 10 per cent in 1979–82.

At the same time, the trend in monetary growth was downwards. The relationship between money and prices or nominal GNE deteriorated noticeably in this period. During the late 1970s and early 1980s, a number of innovations had been affecting M1. Cash management packages had become available to larger enterprises; this reduced their demand for working balances. In 1979 banks began to offer savings accounts with interest computed daily; this provoked some (modest) flight out of M1. In these earlier episodes, the Bank of Canada felt able to adjust the monetary targets to take account of the shifts. But in 1981 the banks began to offer chequeable savings accounts which paid daily interest. This led to a large flight out of M1, sharply reducing its growth to below both the target rates and those which would have been expected on the basis of past relationships, from mid 1981 to late 1982. Taking M1 at face value would have meant an easing of policy at a time when inflation was high. On this occasion, the extent of the adjustment required could not be estimated with confidence, and the Bank of Canada abandoned monetary targeting.

In the post-targeting period, growth of M1 has been quite volatile, and at times high. But, on the whole, Canada has been quite successful in this period in reducing inflation to very moderate levels; by 1985, it was down to 3 per cent, and has remained around that level since.

(c) France

The French authorities began to focus increasingly on monetary aggregates in the early 1970s. In common with most countries, the background to this was rising inflation and monetary growth. By 1975, monetary growth in France had reached around 20 per cent per annum, and the inflation rate around 15 per cent.

From 1973 to 1976, the authorities operated with internal targets for the broad monetary aggregate M2;^[4] from 1977, targets were announced publicly.

M2 was chosen as the target aggregate initially and maintained over most years because it was considered best suited to the method of monetary control in France. M1 was not favoured because it did not include all transactions balances. The monetary base was found not to have a sufficiently close relationship with the broader aggregates M2 or M3. M2 had the further advantage that it fitted well into the counterparts analysis which was favoured by the Banque de France (see Raymond (1983a)).

The French authorities had a range of instruments at their disposal for monetary control. The established practice of the Banque de France in its market operations was to meet any volume of liquidity demanded by the banking system, while setting the price of liquidity – the “intervention rate” – which in turn provided influence over deposit and lending rates. Reserve ratios were in place on banks and there were also direct controls over selected bank interest rates.

But interest rate changes were not, by and large, used specifically to achieve monetary control. Until very recently, monetary control has relied primarily on the imposition of credit ceilings on banks (the encadrement du credit). There is a long history of quantitative targets in France, in the form of credit controls. These had been used on several occasions through the 1950s and 1960s, and had come into force again at the end of 1972.

The focus on bank credit expansion as a control to monetary growth was in one sense appropriate, given the structure of the French economy, with very low rates of internal funding by firms, and limited sources of equity and long-term debt finance. Financial intermediaries' lending policies were, then, the major determinant of the amount of credit extended and, similarly, on the liabilities side there were limited substitutes for bank deposits.

The encadrement was an effective system for restricting bank credit. Each bank was required to meet the target growth rate for the year by complying with a series of monthly “norms”. Penalties for overshooting were severe and involved the lodgment of additional reserves with the Banque de France at an increasing rate. There was also an implicit penalty for undershooting, in that any shortfall could not be recouped in subsequent years, which meant loss of market share. In response to these constraints, banks became effective in controlling their credit growth.

Only a proportion of bank credit was publicly controlled, however. Lending to selected sectors (agriculture, exports, housing and certain industries) was not subject to strict credit ceilings.^[5] The French authorities'approach was to make annual forecasts for the change in each of the counterparts of M2 other than controlled bank credit, and based on these forecasts, to set ceilings for the growth of bank credit which were consistent with the target for M2. If the forecasts were wide of the mark, the target for money was likely to be missed.

Following the introduction of credit controls and internal monetary targets in the early 1970s, M2 growth was lowered to around 13 per cent by 1977 and the inflation rate came down to about 10 per cent. Between 1977 and 1980 the targeted growth rates for M2 were gradually set lower and over this period the growth of M2 did, on the whole come down, to around 11 per cent in 1980, though the deceleration was not monotonic and money overran the targets on a cumulative basis. Inflation tended to rise again, particularly from 1979, as it did in other countries.

The Mitterand government adopted generally more expansionary policies in 1981 and early 1982, and the target growth rate for money was actually raised between 1981 and 1982. In the event, however, this was undershot, because policy reversed course sharply in the middle of 1982.

This tightening of the stance of monetary policy was part of a re-orientation of policies generally against inflation, which had reached around 14 per cent, and a deterioration in France's external position. The policy response involved, in addition to monetary tightening, fiscal consolidation, exchange rate realignment and a prices and incomes policy.

In subsequent years, the target growth rates for money were lowered quite sharply, to 10 per cent in 1983 (revised to 9 per cent during the year) and 5½ – 6½ per cent in 1984. For 1984, the target was expressed in terms of M2R, which excludes non-resident bank deposits. The outcomes were fairly close to these targets.

As in most cases of quantitative rationing, restrictive policies encouraged the growth of the unregulated areas, and when overshooting of money growth occurred, it was primarily due to underestimation of the uncontrolled counterparts, in particular unrestricted credit. In 1986, the authorities removed the credit ceilings, with their associated monthly “norms”, which had been in force since 1972. While there was still an emphasis on controlling bank credit as the major counterpart to monetary growth, the new arrangements had less of the inflexibility for individual banks that was inherent in the previous system, and represented a step in the direction of more market-oriented control techniques. By early 1987, the Banque de France had moved much further in this direction, with the complete removal of quantitative credit controls and much more emphasis on short-term interest rates as the monetary policy tool. (For further discussion of these issues, see OECD surveys.)

The reforms of the French financial system encouraged in particular the development of markets for both government and commercial paper, and the share market. Part of the sharp lowering of money growth after 1983 was consequently due to the effects of disintermediation, as households and firms increased their holdings of non-monetary assets. The stability of established demand for money functions was weakened by this process and in 1985 the monetary aggregates were redefined by the French authorities, on the basis of functional as opposed to institutional criteria, to take account of recent innovations. Targeting continued with the newly defined aggregates but, in view of the substitutability between financial assets, the authorities adopted dual targets for M2 and M3 from 1987.

In terms of the inflation objective, the mix of policies was broadly successful in the 1980s. Inflation did decline, to around 3 per cent by 1987. Monetary growth declined pari passu.

How much of the decline is due to reasonably successful prosecution of monetary targeting per se is, of course, unclear. The discipline of the EMS, and prices and incomes policies, no doubt played their part, while the effect of disintermediation masked the underlying rate of monetary growth.

(d) Italy

The Italian experience with targeting is unique in that, until relatively recently, the focus was exclusively on credit. Thus if monetary targeting were defined purely as the pursuit of monetary aggregates, much of Italy's experience would be excluded. But it is more constructive not to draw the distinction so sharply, and we consider Italy's use of credit targets.

The commencement of Italy's experience with credit targeting is usually dated as April 1974. The economic background against which the introduction of targeting credit took place was one of rapid price inflation, strong growth in output, rapid expansion of the financial system, a very large public deficit and a tendency toward balance of payments and currency crises. As part of the conditions attached to an IMF standby facility, Italy agreed to limit domestic credit expansion to grow by 22.4 billion lire (around 17½ per cent) in the year to March 1975.^[6]

In public announcements of targets, the Italian authorities initially favoured total domestic credit (TDC), which includes bank credit, bond issues by the private sector and the public sector's deficit. The choice of TDC, as opposed to a monetary aggregate, was not only because the IMF traditionally favoured a credit concept for analytical and policy purposes. It was also because of a perceived high degree of substitutability between money and non-money assets, not least because money (other than currency) was interest-bearing (see OECD (1979)).

The principal control technique for pursuing the credit targets was direct ceilings on bank credit expansion. Banca d'Italia set ceilings for growth in bank credit to the private sector so as to be consistent with the target for growth in TDC, given forecasts of the public sector's deficit. The target rate of growth applied to each individual bank. There were monthly objectives within the annual targets, and exceeding the targets carried the penalty of having to lodge with Banca d'Italia additional reserves on a sliding scale which became quite punitive as overruns grew bigger. In addition, there were portfolio restrictions on banks which effectively forced them to take up government debt, as well as variable reserve ratio requirements.

Monetary policy was not directed exclusively towards internal objectives. In fact, monetary policy tightenings were usually at a time of balance of payments deterioration and downward pressure on the lira, when there were real risks of a depreciation-inflation spiral (especially given the prevalence of wage indexation in Italy). The direct controls over banks were supplemented by other direct controls on foreign currency transactions. For example, an import deposit arrangement which was in force on a number of occasions forced importers to lodge a proportion of the relevant funds with Banca d'Italia for several months.

In the first year the TDC target was met. However, the following year saw a large overshoot and in the remaining years of the 1970s there was little success in meeting targets. Perhaps the principal factor behind this was the persistently large public deficit in Italy, which averaged 8 per cent of GDP in the 1970s. Not only were the deficits large, but the forecasts for the deficit were systematically exceeded. During the years of targeting in the 1970s, only in 1975 did the PSBR come in under the expectation.

Given the strategy of setting ceilings for bank credit on the basis of the expected public sector deficit, bank credit to the private sector did not automatically decline to make way for the public sector's greater than expected requirements. Indeed, control over private sector credit was itself only slightly more accurate than was control over public credit.

Nor was there any lasting success in reducing inflation. The rate of growth of the CPI fell sharply in 1975 and early 1976, from a peak of over 25 per cent to a low of around 12 per cent, as the economy endured a severe recession. But the recovery in 1976 was very strong, and inflation rose again to over 20 per cent by early 1977. Inflation temporarily declined again in the economic slowdown in 1977–78, but was back at 20 per cent by 1980.

In the early 1980s, there was a marked change in the policy approach. Italy, like other countries, moved towards more market-based means of implementing monetary control. Intervention techniques of Banca d'Italia in government securities markets were refined and Banca d'Italia was freed of its former obligation to purchase government securities in excess of the market's demand. This contributed to upward flexibility of interest rates and real interest rates rose to historically high levels. Over the rest of the 1980s, nominal interest rates at times came down, but were kept well above the rate of inflation, so that real rates were consistently positive. This tight (at least by previous Italian standards) monetary policy was supported by incomes policies which effectively provided for less than full indexation of wages.

It could not be claimed that Italy was noticeably more successful in achieving targets for credit in this period. Public sector borrowing continued to exceed forecasts. However, the restriction on credit to the private sector meant that it grew, on average, more slowly than credit to the public sector. By the mid 1980s, the share of private credit in the total had declined sharply. At the same time, the development of government securities markets and increased take-up of government debt by the public were changing the relationship between the development of credit and the public's holdings of monetary assets. The focus of the target shifted to include targets for M2 and credit to the private sector from 1985; the targets for TDC were subsequently dropped. Generally speaking, the growth of money and private sector credit has declined since then.

While there was little progress in fiscal consolidation in the 1980s, the combination of monetary and incomes policies, together with the absence of external price shocks and lower world inflation generally, saw a decline in inflation from a high of around 20 per cent in 1980–81 to about 5 per cent by 1988. Real output, after a prolonged period of weakness from 1981 to 1983, has grown continuously at moderate rates since then.

(e) Japan

The Bank of Japan began to focus internally on monetary growth around 1975,^[7] after a large increase in inflation – to over 20 per cent – in 1974, which had been associated with rapid monetary growth.

A tightening of monetary policy at this time had quick results – there was a marked slowing in economic activity, and inflation fell to 5 per cent by the end of 1975. Economic recovery was subsequently able to proceed, though growth was lower than had been seen in the 1960s and early 1970s. Thus, by the time the Bank of Japan began to announce publicly quarterly “forecasts” for growth in M2 (changed to M2+CDs from 1979) in July 1978, reasonable stability in prices and output had already been restored.

The forecasts were seen as contributing to improved public understanding of the effect on the money stock of measures being taken to liberalise financial markets in Japan, and in particular, the likely impact of a new bidding system for selling the large volume of government securities that was required to fund an expanded fiscal deficit.^[8]

The method of monetary control in Japan has been a combination of central bank market operations aimed at inter-bank interest rates and “window guidance”, a form of “suasion” encouraging banks to alter their volume of private lending. The latter was particularly important in early years of Japanese development, when corporations needed large volumes of external finance and banks were the most important source of that finance. In more recent years, the exercise of control over market interest rates has become increasingly important as Japanese financial markets have been liberalised, corporations have become more self-funding and more aggressive in funds management, and the relative importance of banks has declined.

The transmission process for monetary policy is now seen as operating on three levels. The first is through bank portfolio adjustment (substitution between lending in the inter-bank market and lending to the non-bank public); the second is through portfolio shifts by the non-bank private sector (substitution between open market securities with flexible interest rates and bank deposits with largely regulated interest rates) – “disintermediation”; and the third is the traditional interest rate-expenditure link.

The “forecasts” of the Bank of Japan have some differences with the targets generally set in other countries. Forecasts are announced in the first month of each quarter, and refer to the growth of M2+CDs between the current quarter and the same quarter of the previous year. This means that while the forecasts are of annual growth in the money stock, they are announced after three quarters of the year have passed. Events in the final two months of the fourth quarter will have a relatively small weight in determining the outcome. Consequently, the outcomes have invariably been very close to the forecasts.

The announcements are also carefully referred to as “forecasts” rather than “targets” and the forecasts are specified rather approximately, either as a range of outcomes or as “about x %”. The Bank of Japan considers that limiting the quantitative announcements to short-term forecasts is the most appropriate approach to take, given the lack of precision with which monetary growth projections can be made and the need to be flexible in response to changing circumstances. Also, the Bank of Japan has made quite clear that while it emphasises the money stock, it is “not blind to other indicators” and that the money stock is considered in an overall framework which includes prices, output, the balance of payments, interest rates at home and abroad, movements in the foreign exchange market, and attitudes of financial intermediaries to lending.^[9]

M2+CDs has been the focus of monetary control because of empirical evidence that suggests that it has the closest relationship with future income and expenditure. M1 was found to have the best correlation with current income, but causality seemed to run from income to money.

The recognition of long lags in the transmission process was also the basis for the Bank of Japan seeking only to control money supply growth averaged over long periods. Short-term deviations from desired growth paths were not considered to be in need of correction.

The initial quarterly forecasts in 1978 did not envisage any reduction in monetary growth, with inflation already low at 5 per cent. However, with new inflationary pressures from the second oil price shock, monetary policy was progressively tightened in Japan over 1979 and the first half of 1980. Forecast and actual monetary growth were reduced.

In 1986 economic growth slowed in Japan, under the weight of a strong appreciation in the yen, and the Bank of Japan moved to ease monetary conditions. By the end of 1987 monetary growth had risen to its highest level since the second oil price shock and this provoked some debate in Japan over whether it had been excessive. The Bank of Japan undertook an examination of this issue and concluded that the demand for money had been higher due to historically low interest rates and faster growth in income associated with economic recovery, but also that there seemed to be a shift in the money demand function resulting from continued financial liberalisation.

Superficially, the Japanese experience with monetary “targeting” could perhaps be regarded as something of a model of monetarist practice. Certainly, the simple results speak for themselves. Since 1975, monetary growth has been lower on average and has been more stable than in the 1960s and early 1970s. Inflation has fallen to negligible proportions. Real GDP has grown fairly consistently since the mid 1970s, though at rates considerably below those seen in the 1960s and early 1970s.

Suzuki (1985), however, attributes this performance to the close tracking in Japan of expectations of inflation to the actual path of inflation and the flexibility of real wages. He argues that with fewer “mistakes” in expectations of inflation, and rapid adjustment of wages accordingly, the Japanese economy “has been gradually moving down along the long run Phillips curve” (see Suzuki (1985) p 4).

In terms of practice, the Japanese approach has been some distance from strict adherence to a pre-determined money growth rule. This is openly acknowledged by the Bank of Japan, which regards its approach as pragmatic.

(f) Switzerland

The decision of the Swiss National Bank in early 1975 to adopt a target for M1 reflected concern over inflation in Switzerland, which had risen to around 10 per cent by 1974, and a perception of a medium-term relationship between M1 and the price level. The latter was described by Schiltknecht (1983 p 72) as “the casual observation that there had been a fairly close relationship between M1 and the consumer price index and that growth rates of M1 had fluctuated around 3 per cent during periods of stable prices”.

At the time of the announcement of the first target for M1, the Swiss economy was contracting after very tight policies over 1973–74 to fight inflation. Real GDP fell by 5 per cent in 1974. Monetary policy eased substantially in 1975, and the target for M1 growth was set at 6 per cent for the year, which was noticeably higher than the outcome for the previous year (though not higher than the likely rise in nominal GDP).

For the first few years of targeting, until 1978, the actual growth of M1 was close to target on a cumulative basis, though there were over or underruns in individual years.

The aim of policy was clearly to contain inflation and this was achieved with the increase in the CPI down to around 1 per cent in 1977. However, while the targets for money allowed moderate growth in M1, Switzerland appears to have had a poorer real growth performance than most economies in the mid 1970s. Real output fell again in 1975 (by 3 per cent), recovered slightly in 1976, then fell again in 1977 and showed little growth in 1978. Real GDP remained below its 1974 level until 1980.

While inflation was the medium term objective, the Swiss National Bank was prepared to depart temporarily from monetary targets when pressures on the external front intruded. In 1978, with the Swiss Franc under upward pressure, the SNB suspended the target for M1 and adopted an exchange rate target instead (see Schiltknecht (1983)).

The resulting intervention in the foreign exchange market, which was on a massive scale, was associated with an enormous increase in M1 (over 20 per cent on a year-ended basis in December 1978) and a decline in domestic interest rates (some rates were below zero in the final months of 1978 and early months of 1979). Even so, the real effective exchange rate rose significantly (about 30 per cent), and contributed to fairly weak performance in output for the year.

No monetary target was set in 1979. By 1980, the exchange rate situation had stabilised, and monetary targeting resumed, this time with the monetary base as the focus. Switzerland is unique in that it is the only country which adopted a procedure of monitoring the monetary base and forecasting the “money multiplier” as a control method for its intermediate monetary target. After the experience of the late 1970s with exchange rate fluctuations, however, the SNB concluded that exchange rate expectations affected the demand for M1. The forecasts of the money multiplier between the base and M1 had become increasingly unreliable apparently, for this reason. The SNB therefore decided to target the monetary base itself, though the target is still derived with reference to desirable growth in M1 (see Schiltknecht (1983) pp 74–75).

The targets have been undershot more often than overshot during the 1980s. The monetary base declined in both 1980 and 1981, compared with a target of 4 per cent growth in each, as the SNB sought to reverse the monetary expansion of 1978–79, and to counter rising inflation (much of which came through imports after the second OPEC oil price rise).

For most of the period since 1982, growth of the base has been remarkably stable and close to target in each year. As the OECD survey for 1987/88 noted, Switzerland has not, for the most part, been affected by deregulation, which disrupted the demand for money in so many other countries, and since 1980 has not been subject to large external shocks which complicated policy-making in the 1970s. In 1988, the monetary base fell well below target, but this was a reflection of a decline in required reserve ratios, during the year.

The achievements on inflation are remarkably good. In the wake of the second OPEC shock, inflation measured using the CPI reached only about 7 per cent, and since 1983 has averaged 2 per cent.

(g) United Kingdom

The U.K's experience with quantitative targets really began in the mid 1960s, when balance of payments difficulties led to temporary funding from the IMF, with a provision that targets for domestic credit expansion be put in place.

As a result of this, the U.K. started the 1970s with restrictions on bank lending in place. But with the changing intellectual climate, there was an increased focus on money, and the Bank of England took the first steps towards monetary targeting in the early 1970s. Money was used as an internal guide for policy from about this time, and the Bank of England prepared to move away from direct controls over the banks.

Early on, there was apparently considerable confidence that interest rates could be used to influence the growth of monetary aggregates (see Richardson (1978)), and as Fforde (1983 p 53) notes, monetary policy “remained along with fiscal policy essentially ‘Keynesian’ in outlook”.

The competition and credit control arrangements were removed in 1971, facilitating an increasing role for liability management by banks. Growth in M3 increased sharply, reaching rates of 25 per cent in 1972 and 1973. This was associated with expansionary fiscal policy and rapid growth in real output, and was followed by a large increase in the rate of inflation.

Through this period, interest rates moved upwards, but this, of itself, was not sufficient to stem the boom. Eventually, fiscal policy had to be tightened and direct controls were reimposed on the banking system.

Although there were already doubts about the relationship between M3 and nominal GDP, the focus of public attention on M3 growth, uncertainty over the overall economic situation, particularly in the foreign exchange market, and flagging public confidence in policies generally, constituted a powerful case for a public announcement of an M3 target (see Fforde (1983)); the announcement was made in July 1976.

It is clear from the U.K. literature that fiscal policy and monetary expansion are viewed as closely linked. This is consistent with the long-standing focus on M3, where the “counterpart” type of analysis shows the accounting identities which link the size of the public deficit, sales of debt to the public, expansion of bank lending and growth of deposits. (The use of the sterling component of M3 as the target from 1977 was a refinement, so as to provide a closer link to domestic credit expansion, targets for which were again in force by that time under an IMF facility.) It is clear in this context how the monetary targets could be referred to as providing “the framework of stability within which other policy objectives can be more easily achieved” (Richardson (1978) p 34).

But the targets were not adhered to rigidly. After an initial undershoot in 1976, monetary policy was eased in 1977 because of exchange rate considerations and the monetary target was overshot by a large margin. The Bank of England was also still inclined to see a countercyclical role for monetary policy (e.g. Richardson (1978)).

The Conservative government elected in 1979 introduced the Medium Term Financial Strategy, which provided for gradual reductions in £M3 growth over a run of years, in the context of a concurrent reduction in the public deficit. At the same time though, exchange controls were abolished, which seriously undermined the effectiveness of direct controls on bank lending, since foreign sources of credit were now available to U.K. residents. The removal of the “corset” in 1980 saw a rapid increase in £M3 as “reintermediation” took place. Bank credit expanded rapidly, as banks moved into mortgage lending. In both 1980 and 1981, £M3 overshot its target, even as real activity weakened sharply and the inflation rate fell.

The disruption to the relationship between £M3 and nominal GDP proved not to be temporary. Forward targets for monetary growth originally set down in the Medium Term Financial Strategy in 1980 were revised, and the status of some of the targets downgraded to that of “illustrative ranges”. While reasonable control was exerted over £M3 in 1982 and 1983 by the policy of “overfunding”,^[10] the authorities were sufficiently uncertain about the validity of £M3 that in 1982 and 1983, supplementary targets for M1 and PSL2^[11] were announced in addition to the targets for £M3.

Similar problems occurred in the mid 1980s. £M3 recorded a large cumulative overrun of its targets in 1984, 1985 and 1986. This rapid growth was not reflected, however, in nominal income, and inflation remained at around the 5 per cent level it had reached in 1983. During this period, it is clear that once again the targets were considered in the light of all the other economic evidence. The Governor of the Bank of England confirmed this in a 1984 lecture (see Leigh-Pemberton (1984)).

Even so, announced targets still remained in place. From 1986, targets for M0, a measure of the monetary base which was thought to be more stable in its relationship to the economy than M3, as well as to lead the economy, were announced, and after 1986, the targets for M3 were dropped. In both 1986 and 1987, the targets for M0 were achieved. In 1988, M0 was running substantially above target, as the U.K. economy had grown rapidly. A pronounced tightening of monetary policy had been put in place, the full effects of which were yet to be seen at the time of writing.

(h) United States

The U.S. Federal Reserve had been following internal “tolerance bands” for money and credit since 1970.^[12] Open market operations took account of deviations of the aggregates from paths thought to be desirable from a medium-term perspective. But money and credit took their place alongside other factors, such as “credit conditions”, and the U.S. dollar exchange rate.

From 1975, the Fed reported publicly target ranges for the monetary aggregates. This was in response to a resolution passed by the U.S. Congress calling for the Fed to report quarterly on its monetary intentions for the ensuing year. This effectively meant that the targets were in principle “updated” each quarter, though in practice they did not actually change from quarter to quarter. Later, the “Humphrey-Hawkins” Act (1978) required the Fed to testify twice yearly to Congress, and announce its monetary targets.

The extent to which the targets gained additional weight in monetary policy decisions has been questioned. While the Fed clearly viewed the federal funds rate as an instrument with which to exert influence over the monetary aggregates, the operating procedures in place at this time (for a description see OECD (1979)) still prevented there being too much short-term movement in the funds rate. As a result, it has been suggested that while funds rate movements were usually in the direction consistent with returning the monetary aggregates to their targeted paths, they were usually insufficient in size to do so. This could be, and was, defended on the grounds that any given movement in the money stock may be noise, or a shift in the money-income relationship, in which case a monetary policy reaction is not called for, or a genuine piece of information requiring a response. In such cases, it was argued, a sensible reaction would be to steer a middle course: resist the movement in money, but not to the extent that unswerving adherence to the target would suggest. It is also clear that the targets had not been intended by the Congress as rigid constraints on the Federal Reserve (see Axilrod (1983) p 33 for quotation of the relevant legislation).

Nonetheless, targets were in place. At the time of their introduction in March 1975, monetary growth had already been brought down considerably from levels of the early 1970s. The initial targets, if anything, allowed for a slight increase in monetary growth. This was against the background of a contraction in output of 2½ per cent in 1974 and some moderation in inflation. The Fed opted for multiple targets, with ranges for growth in M1, M2 and M3, as well as a measure of credit, being announced. At this stage, however, M1 was the focus of most attention.

In 1975 and 1976, M1 growth fell within the target range. M2 and M3 grew only slightly faster than their targets in 1976, while credit grew slower than its target in both years. Output growth recovered rapidly and inflation fell.

In 1977, with the economy strengthening, the monetary aggregates tended to be at or above the top ends of their respective growth ranges. The U.S. dollar was also under downward pressure at this time, and from the middle of 1977 through 1978, nominal interest rates rose. Growth in the monetary aggregates slowed in 1978 and 1979 (though credit accelerated) and domestic output growth dropped away. Inflation, however, rose strongly through this period, especially after the oil price shock and by late 1979, inflation was running at an annual rate of around 15 per cent and set to rise further. Inflationary expectations were firmly entrenched.

October 1979 is frequently regarded as a watershed in the conduct of U.S. monetary policy at the operational level. The Fed announced (somewhat dramatically, on a Saturday) a switch from the federal funds rate to the non-borrowed component of bank reserves as the operational target of monetary policy. Henceforth, this particular quantity would be monitored much more closely, and policy reaction would be larger and faster than previously, with the federal funds rate free to move in a much wider band.^[13]

Following this, interest rates were indeed much more volatile and also on average much higher. The standard deviation of the quarterly change in the U.S. 3-month commercial bill rate for the period spanning the December quarter 1979 to the September quarter 1982 was higher by a factor of 3 or 4 than in the 3-year periods either side.

Despite this greater flexibility in interest rates, the Fed was not noticeably more successful in achieving monetary targets. In 1980, all the monetary aggregates overran their target ranges; in 1981, M1 ran under its target while M2 and M3 ran over.

Monetary growth was also more volatile in this period. This was the focus of considerable criticism of the Fed. McCallum (1985), for example, argues that the operating procedures adopted by the Fed were not well designed, and that both interest rates and monetary growth were more volatile than necessary.

But whatever the criticism, monetary policy was sufficiently tight that the U.S. economy slowed sharply during 1981–82, and inflation started to decline.

About this time, the effects of financial deregulation and innovation were being felt. The U.S. experience in this area is rich, and far too complex to be covered in detail here. It included the growth of money market mutual funds (MMMFs) since the late 1970s; the January 1981 introduction of Negotiable Order of Withdrawal (NOW) accounts (interest bearing and “chequeable”); the December 1982 introduction of money market deposit accounts (MMDAs) allowing banks and thrift institutions to offer services which compete with the MMMFs; the January 1983 introduction of super-NOW accounts, with no interest ceilings; and the general spread of cash management services offered by banks and brokers, including services such as “deposit sweeping” (automatic, often daily, transfers between transactions and investment accounts). (For a useful summary of these issues, see the OECD survey of the U.S. economy published in December 1983.)

These sorts of innovations blurred the distinction between “transactions” balances and “savings” balances: the one account could now have both characteristics. This complicated the task for monetary policy, since NOW accounts could attract funds from non-interest transactions deposits or from other interest-earning assets (e.g., savings deposits) included in M2. To the extent that they came from the former, the demand for M1 in the aggregate would be relatively unaffected, as NOW accounts were included in the redefined M1. Thus, in targeting M1 growth virtually no adjustment would have been needed. However, to the extent that they came from the latter, the demand for M1 would have been increased and an “easier” M1 growth target which simply accommodated the increased demand for M1 would have been appropriate. It was thus difficult for the authorities to judge ex-ante how the M1 growth target should be set or adjusted.^[14] Similar issues arose with the introduction of MMMFs, and subsequently MMDAs and super-NOW accounts.

In the second half of 1982, the confluence of a severe domestic recession, and the emergence of a crisis in international financial markets with the Mexican debt impasse in August 1982, was followed by an easing of U.S. monetary policy. The operating procedures based heavily on non-borrowed reserves were altered, so as to accommodate more fully shocks to the demand for money. At the same time, recognising the technical difficulties of M1, the Fed began to “de-emphasise” it, though targets were still announced.

M1 growth thereafter rose rapidly, but the traditional relationships between money, especially M1, and nominal GNP and/or prices appeared to have shifted. Despite variable and, at times, quite rapid growth in money during the remainder of the 1980s, inflation continued to drift down in the U.S. until as late as 1986, and the velocity of M1, which had trended upwards continually for two decades, reversed course sharply between 1980 and 1987. By 1987, M1 targets were dropped altogether.

The other aggregates continued to be targeted, and the Fed paid more attention to a new credit measure, debt of the private non-finance sector. The record of achieving the monetary and credit targets for M2 and M3 from 1983 was mixed. While targets for monetary aggregates are still announced (if for no other reason than that it is a requirement under law), it is also clear that the Fed looks more broadly than just at the financial aggregates in deciding the stance of monetary policy (see Volcker (1985)).

(i) West Germany

Like other countries, Germany had seen an increase in monetary growth in the early 1970s, accompanied by a rise in inflation.

By 1974, monetary policy had been tightened very sharply. Short-term interest rates were at very high levels, even in real terms. The rate of growth of Central Bank Money, the Bundesbank's preferred measure, had declined from 13 per cent in 1972 to 6 per cent in 1974. Real output growth had slowed sharply.

Against this background, the Bundesbank introduced a monetary target in late 1974 for growth through 1975. As in some other countries which had already seen a sharp slowing in monetary growth as well as a deep recession, the German authorities felt able to countenance an increase in the rate of growth of money over the result for 1974; the target for 1975 was set at 8 per cent.

The preferred aggregate was Central Bank Money (CBM), comprising currency in circulation and banks' required reserves (at constant reserve ratios) on deposits. This was preferred over M1 and M2 because of concern over the effects of portfolio shifts on those aggregates, whereas CBM has more of the characteristics of M3, which internalises those shifts. CBM was preferred to M3 itself because the weightings attached to the various types of deposit (basically the reserve ratios applicable to those deposits) in CBM were thought to be more appropriate than the weights in simple summation aggregates.

The setting of the targets paid attention to the state of the economy and the prevailing rate of inflation. The authorities were quite explicit about this, describing the setting of the target along the following lines. The required monetary growth depended on assumptions made about the rate of inflation, the growth of output, and the change in velocity. On inflation, they had some notion of an “unavoidable” increase in prices. In general, this was related to the rate of inflation in the year immediately preceding, though if that was seen as excessively high, a downward adjustment was made.

The assumption about the growth of output comprised two parts: capacity growth, and the growth due to a change in the utilisation of capacity. The determination of the first was fairly straightforward; a projection was also made of the latter, based on the state of the business cycle. Finally, the change in velocity had a long-term component and a cyclical component which, it was argued, was itself related to the degree of capacity utilisation.

The method of influencing CBM was through the terms at which the Bundesbank made available liquidity to the banks. Though the Bundesbank can change reserve ratios, such changes were not often used as policy tools in the targeting period, and the focus was mainly on “structuring interest rates and supply and demand conditions in the money market in line with … monetary objectives” (Deutsche Bundesbank (1982) p 90).

Following the first year, when a December to December target was missed, the Bundesbank expressed the targets for 1976 to 1978, which were maintained at 8 per cent, on an annual average basis, so as to minimise the effects of fluctuations in the month at either end of the “target corridor”. Even so, the Bundesbank was not particularly successful in achieving these targets.

The reason was essentially that during each year, an assessment of progress compared with the target was made in the light of economic circumstances. Inflation and output were considered; so also was the Deutschemark exchange rate. In 1977–78, for example, CBM growth was allowed to run well ahead of target, as the Bundesbank sought to moderate the rise in the Deutschemark against the U.S. dollar. Such rise as there was in the exchange rate was seen as dampening inflationary pressure.

In the following year, the Bundesbank, for the first time, set a range of target outcomes for CBM designed to give some degree of discretion, with the outcome towards the higher or lower end of the range depending on developments during the year. With exchange rate pressures easing and the second OPEC price shock affecting domestic prices, monetary policy tightened sharply and growth in CBM slowed, coming in at the lower end of the target range. The tightening continued through 1980 and 1981, and CBM growth for both years was at the low end of the target band.

By 1982, real output was quite weak, and there were signs that inflation was coming down. Monetary policy was eased, and CBM growth ran at higher levels, though within the target range. In subsequent years monetary targets were achieved, with generally firm monetary policies, moderate growth in output and declining inflation.

The strength of the Deutschemark against the U.S. dollar in 1986 and 1987 was nominated by the Bundesbank as the key factor in the toleration of above-target monetary growth in those years, against a background of negligible inflation. By 1988, however, there were signs of a resurgence of inflationary pressure, and monetary policy began to tighten. At the same time, the focus of the target shifted to M3 (though given the similarities between CBM and M3, this shift was only of a technical nature).

4. Summary of Experiences

This section draws together what seem to us to be the important aspects of the experiences of the nine countries above, under a number of headings.

(a) Choice of Aggregate and Expression of Targets

It is interesting to note the preponderance of relatively broadly-defined aggregates. Only Switzerland and Canada had exclusively narrow targets; the U.S. gave M1 prominence over the other targets initially, until it was “de-emphasised” in the early 1980s, and subsequently dropped altogether.

The remaining countries adopted targets for aggregates which were broader, usually either on the grounds of internalising portfolio shifts resulting from interest rate changes, or of the usefulness of the counterparts approach to forecasting and analysis. It is no coincidence, of course, that the countries which stressed the latter (such as the U.K., France, Australia) also tended, initially at least, to have direct controls on at least part of the private sector's credit expansion.

Two of the countries abandoned targeting permanently in the 1980s. Of the remaining seven, all have changed the aggregate which is targeted, on at least one occasion. In one or two cases, the changes were of technical significance only (as in the case of West Germany's shift from Central Bank Money to M3). But in most cases, the changes in aggregates were not merely technical, but more fundamental. These were generally made because there had been a shift or weakening in the established relationship between the original aggregates chosen as targets and nominal income and/or prices. As mentioned, this was the reason for the “de-emphasis” of M1 in the U.S.; it also accounts for the shift from M2 to a redefined M3 and then to multiple targets in France; from £M3 to M0 in the U.K.; and, to a certain degree, for the shift from credit targets to the focus on money in Italy. In most cases, the underlying shifts in money demand were the result of financial innovations and changes in behaviour following deregulation. (The only exception is Switzerland, where exchange rate expectations destabilised demand for M1 in the late 1970s.)

It is doubtful whether such changes to aggregates had been envisaged when countries first began to target in the mid 1970s. Indeed, widespread shifts in the aggregate which is targeted arguably undo some of the value of targeting, particularly the announcement effects.

In five of the nine countries, target bands were preferred to single figure targets most of the time. The motive for this was presumably that a single figure was more difficult to hit, while it was easier to fall within a band of reasonable width. On the other hand, another view, held at least by the Swiss, holds that the loss of reputation in missing a band is far more than that in missing a single figure, so the band is not to be preferred (see Schiltknecht (1983)).

(b) Monetary Control Methods

The use of direct credit controls in some countries which tended to have broadly defined targets has already been noted.

Such countries generally have found that the effectiveness of such controls diminished over time, or had undesirable side effects such as freezing market shares of individual financial institutions or actually reducing the share of one group of institutions in favour of another. Over time, these countries, such as the U.K., France, Italy and Australia, have tended to move more in the direction of market operations to achieve monetary control. It is fair to say that the majority of countries now view short-term interest rates, influenced through market operations, as, effectively, the instrument of monetary policy.

Switzerland is the only professed follower of a monetary-base control strategy among the major countries (apart from the U.S.'s brief focus on reserve quantities between 1979 and 1982). What is of considerable interest here is that both Canada and Switzerland were relatively successful in achieving targets, with control techniques that were conceptually quite different in the two countries – Switzerland using the money base-money multiplier approach and Canada using short-term interest rates. Criticism of central banks' reluctance to embrace strict quantity control of the monetary base (e.g. Sumner (1980) and references cited there) may be misplaced if using short-term interest rates works just as well. But it is also possible that the optimal control technique depends upon the institutional framework and the source of disturbances.

(c) Outcomes Compared with Targets

It is conventional in assessing the experience of monetary targeting to compare targets and outcomes. Table 1 presents the data for the nine countries considered in this paper. Altogether (including all the targets in countries where there were multiple targets) there were 140 individual annual targets^[15] over the period from 1975 to 1987. Less than half of these were achieved.^[16]

Most countries also appear to have allowed some degree of “base drift” – that is, the targets set in year t have tended to use the actual outcome in year t−1, rather than the targeted outcome for that year, as their base. It is impossible to be dogmatic about this, since it is not known with certainty what target would have been set had the outcome in t−1 been different. But judging simply by the announced target ranges, it seems that monetary policy did not usually try overtly to reverse target “misses” in following years. The exception appears to be Switzerland, which regained some “lost ground” in the early 1980s after monetary targets were suspended in the late 1970s.

Of course, there can be any number of valid reasons why a target might be missed. Shocks to the demand for money, for example, are conventionally thought of as a defensible reason for allowing monetary growth to diverge from its targeted path.

It is interesting to note, however, that “misses” above targets were far more numerous than those on the low side: overruns outnumbered underruns by four to one. (These statistics give a heavy weighting to the U.S. because it had four targets in almost every year, but interestingly enough, if the U.S. is excluded the outcome is almost identical.) In several countries (Australia, France, Italy and the U.K.), cumulative monetary growth over the targeting period was in excess of what was implied by the successive targets, even taking the top of the range for each year. This was also true for M1 in the U.S., though not for M2 and M3. In some of the other countries, there were cumulative overruns for several years, with money on target at other times.

Perhaps this can be explained by a number of serially correlated shocks to the demand for money. This could not be ruled out, particularly during the deregulation of the 1980s, when many countries appear to have experienced money demand shocks. But that cannot be the explanation for why Germany or the U.S. overran targets in the late 1970s, or why Italy exceeded its targets in virtually every year.

There were other technical factors which were important. In a number of countries, procedures for monetary control were not well developed. Several countries did not have sufficient flexibility in interest rates to prevent the funding requirements of budget deficits impinging on monetary growth. The set of parameters within which the projections or targets were framed were frequently unreliable (Italy and France, for example). Foreign exchange inflows proved to be problematic for some countries with relatively inflexible exchange rates (Australia until the end of 1983). Countries with direct controls saw the effectiveness of those controls eroded (such as France, Australia).

But the answer as to why targets were not met more frequently is only partly to be found in technical factors which distorted the demand for money, or rendered control of the money stock difficult, important as these were in individual cases.

The other part of the answer is the way in which central banks approached targeting. It is doubtful that central banks themselves saw monetary targets as monetary rules – that is, as binding policy rules to be achieved at all costs. All central banks reserved the discretion to diverge from targets if circumstances were deemed to require it. Thus the Bundesbank and the Swiss National Bank permitted rapid monetary growth in the late 1970s because of exchange-rate considerations; the Bundesbank did so again in 1986 and 1987; the U.K. did so in 1977. Given that monetary targets were not seen as monetary rules, comparison of targets and outcomes on its own is inadequate.

Ultimately, it may be that the assessment of monetary targeting as practiced cannot be divorced from the assessment of monetary policy generally. In particular, whether monetary policy in the targeting era contributed to acceptable macroeconomic results may be a more important question than whether particular targets were achieved. The only objective measure of that performance is observed results, for the intermediate and ultimate objectives of policy.

(d) Outcomes for Intermediate Objectives

On the intermediate objectives, this paper started by suggesting that there were to be three principal features of monetary targeting: money growth should be low, it should be steady and targets should be publicly known in advance.

The third of these criteria has obviously been fulfilled in each of the countries studied here, (although the contrary case could perhaps be argued for Japan because of the relatively short period which monetary growth projections cover).

Did money growth slow down under targeting? Table 2 presents some results. Compared to the period from 1971 to the commencement of targeting, the average rate of growth of the monetary aggregates which were the focus of monetary targets did decline in the targeting period. The only exception was M1 in the United States, where growth after targeting commenced and up to 1984 (this excludes the very high rates of growth in M1 in 1985 and 1986 while the distortions to the statistics were at their worst) was slightly higher than the average from 1971 to 1974.

But the early 1970s is in many respects a favourable point of comparison, since monetary growth rates were unusually high for part of that period. If we compare average growth rates for money in the targeting period with the corresponding figures from the 1960s, we find that only in West Germany, Japan and Switzerland was growth of the money stock lower on average in the targeting period, while for Australia, Canada, the U.K. and the U.S. it was higher.^[17]

The variability (measured as the standard deviation of the quarterly percentage change) of money declined in the targeting period compared with the period from 1971 to the commencement of targeting in five of the eight countries for which data were available for a long enough period to make a comparison. Of the three exceptions, the U.S. saw an increase in the variability of M1, though this was affected by the big shifts in M1 in the 1980s. Variability of M3 did decline in the targeting period. The remaining exceptions were Switzerland and Canada, which, perhaps paradoxically, were the most successful in achieving targets. Again, the early 1970s is a favourable comparison point. The only country which achieved variability in monetary growth lower than in the 1960s was Japan.

(e) Outcomes for Final Objectives

So much for the intermediate objectives. Did the targeting period see an improvement in the “ultimate objectives” of monetary policy, such as inflation?

Most countries which targeted saw inflation fall during the targeting period. The most remarkable case was Japan, where in the period since “targeting” began, monetary growth and inflation have both averaged a rate 7 percentage points lower than in the period from 1971 to the commencement of targeting.

The two exceptions make an interesting pair. Italy saw inflation higher, on average, during the targeting period (though to be fair it has made a lot of progress in the mid 1980s in reducing inflation) and was repeatedly unable to prevent its target from being exceeded. Canada, on the other hand, was able to achieve its target range on almost every occasion (and the only “misses” were when monetary growth was lower than targeted), but had no success in reducing inflation. In fact, Canada only achieved a significant decline in inflation after targeting had been abandoned because of instability in the demand for M1. Australia is also an interesting case in that inflation has been lower on average after monetary projections were discontinued than in either the targeting or immediate pre-targeting periods.

It is also apparent from the data in Table 2 that the variability of prices was lower in the targeting period than just before it in most countries. Compared to the 1960s, however, inflation was still higher for the targeting period on average for five of the nine countries.

These averages disguise, of course, a significant reduction in inflation in all countries since 1982. By 1987, inflation had declined to levels not seen since before the acceleration in prices in the early 1970s. There can be little doubt that tight monetary policies helped to bring this about – the level of real interest rates in all countries, and the depth of the 1982 recession, in the U.S. in particular, testify to that. In countries such as Italy and France, where there has certainly been much slower monetary growth in line with targets in this period, the additional discipline of the European Monetary System may have added to the discipline of monetary targets.

But the world economy has not been subject to any large price shocks in the 1980s, so policies have been able to work in a more favourable environment. The weakness of oil and other commodity prices in 1986 has doubtless helped overall inflation performance, independently of monetary policy.

5. Conclusion

Clearly, monetary targeting, as adopted in one form or another by a number of countries, has coincided with a decline in rates of inflation from those that prevailed prior to the introduction of targeting in the mid 1970s. How much the use of targets themselves contributed to that decline is less clear.

That monetary authorities had a revealed preference for the announcement of targets in the mid 1970s suggests that they perceived a role for monetary aggregates both as indicators to be used in the policy-making process and as benchmarks which could play a role in anchoring expectations and against which they were prepared, to some extent, to be judged. There is also more than a hint of an implicit discipline on fiscal policy in the announcement of targets for monetary growth in some countries.

But several facts point to monetary targets as being, at least from the practitioners' points of view, an appropriate tactical response to a set of circumstances, and not a complete revolution in the execution of monetary policy. Targets were never interpreted as monetary rules (that probably would be a revolution). Overruns were not usually reversed; neither were underruns, but they were far less frequent. The aggregates concerned were switched when considered necessary, and the targets themselves were ignored at times when other considerations intruded. Some countries eventually abandoned targeting altogether.

Much of the progress in reducing inflation in a number of countries occurred in the 1980s, long after the major countries started targeting, and only after a serious world-wide recession in the early part of the decade. The decline in inflation was also helped with very weak prices for oil and other commoditites in world markets.

Of course, a number of countries still have monetary targets in place. But equally, most of those countries either explicitly follow a strategy based on assessment of a range of indicators or have additional monetary policy objectives such as exchange rates. This suggests that the role of monetary targets is one of being one part of a more widely-focussed approach to making monetary policy.

The difficulties in using monetary aggregates are well-known, and may be sufficient justification for not making monetary targets the sole or even the primary focus for monetary policy, and may even be reason enough to not have an announced monetary objective at all (though many would of course dispute that.) The nagging question is, with what is such an objective to be replaced? Few would argue for an interest rate rule (though it is not hard to see a role for interest rates as operating objectives in a policy designed to stabilise nominal GDP – see Edey (1989)).

In recent years, a number of writers have supported making nominal GDP the direct target for policy (for example, Tobin (1980), Meade (1978), McCallum (1984)). The principal advantage of a nominal income target for monetary policy, as opposed to a monetary target, is that it prevents monetary policy being blown off course by unanticipated shifts in money demand.

There are, of course, some practical difficulties with this. First, unless there are good advance indicators the authorities will not have information about income or prices until after the event. Second, the information may be subject to a substantial margin of error (i.e. data are subsequently revised). In practice, of course, policymakers frequently make decisions on the basis of partial indicators which are more timely than GDP statistics. But these rarely all give the same message, and the authorities will inevitably react with some lag. On top of this there will be some lag in the response of the economy to the policy action.

There is also a literature on the issue of exchange rate objectives for monetary policy. An early proposal was by Ethier and Bloomfield (1978). More recently, Williamson (1983) and Williamson and Miller (1988) have received considerable exposure (see also Argy, McKibbin and Siegloff (1988)).

No doubt discussion and development of these ideas will continue. The optimal strategy for monetary policy may depend on the circumstances and may vary from place to place and from time to time. What will be much less variable is the need to judge monetary policy's conduct, in the final analysis, on bottom line results.

Footnotes

* Macquarie University and ** Reserve Bank of Australia. This paper was originally commissioned for the Economic Record, and is forthcoming in that Journal. The cooperation of the Editors of the Record in allowing it to be reproduced here is gratefully acknowledged. [*]
[**]

By monetary targeting in this paper, we mean targets for growth in a measure (in some cases more than one measure) of the money stock which were publicly announced in advance. Thus the internal targets adopted by some countries earlier in the 1970s are not treated here as monetary targets in the full sense. We do, however, include the case of Italy, which for most of the period in question announced targets for credit, rather than money. [1]

This is by no means an exhaustive list of the countries which have used monetary targets. The Netherlands, Spain and Brazil have each pursued a monetary objective at one time or another. South Africa has targets in place at present. But the nine countries listed here had the most persistent use of targets over the period in question, and there is more material on which to draw to assess the experience. [2]

See for example Blundell-Wignall and Thorp (1987), Stevens, Thorp and Anderson (1987). [3]

See OECD Survey of France (1987), also Cobham and Serre (1986). [4]

From 1980, a system was introduced whereby favoured lending received a weighting in the credit ceilings, but a lower weight than ordinary lending. [5]

In fact, Italy had imposed ceilings on bank credit prior to this. A limit of 12 per cent had been placed on the increase from March 1973 to March 1974 in bank loans to customers with an indebtedness of more than 500 million lira. There were also credit expansion limits as part of funding agreements between Italy and the EEC, and between Banca d'Italia and the Bundesbank. [6]

The transition to monetary growth control is placed at “around 1975” by both Shimamoto (1983) and Suzuki (1985). [7]

See Bank of Japan Annual Report 1978. [8]

See Shimamoto (1983) and also Suzuki (1985, 1986). [9]

“Overfunding” refers to the practice of selling more government securities than required to finance the public sector deficit. This placed continual pressure on bank liquidity which could only be relieved by purchases of private sector securities by the Bank of England. [10]

PSL2 is a broader aggregate which comprises sterling assets held by the private sector with a maturity of up to one year, including money market instruments and liquid deposits with non-bank intermediaries, as well as currency and short-term bank deposits. [11]

The gradual increase in interest in the money stock in the U.S. is discussed by Wallich and Keir(1978). [12]

The details of the changes to practices can be found in Axilrod and Lindsey (1981), Axilrod (1981). The issue of reserve asset vs interest rate control has provoked a large literature. (One example is Sivesind and Hurley (1980).) The basic conclusion of this literature is that controlling the monetary base will be superior if money demand disturbances and errors in forecasting income are dominant. If, however, money supply disturbances are dominant, the interest rate is the preferred method of control. [13]

In the event, any increase in the demand for M1 which came from shifts out of savings accounts may have been exactly neutralised by the reduction in the demand for M1 which came from continuing financial innovation (see on this, Pierce (1984)). [14]

This excludes Japan, where targets were not announced a year in advance. [15]

“Achieving” the target is defined as any outcome within the target range, or within one percentage point of a “point” target. [16]

We were unable to locate data for the 1960s for France and Italy. [17]

References

General

Argy, Victor (1983a), “Choice of Intermediate Money Target in a Deregulated and an Integrated Economy with Flexible Exchange Rates”, IMF Staff Papers, Vol. 30, No. 4, December, pp 727–754.

Argy, Victor (1983b), “The Design of Monetary and Fiscal Policies – The Experience of Four Conservative Administrations”, Centre for Studies in Money, Banking and Finance Working Paper No. 8315A, Macquarie University, December.

Argy, Victor (1987), “International Financial Liberalisation: the Australian and Japanese Experiences Compared”, Bank of Japan Monetary and Economic Studies, Vol. 5 No. 1, May, pp 105–167.

Argy, Victor (1988a), “A Post-War History of the Rules vs Discretion Debate”, Banca Nazionale del Lavoro Quarterly Review, June, pp 147–177.

Argy, Victor (1988b), “Monetary-Fiscal Exchange Rate Policy Rules: A Survey”, Centre for Studies in Money, Banking and Finance Working Paper No. 8854B, Macquarie University, June.

Argy, Victor, McKibbin, Warwick and Siegloff, Eric (1988), “Exchange Rate Regimes for a Small Economy in a Multi-Country World”, Centre for Studies in Money, Banking and Finance Working Paper No. 8852B, Macquarie University, June.

Atkinson, Paul, Blundell-Wignall, Adrian, Rondoni, Manuela and Ziegelschmidt, Helmut (1984), “The Efficacy of Monetary Targeting: The Stability of Demand for Money in Major OECD Countries”, OECD Economic Studies, Paris, No. 3, pp 145–176.

Atkinson, Paul and Chouraqui, Jean-Claude (1986), “The Formulation of Monetary Policy: A Reassessment in the Light of Recent Experience”, OECD Department of Economics and Statistics Working Papers, No. 32, March.

Bank for International Settlements (1974–1987), Annual Report.

Barro, Robert J. and Gordon, David B. (1983), “Rules, Discretion and Reputation in a Model of Monetary Policy”, Journal of Monetary Economics, Vol. 12, No. 1, pp 101–121.

Benavie, Arthur and Froyen, Richard T. (1983), “Combination Monetary Policies to Stabilise Price and Output under Rational Expectations”, Journal of Money, Credit and Banking, Vol. 15, No. 2, May, pp 186–198.

Blinder, Alan S. (1987), Hard Heads, Soft Hearts: Tough-Minded Economics for a Just Society, Addison-Wesley.

Boughton, James M. (1979), “Demand for Money in Major OECD Countries”, OECD Economic Outlook Occasional Studies, Paris, January, pp 35–57.

Chouraqui, Jean-Claude and Price, Robert W.R. (1984), “Medium-Term Financial Strategy: The Co-ordination of Fiscal and Monetary Policies”, OECD Economic Studies, Paris, No. 2, pp 7–49.

Courakis, Anthony S. (1981), “Monetary Targets: Conceptual Antecedants and Recent Policies in the U.S., U.K. and West Germany”, in Courakis, Anthony S. (ed.), Inflation, Depression and Economic Policy in the West, Mansell Publishing, London, pp 203–216.

Davis, Richard G. and Meek, Paul (1983), “Monetary Targeting – Variations on a Common Theme”, in Meek, Paul (ed.), Central Bank Views on Monetary Targeting, Federal Reserve Bank of New York, April, pp 1–5.

Dotsey, Michael and King, Robert G. (1983), “Monetary Instruments and Policy Rules in a Rational Expectations Environment”, Journal of Monetary Economics, Vol. 12, No. 3, pp 357–382.

Edey, Malcolm (1989), “Monetary Policy Instruments: A Theoretical Analysis”, this volume, pp 147–161.

Ethier, Wilfred and Bloomfield, Arthur I. (1978), “The Reference Rate Proposal and Recent Experience”, Banca Nazionale del Lavoro Quarterly Review, September, pp 211–232.

Foot, M.D.K.W. (1981), “Monetary Targets: Their Nature and Record in the Major Economies”, in Griffiths, Brian and Wood, Geoffrey E. (eds.), Monetary Targets, MacMillan, London, pp 13–46.

Friedman, Benjamin M. (1977), “The Inefficiency of Short-Run Monetary Targets for Monetary Policy”, Brookings Papers on Economic Activity, No. 2, pp 293–346.

Friedman, Benjamin M. (1982), “Time to Reexamine the Monetary Targets Framework”, Federal Reserve Bank of Boston, New England Economic Review, March/April, pp 15–23.

Friedman, Milton (1960), A Program for Monetary Stability, Fordham University Press, New York.

Friedman, Milton (1968), “The Role of Monetary Policy”, American Economic Review, Vol. 58, No. 1, March, pp 1–17.

Goodhart, Charles A.E. (1989), “The Conduct of Monetary Policy”, Economic Journal, Vol 99, June, pp 293–346.

Hoskins, W. Lee (1985), “Foreign Experiences with Monetary Targeting: A Practitioner's Perspective”, Contemporary Policy Issues, Vol. 3, No. 5, Fall, pp 71–83.

Isard, Peter and Rojas-Suarez, Liliana (1985), “Velocity of Money and the Practice of Monetary Targeting: Experience, Theory and the Policy Debate”, Staff Studies for the World Economic Outlook, International Monetary Fund, Washington D.C., July, pp 73–114.

Johnson, Karen H. (1983), “Foreign Experience with Targets for Money Growth”, Federal Reserve Bulletin, October, pp 745–754.

Jonson, P.D. and Rankin, R.W. (1986), “On Some Recent Developments in Monetary Economics”, The Economic Record, Vol. 62, No. 178, pp 257–267.

Kaldor, N. (1985), “Lessons of the Monetarist Experiment”, in van Ewijk, C. and Klant, J.J. (eds.), Monetary Conditions for Economic Recovery, Financial and Monetary Policy Studies II, Martinus Nijhoff Publishers, Dordrecht, pp 243–262.

Lamfalussy, A. (1985), “What Role for Monetary Policy Today?”, in van Ewijk, C. and Klant, J.J. (eds.), Monetary Conditions for Economic Recovery, Financial and Monetary Policy Studies II, Martinus Nijhoff Publishers, Dordrecht, pp 21–38.

Lane, Timothy D. (1985), “The Rationale for Money-Supply Targets: A Survey”, The Manchester School of Economic and Social Studies, No. 2, June, pp 179–207.

Laney, Leroy O. (1985), “An International Comparison of Experiences with Monetary Targeting: A Reaction Function Approach”, Contemporary Policy Issues, Vol. 3, No. 5, Fall, pp 99–112.

McClam, Warren (1978), “Targets and Techniques of Monetary Policy in Western Europe”, Banca Nazionale del Lavoro Quarterly Review, March, pp 3–27.

Meade, James E. (1978), “The Meaning of Internal Balance”, Economic Journal, Vol. 88, September, pp 423–435.

OECD (1979), Monetary Targets and Inflation Control, OECD Monetary Studies Series, Paris.

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

Sivesind, Charles and Hurley, Kevin (1980), “Choosing an Operational Target for Monetary Policy”, Quarterly Journal of Economics, Vol. 94, February, pp 199–203.

Sumner, M.T. (1980), “The Operation of Monetary Targets”, Carnegie-Rochester Conference Series on Public Policy, Vol. 13, pp 91–130.

Tobin, James (1980), “Stabilisation Policy after Ten Years”, Brookings Papers on Economic Activity, No. 1, pp 19–72.

Williamson, John (1983), “The Exchange Rate System”, Policy Analysis in International Economics, No. 5, Institute for International Economics, Washington, D.C., September.

Williamson, John and Miller, Marcus M. (1988), “The International Monetary System: an Analysis of Alternative Regimes”, European Economic Review, Vol. 32, No. 5, June, pp 1031–1048.

Australia

Agnew, R.W. (1985), “The Rise and Demise of Monetary Targeting in Australia”, Macquarie University Bulletin of Money, Banking and Finance, No.3, pp 27–41.

Battellino, Ric and McMillan, Nola (1989), “Deregulation and Changes in the Behaviour of Financial Intermediaries and Their Implications for Financial Aggregates”, this volume, pp 124–146.

Blundell-Wignall, Adrian and Thorp, Susan (1987), “Money Demand, Own Interest Rates and Deregulation”, Reserve Bank of Australia Research Discussion Paper No. 8703, May.

Budget Papers (1976/77–1984/85), Australian Government Publishing Service, Canberra.

Jonson, P.D. (1987), “Monetary Indicators and the Economy”, Reserve Bank of Australia Bulletin, December.

Johnston, R.A. (1985), “Monetary Policy – The Changing Environment”, T.A. Coghlan Memorial Lecture, Reserve Bank of Australia Bulletin, June.

Keating, P. (1985), “Review of the 1984–85 M3 Projection”, Treasurer's Press Release No. 8, 29 January.

Lynch, P. (1976), “The Economy and Economic Policy”, Treasurer's Press Release No. 52, 4 March.

Macfarlane, I.J. (1984), “Methods of Monetary Control in Australia”, in Juttner, D. Johannes and Valentine, Tom (eds.), The Economics and Management of Financial Institutions, Longman – Cheshire, Melbourne.

Porter, Michael G. (1982), “Monetary Targeting”, Australian Financial System Inquiry, Commissioned Studies and Selected Papers, Part I, pp 29–119, Australian Government Publishing Service.

Reserve Bank of Australian (1974/75 – 1987/88), Annual Report.

Reserve Bank of Australia Bulletin, May.

Stevens, Glenn, Thorp, Susan and Anderson, John (1987), “The Australian Demand Function for Money: Another Look at Stability”, Reserve Bank of Australia Research Discussion Paper No. 8701, January.

Canada

Bank of Canada (1975–1987), Annual Report.

Bouey, Gerald K. (1975), “Remarks at the 46th Annual Meeting of the Canadian Chamber of Commerce”, Bank of Canada Review, October, pp 23–29.

Bouey, Gerald K. (1982a), “Monetary Policy – Finding a Place to Stand”, Per Jacobsson Lecture, reprinted in Bank of Canada Review, September, pp 3–17.

Bouey, Gerald K. (1982b), “Recovering from Inflation”, Bank of Canada Review, December, pp 3–14.

Freedman, Charles (1981), “Monetary Aggregates as Targets: Some Theoretical Aspects”, Bank of Canada Technical Report No. 27.

Freedman, Charles (1983), “Financial Innovation in Canada: Causes and Consequences”, American Economic Association Papers and Proceedings, Vol. 73, No. 2, pp 101–106.

Freeman, G.E. (1981), “A Central Banker's View of Targeting”, in Griffiths, Brian and Wood, Geoffrey E. (eds.), Monetary Targets, MacMillan, London, pp 191–201.

Sellon, Gordon H. Jr. (1982), “Monetary Targets and Inflation: The Canadian Experience”, Federal Reserve Bank of Kansas City Economic Review, April, pp 16–31.

Thiessen, Gordon G. (1983), “The Canadian Experience with Monetary Targeting”, in Meek, Paul (ed.), Central Bank Views on Monetary Targeting, Federal Reserve Bank of New York, pp 100–104.

France

Aftalion, Florin (1983), “The Political Economy of French Monetary Policy”, in Hodgman, D.R. (ed.), The Political Economy of Monetary Policy: National and International Aspects, Monetary Policy Conference Series No 26, Federal Reserve Bank of Boston, pp 7–25.

Aftalion, Florin (1987), “Factors Affecting French Monetary Policy”, in Hodgman, D.R. and Wood, G.E. (eds.), Monetary and Exchange Rate Policy, MacMillan, London, pp 49–68.

Argy, Victor (1983c), “France's Experience with Monetary and Exchange Rate Management, March 1973 to End 1981”, Banca Nazionale del Lavoro Quarterly Review, December, pp 387–411.

Chouraqui, Jean-Claude (1981), “Monetary Policy and Economic Activity in France”, in Courakis, Anthony S. (ed.), Inflation, Depression and Economic Policy in the West, Mansell Publishing, London, pp 203–216.

Cobham, David and Serre, Jean-Martin (1986), “Monetary Targeting: A Comparison of French and U.K. Experience”, Royal Bank of Scotland Review, No. 149, pp 24–42.

Galbraith, James K. (1982), “Monetary Policy in France”, Journal of Post-Keynesian Economics, Vol. 4, No. 3, pp 388–403.

OECD (1974), Monetary Policy in France, Monetary Studies Series.

OECD (1975–1987), Economic Surveys, France.

Raymond, Robert (1983a), “The Formulation and Implementation of Monetary Policy in France”, in Meek, Paul (ed.), Central Bank Views on Monetary Targeting, Federal Reserve Bank of New York, pp 105–114.

Raymond, Robert (1983b), “Discussion”, in Hodgman, D.R. (ed.), The Political Economy of Monetary Policy: National and International Aspects, Monetary Policy Conference Series No 26, Federal Reserve Bank of Boston, pp 26–33.

Italy

Barbato, Michele (1987), “The Evolution of Monetary Policy and its Impact on Banks”, Review of Economic Conditions in Italy, No. 2, pp 165–208.

Ceranza, C. and Fazio, A. (1983), “Methods of Monetary Control in Italy: 1974–1983”, in Hodgman, D.R.(ed.), The Political Economy of Monetary Policy: National and International Aspects, Monetary Policy Conference Series No 26, Federal Reserve Bank of Boston, pp 65–87.

Cotula, F. and Micossi, S. (1977), “Some Considerations on the Choice of Intermediate Monetary Targets in the Italian Experience”, Actes du seminare des Banque Centrales et des Institutions Internationales, Cohiers Economiques et Monetaries No. 6, Banque de France, Paris, pp 141–161.

Fazio, Antonio (1979), “Monetary Policy in Italy from 1970 to 1978”, Kredit und Kapital, Vol. 2, pp 145–179.

OECD (1974–88), Economic Surveys, Italy.

Saini, Krishen G. (1984), “Economic Policy Management in Italy”, Review of Economic Conditions in Italy, No. 3, pp 345–363.

Sarcinelli, Mario (1981), “The Role of the Central Bank in the Domestic Economy”, Review of Economic Conditions in Italy, No. 3, pp 433–454.

Vaciago, Giacomo (1977), “Monetary Policy in Italy: The Limited Role of Monetarism”, Banca Nazionale del Lavoro Quarterly Review, Vol. 30, pp 333–348.

Vaciago, Giacomo (1983), “Discussion”, in Hodgman, D.R. (ed.), The Political Economy of Monetary Policy: National and International Aspects, Monetary Policy Conference Series No 26, Federal Reserve Bank of Boston, pp 89–92.

Vaciago, Giacomo (1985), “Monetary Policy with Credit Targets: The Italian Experience”, Greek Economic Review, Vol. 7, No. 1, pp 1–33.

Japan

Bank of Japan (1974–1987), Annual Report.

Bank of Japan (1988), “Recent Growth of Money Stock”, Research and Statistics Department Special Paper No. 161, February.

Hutchison, Michael M. (1986), “Japan's ‘Money Focussed’ Monetary Policy”, Federal Reserve Bank of San Francisco Economic Review, No. 3, pp 33–46.

Shimamoto, Reiichi (1983), “Monetary Control in Japan”, in Meek, Paul (ed.), Central Bank Views on Monetary Targeting, Federal Reserve Bank of New York, pp 80–85.

Suzuki, Yoshio (1985), “Japan's Monetary Policy Over the Past 10 Years”, Bank of Japan Monetary and Economic Studies, Vol. 3, No. 2, September, pp 1–9.

Suzuki, Yoshio (1986), Money, Finance and Macroeconomic Performance in Japan, Yale University Press, New Haven.

Switzerland

Burdekin, Richard C.K. (1987), “Swiss Monetary Policy: Central Bank Independence and Stabilisation Goals”, Kredit und Kapital, Vol. 4, pp 454–466.

OECD (1976–1988), Economic Surveys, Switzerland.

Schiltknecht, Kurt (1981), “Targeting the Base – the Swiss Experience”, in Griffiths, Brian and Wood, Geoffrey E. (eds.), Monetary Targets, MacMillan, London, pp 211–225.

Schiltknecht, Kurt (1983), “The Pursuit of Monetary Objectives” in Meek, Paul (ed.), Central Bank Views on Monetary Targeting, Federal Reserve Bank of New York, pp 72–79.

Swiss National Bank (1983), Functions, Instruments and Organisation, Zurich. Union Bank of Switzerland (1987), The Swiss Economy 1946–1986., Zurich.

United Kingdom

Argy, Victor (1983d), “The United Kingdom's Experience with Monetary and Exchange Rate Management”, Centre for Studies in Money, Banking and Finance Working Paper No. 8313A, Macquarie University, November.

Bank of England (1975–1987), Annual Report.

Bank of England (1983), “The Nature and Implications of Financial Innovation”, Bank of England Quarterly Bulletin, September, pp 358–362.

Fforde, John (1983), “The United Kingdom – Setting Monetary Objectives”, in Meek, Paul (ed.), Central Bank Views on Monetary Targeting, Federal Reserve Bank of New York, pp 51–59.

Leigh-Pemberton, Robin (1984), “Some Aspects of U.K. Monetary Policy”, Lecture at the University of Kent at Canterbury, Bank of England Quarterly Bulletin, December, pp 474–481.

Leigh-Pemberton, Robin (1986), “Financial Change and Broad Money”, Loughborough University Banking Centre Lecture in Finance, Bank of England Quarterly Bulletin, December, pp 499–507.

Richardson, Gordon (1978), “Reflections on the Conduct of Monetary Policy”, Bank of England Quarterly Bulletin, March, pp 31–37.

United States

Axilrod, Stephen H. (1981), “New Monetary Control Procedure: Findings and Evaluation from a Federal Reserve Study”, Federal Reserve Bulletin, April, pp 277–90.

Axilrod, Stephen H. (1983), “Monetary Policy, Money Supply, and the Federal Reserve's Operating Procedures”, in Meek, Paul (ed.), Central Bank Views on Monetary Targeting, Federal Reserve Bank of New York, pp 32–41.

Axilrod, Stephen H. (1985), “U.S. Monetary Policy in Recent Years: An Overview”, Federal Reserve Bulletin, January, pp 14–24.

Axilrod, Stephen H. and Lindsey, David E. (1981), “Federal Reserve System Implementation of Monetary Policy: Analytical Foundations of the New Approach”, American Economic Association Papers and Proceedings, Vol. 71, No. 2, May, pp 246–252.

Davis, Richard G. (1977), “Monetary Objectives and Monetary Policy”, Federal Reserve Bank of New York Quarterly Review, Spring, pp 29–36.

Friedman, Benjamin M. (1988), “Lessons on Monetary Policy from the 1980s”, Journal of Economic Perspectives, Vol. 2, No. 3, Summer, pp 51–72.

McCallum, Bennett T. (1984), “Monetarist Rules in the Light of Recent Experience”, American Economic Association Papers and Proceedings, Vol. 74, No. 2, May, pp 388–391.

McCallum, Bennett T. (1985), “On Consequences and Criticisms of Monetary Targeting”, Journal of Money, Credit and Banking, Vol. 17, No. 4, November, pp 570–597.

Pierce, James L. (1978), “The Myth of Congressional Supervision of Monetary Policy”, Journal of Monetary Economics, Vol. 4, April, pp 363–370.

Pierce, James L. (1984), “Did Financial Innovations Hurt the Great Monetarist Experiment?” American Economic Association Papers and Proceedings, Vol. 74, No. 2, May, pp 372–376.

Poole, William (1975), “The Making of Monetary Policy: Description and Analysis”, Federal Reserve Bank of Boston, New England Economic Review, March/April, pp 21–30.

Poole, William (1988), “Monetary Policy Lessons of Recent Inflation and Disinflation”, Journal of Economic Pespectives, Vol. 2, No. 3, Summer, pp 73–100.

Volcker, Paul A. (1977), “A Broader Role for Monetary Targets”, Federal Reserve Bank of New York Quarterly Review, Spring, pp 23–28.

Volcker, Paul A. (1978), “The Role of Monetary Targets in an Age of Inflation”, Journal of Monetary Economics, Vol. 4, April, pp 329–339.

Volcker, Paul A. (1985), Statement to Congress on the Mid Year Review of Monetary Policy Objectives for 1985, 17 July 1985, rprinted in Federal Reserve Bulletin, September, pp 690–697.

Wallich, Henry C. and Keir, Peter M. (1978), “The Role of Operating Guides in U.S. Monetary Policy: A Historical Review”, Kredit und Kapital, Vol. 11, No. 1, pp 30–52.

West Germany

Argy, Victor (1983e), “The West German Experience with Monetary and Exchange Rate Management, 1973 to 1981”, Centre for Studies in Money, Banking and Finance Working Paper No. 8311 A, Macquarie University, June.

Bockelman, H. (1977), “Quantitative Targets for Monetary Policy in Germany”, Actes du seminaire des Banques Centrales et des Institutions Internationales, Cohiers Economiques et Monetaries No. 6, Banque de France, April, pp 11–24.

Deutsche Bundesbank (1974–1987), Annual Report.

Deutsche Bundesbank (1982), Monetary Policy Instruments and Functions, Special Series No. 7, Frankfurt.

Deutsche Bundesbank (1985), “The Longer-term Trend and Control of the Money Stock”, Monthly Report, Vol. 37, No. 1, January, pp 13–26.

Deutsche Bundesbank (1988), “Methodological Notes on the Monetary Target Variable M3”, Monthly Report, Vol. 40, No. 3, March, pp 18–21.

Trehan, Bharat (1988), “The Practice of Monetary Targeting: A Case Study of the West German Experience”, Federal Reserve Bank of San Francisco Economic Review, No. 2, Spring, pp 30–44.