A key feature of Australia’s economic performance in the 1970s and for most of the 1980s was relatively high inflation. In the 1970s, inflation rose more sharply than in other OECD countries and, in the 1980s, it came down more slowly than elsewhere. The high inflation brought with it significant real costs. It interacted with the tax system and distorted investment incentives, and it increased uncertainty surrounding relative price signals. As a result, investment flowed to property rather than to more productive uses (see Chapter 2).
In the early 1990s, inflation fell to low levels. Monetary policy had been firm enough in the late 1980s to reduce underlying inflation in the face of strong economic activity, and this modest success was followed by an unexpectedly severe recession in 1990/91 that reduced inflation further. In response, the authorities moved to lock in low inflation by adopting an inflation-targeting framework aimed at achieving underlying inflation of between 2 and 3 percent, on average, over the economic cycle. Following its introduction, the main challenge was to strengthen the credibility of the authorities’ commitment to price stability. This was addressed by building a track record of low inflation and by enhancing the transparency and accountability of the Reserve Bank of Australia.
This chapter first outlines Australia’s inflation performance and monetary policy before inflation targeting. Inflation targeting is then discussed and contrasted with inflation targeting in other countries, particularly New Zealand, which pioneered inflation targeting. The experience in securing low inflation and the associated benefits are then outlined. Finally, the approach is assessed and some comments are presented on outstanding issues.
Australia’s Inflationary Period: The Early 1970s to the Mid-1980s
Australia’s inflation performance in the 1960s was comparable with that of other OECD countries, with inflation averaging about 2½ percent annually (Figure 5.1). This low rate was primarily due to sound monetary and fiscal policies that supported a fixed exchange rate, and to the absence of external inflationary shocks.1 In the 1970s, inflation rose sharply, peaking at almost 18 percent in 1975. The rise was sharper than in other OECD countries and inflation was slower to fall than elsewhere in the late 1970s. Inflation rose again in the early 1980s and remained significantly higher than rates in the G-7 countries’ through most of the 1980s.
Consumer Price Inflation
(Annual percentage change)
Sources: Australian Bureau of Statistics; and IMF, World Economic Outlook.1Headline consumer price inflation rate.Consumer Price Inflation
(Annual percentage change)
Sources: Australian Bureau of Statistics; and IMF, World Economic Outlook.1Headline consumer price inflation rate.Consumer Price Inflation
(Annual percentage change)
Sources: Australian Bureau of Statistics; and IMF, World Economic Outlook.1Headline consumer price inflation rate.Australia’s propensity for high inflation in the 1970s and 1980s stemmed from the interaction of commodity price booms in the early 1970s and early 1980s with the wage-bargaining system. The heavy reliance of the export sector on commodities, unlike most other OECD countries, meant that the commodity price boom in 1971/72 increased the terms of trade to historically high levels (Figure 5.2). Prices, wages, and incomes in the booming sectors rose and inflation increased significantly to more than 7 percent, even before the inflationary shock from the first oil crisis struck in 1973. In 1980, Australia experienced a further sharp rise in commodity prices, but the increase in export prices was largely offset by higher oil prices (the second oil shock), and the terms of trade did not rise significantly. Nonetheless, increasing awareness of Australia’s rich endowment of energy resources led to an investment boom in the energy sector.
Terms of Trade and Effective Exchange Rate
(1990 = 100)
Source: Australian Bureau of Statistics.1Reserve Bank of Australia nominal trade-weighted index.Terms of Trade and Effective Exchange Rate
(1990 = 100)
Source: Australian Bureau of Statistics.1Reserve Bank of Australia nominal trade-weighted index.Terms of Trade and Effective Exchange Rate
(1990 = 100)
Source: Australian Bureau of Statistics.1Reserve Bank of Australia nominal trade-weighted index.The centralized wage-bargaining system, supported by the notion of “comparative wage justice,” made it difficult to contain wage pressures to the booming sectors. With the increase in prices and wages in the booming sectors in the early 1970s and early 1980s, the wage-bargaining system ensured that similar wage demands by unions were granted in other sectors, thereby raising the economy-wide wage level. In the early 1970s, this phenomenon was compounded by the implementation of equal pay for females. As a result, average nonfarm wages rose by more than 90 percent in the four years to end-1975 (Figure 5.3).2
Consumer Price and Wage Inflation
(Annual percentage change)
Source: Australian Bureau of Statistics.Consumer Price and Wage Inflation
(Annual percentage change)
Source: Australian Bureau of Statistics.Consumer Price and Wage Inflation
(Annual percentage change)
Source: Australian Bureau of Statistics.The policy response to the commodity price shocks was not strong enough to contain the inflationary pressures during these episodes. In large part, this was because monetary policy was hindered by the Reserve Bank’s lack of effective independence and by the fixed exchange rate, which was in place until the Australian dollar was floated in December 1983. Although the Reserve Bank Act gives the bank independence to formulate and implement monetary policy, effectively the bank was not independent until the mid-1980s (Box 5.1). Monetary policy in the 1970s and early 1980s focused on achieving an M3 target, which was adopted in 1976 in an effort to reduce inflation. The targets were announced regularly but were seldom achieved partly because of the inconsistency of the targets with the fixed exchange rate, given the commodity price swings, but also because of the Reserve Bank’s lack of control over monetary policy.3 M3 targeting was abandoned in 1985, because financial market deregulation in the early 1980s meant that any existing stable relationship between money and nominal income broke down.4
Evolution of the Reserve Bank’s Effective Independence
The Reserve Bank Act gives the bank the power to determine monetary policy independently, consistent with the broad objectives of the Reserve Bank Act of 1959 (namely, stability of the currency of Australia; the maintenance of full employment in Australia; and the economic prosperity and welfare of the people of Australia). The bank can take the necessary action to implement policy changes. However, in the event of a dispute between the Reserve Bank and the government over monetary policy, the government may override the Reserve Bank by tabling objections before both Houses of Parliament. This override provision has never been used.
From the 1950s until the mid-1980s, monetary policy was implemented as though the Reserve Bank had no independence. This arose because the Reserve Bank did not have the instruments of monetary policy at its disposal. Before deregulation of the Australian financial markets in the mid-1980s, the main way of influencing monetary policy was by changing interest rates on government securities used to finance the deficit. This was almost entirely in the hands of the Treasurer. Bank lending and deposit rates were also centrally determined, ostensibly by the Reserve Bank, but changes required prior approval by the Treasurer. The monetary policy instrument that was wholly at the Reserve Bank’s disposal was the Statutory Reserve Deposit ratio, but this was not an effective instrument because it had little effect on monetary conditions as interest rates and the exchange rate were fixed.
It was not until financial deregulation was completed in the mid-1980s that the Reserve Bank gained a significant degree of effective independence to implement monetary policy. The deregulation swept away interest rate controls, freed up the exchange rate, and made it possible to finance the budget fully at market-determined interest rates. This left open market operations, which effectively determined short-term interest rates, as the only monetary policy instrument. Such operations were entirely in the hands of the bank, therefore giving it the capacity to act independently in accordance with the Reserve Bank Act by the late 1980s.
In August 1996, a joint statement between the Treasurer and Governor was issued to clarify the roles and responsibilities of the Reserve Bank and the government with regard to monetary policy. In the statement, the government reaffirmed the Reserve Bank’s independent responsibility for monetary policy as provided by statute. The statement noted that the government would no longer make parallel announcements when the Reserve Bank makes monetary policy adjustments. The motivation for this change was to “enhance the perception, as well as the reality, of the independence of the Reserve Bank’s decision making” (Costello and Macfarlane, 1996).
The M3 target was replaced with a “checklist” of indicators, such as unemployment, monetary aggregates, the inflation rate, the exchange rate, interest rates, and the balance of payments. Efforts to reduce inflation under the checklist approach were derailed by an exchange rate crisis. Over the two years to end-1986, the trade-weighted exchange rate index depreciated by almost 35 percent, while the terms of trade fell by 20 percent. The fall in the exchange rate was greater than that dictated by the terms of trade because of a reassessment of Australia’s long-term prospects by the financial markets, which began to focus on the structural weakness of the balance of payments and poor competitiveness given the earlier increase in real wages.
Australia’s Success in Reducing Inflation: Late 1980s-Early 1990s
Following the poor inflation record in the 1970s and first half of the 1980s, Australia succeeded in reducing inflation in the late 1980s and early 1990s. The underlying inflation rate fell from almost 10 percent in 1986, to 5 percent by 1990, and to less than 2 percent by the end of 1992.5
The Reserve Bank’s firm monetary policy stance from 1985 until the start of the recession in 1991 (except for a temporary easing in 1987) was important in reducing inflation. Real short-term interest rates averaged about 7 percent over this period, higher than at any time since the Second World War and higher than in any major OECD country. Although the goal of price stability was important in this period, concerns about the exchange rate and the balance of payments, following the exchange rate crisis in 1985/86, were the main focus of monetary policy. There was considerable public debate, however, as to the effectiveness of monetary policy in achieving external sector objectives, with growing recognition that these objectives were better addressed by fiscal and structural policy reforms.
Success in reducing underlying inflation in the late 1980s was achieved despite an asset price boom. The boom was led by an increase in share prices, which rose by 210 percent in the three years to September 1987, before a sharp correction in late 1987 of about 45 percent. This was followed by increases in property prices, with house prices rising by nearly 50 percent in the two years to end-1989. The boom was driven by strong economic growth in the late 1980s and the opportunity presented by the increased availability of credit (as a result of financial sector deregulation) to make highly levered investment in shares and property.6
The impact on consumer prices of the asset price boom in the late 1980s and the depreciation of the exchange rate in 1985/86 was muted in large part by incomes policy in support of monetary policy. In particular, the agreement between the government and the Australian Council of Trade Unions (ACTU) in 1985, which included income tax cuts in return for less than full indexation of wages (see Chapter 6), helped contain the second-round impact of the exchange rate depreciation on wages.
Fiscal consolidation by the Commonwealth government in the late 1980s also supported monetary policy by reducing the threat of a further depreciation of the exchange rate. By 1987/88, the exchange rate had begun to strengthen on the back of success in establishing financial discipline and a recovery in the terms of trade. This appreciation played a key role in insulating the economy from the inflationary impact of a rise in commodity prices—a major advantage of the floating exchange rate.
Other factors also contributed to reducing inflation, particularly the unexpectedly severe recession in 1990/91, along with growing competitive pressures stemming from the increasing openness of the economy, and a fall in world inflation.
Although Australia was successful in reducing inflation to low levels by the early 1990s, the absence of a clear medium-term objective for monetary policy at this stage had the potential to unsettle the financial markets. The Reserve Bank had not clarified the relative importance of price stability to its other statutory objectives, particularly the external sector objectives that had been the focus of monetary policy in the late 1980s.
Australia’s Adoption of Inflation Targeting
The adoption of an inflation target as the objective of monetary policy entailed a shift in the Reserve Bank’s focus in the early 1990s, rather than a discrete change in the policy regime toward the price stability objective. The approach was effectively in place by 1993, and was given increasing prominence in a series of statements by the Governor in 1993 and 1994. He stated that monetary policy would be conducted to maintain an average underlying rate of “around 2–3 percent… over a run of years” (Fraser, 1994). Subsequently, the Reserve Bank defined the target more precisely as achieving underlying inflation of between 2 and 3 percent, on average, over the economic cycle.
The Rationale for Inflation Targeting
The primary reason for the adoption of the inflation-targeting framework was to lock in low inflation. The clear focus on a medium-term inflation target contrasted with the experience of the late 1980s when the objective of monetary policy was not focused on price stability. In adopting the inflation target, the authorities aimed to break with past experience of inflation rising to relatively high levels during economic booms, while falling to low levels only during periods of slow or negative growth.
Various developments in theoretical and empirical research provided support for inflation targeting. One of these was that the primary goal of monetary policy should be reasonable price stability. This follows from research suggesting that in the long run, monetary policy affects only nominal variables, and has no marked impact on real economic activity (i.e., that the long-run Phillips curve is vertical).7
A second development is the insight that any attempt to exploit the existence of short-run nominal rigidities to lower the unemployment rate tends to impart an inflationary bias to monetary policy. This stems from the long-running rules versus discretion debate, which took a new path in the mid-1970s with the development of the concepts of time consistency (or credibility) and inflation bias. The inflation bias results from attempts by the monetary authorities to exploit the inflation-output trade-off (characterized by the short-run Phillips curve) to raise output above its potential level by engineering unexpected inflation. Such policies are eventually fully anticipated by private agents, however, at which point they cease to have an impact on output, and serve only to raise the average rate of inflation. This increase in average inflation is known as inflation bias. Kydland and Prescott (1977), Calvo (1981), and Barro and Gordon (1983) show that adherence to a rule to determine current and future monetary policy will remove the inflation bias and result in lower inflation than under discretionary monetary policy. Inflation targeting can be seen as adherence to a rule.
A third development is research arguing that accountability for the inflation outcome, supported by transparency and openness of the operating framework for monetary policy, can bring important credibility gains (Nolan and Schaling, 1996). Inflation targets can provide a transparent guide to the specific objectives of monetary policy, whose commitment and credibility can then be assessed on the basis of whether policy actions are taken to ensure that the targets are achieved.8 Transparency can be enhanced by the central bank publicly presenting the factors influencing the forecast of inflation, including the model upon which the forecasts are based or the judgments made by the central bank. In this way, the public can assess whether the forecasts are credible and whether any monetary policy action based on the forecasts is appropriate. Furthermore, by communicating to the public the forecast path of inflation that monetary policy seeks to accomplish, inflation targets serve as a coordination device in wage- and price-setting processes and in forming the public’s inflationary expectations. If the targets are credible, they may help anchor inflationary expectations and thereby lower the costs of achieving and maintaining low inflation.
International Trend Toward Inflation Targeting
The adoption of inflation targeting in Australia followed a trend toward this approach in several advanced economies in recent years. New Zealand pioneered the approach by adopting inflation targets in 1990, and was followed by Canada (1991), the United Kingdom (1992), Sweden (1993), Finland (1993), Australia (1993), and Spain (1994) (see Table 5.1).
International Comparison of Inflation Targets
International Comparison of Inflation Targets
| Country | Starting Date | Inflation Target | Caveats | Transparency Measures | Credibility Measures |
|---|---|---|---|---|---|
| New Zealand | March 1990 (modified December 1996) | Annual CPI inflation rate excluding credit services (defined as the 4-quarter rate of change) of 0-3 percent. From 1990 until December 1996 the target range was 0-2 percent. | Significant direct effects of specific exogenous shocks are excluded, including increases in indirect taxes and government charges, changes in the terms of trade, and a crisis such as a natural disaster or a disease-induced fall in livestock numbers. | Publication of monetary policy statement every six months, quarterly release of inflation projections, and occasional press statements by Reserve Bank officials. | The Reserve Bank Act of 1989 defined price stability as the sole objective of monetary policy. Policy Target Agreements are negotiated between the Reserve Bank and the government, subject to a highly public government override. |
| Canada | February 1991 | 12-month CPI inflation rate of 1-3 percent. | In practice, the Bank of Canada excludes food, energy prices, and indirect tax increases from the CPI. | Regular progress monitoring, particularly in the Bank of Canada’s Annual Report, but official inflation forecasts are not made public. | Joint announcement of the inflation target by the Minister of Finance and the Governor of the Bank of Canada. There were no major changes to the BOC’s mandate; it remains responsible for implementation of policy, subject to a highly public government override. |
| United Kingdom | October 1992 (modified May 1997) | Underlying inflation of 2½ percent. | Underlying inflation is measured by the annual rate of change of the retail price index, excluding mortgage interest payments (RPIX). | The Treasury announces the inflation target for the year in its annual budget with the expectation that it will be kept stable. The Bank of England publishes the Inflation Report quarterly (including its inflation forecasts). It also publishes minutes of the Monetary Policy Committee’s monthly meetings after a six-week interval. | In May 1997, the Chancellor gave the Bank of England operational responsibility for setting interest rates, and achieving the inflation target set by government. |
| Sweden | January 1993 | Annual rate of CPI inflation of 2 percent from 1995, within a range of ±1 percent. | None. | Publication of inflation reports three times a year, but reports lack explicit inflation forecasts. | There is no legislated price stability goal; in fact, the Riksbank can change the target at any time. |
| Finland | February 1993 | Underlying CPI inflation rate to be stabilized permanently at about 2 percent by the end of 1995. | In practice, an underlying inflation measure is used which excludes the effects of indirect taxes, government subsidies, house prices, and mortgage interest payments. | Quarterly publications that cover price developments in a broad sense and assess whether the target will remain within reach. | The inflation target was endorsed by the government in March 1993. |
| Australia | 1993 | Underlying inflation rate of 2-3 percent, on average, over the economic cycle. | The underlying inflation measure excludes volatile items such as interest rate charges, fuel, and food. | Monetary policy report every six months. Broad statements on the inflation outlook, but no detailed publication of inflation forecasts by the Reserve Bank. The Governor reports to a parliamentary committee. | In a joint policy statement by the Treasurer and Governor in 1996, the government endorsed the target and acknowledged the Reserve Bank’s independence. |
| Spain | 1994 | Less than 3 percent by early 1997; remain below 3 percent for 1997. Below 2 percent for 1998 and beyond. | None. | Monetary policy report every six months, but no publication of inflation forecasts. The Governor reports to a parliamentary committee. | Target set by the Bank of Spain, which gained independence in 1994 with the primary goal of maintaining price stability. |
International Comparison of Inflation Targets
| Country | Starting Date | Inflation Target | Caveats | Transparency Measures | Credibility Measures |
|---|---|---|---|---|---|
| New Zealand | March 1990 (modified December 1996) | Annual CPI inflation rate excluding credit services (defined as the 4-quarter rate of change) of 0-3 percent. From 1990 until December 1996 the target range was 0-2 percent. | Significant direct effects of specific exogenous shocks are excluded, including increases in indirect taxes and government charges, changes in the terms of trade, and a crisis such as a natural disaster or a disease-induced fall in livestock numbers. | Publication of monetary policy statement every six months, quarterly release of inflation projections, and occasional press statements by Reserve Bank officials. | The Reserve Bank Act of 1989 defined price stability as the sole objective of monetary policy. Policy Target Agreements are negotiated between the Reserve Bank and the government, subject to a highly public government override. |
| Canada | February 1991 | 12-month CPI inflation rate of 1-3 percent. | In practice, the Bank of Canada excludes food, energy prices, and indirect tax increases from the CPI. | Regular progress monitoring, particularly in the Bank of Canada’s Annual Report, but official inflation forecasts are not made public. | Joint announcement of the inflation target by the Minister of Finance and the Governor of the Bank of Canada. There were no major changes to the BOC’s mandate; it remains responsible for implementation of policy, subject to a highly public government override. |
| United Kingdom | October 1992 (modified May 1997) | Underlying inflation of 2½ percent. | Underlying inflation is measured by the annual rate of change of the retail price index, excluding mortgage interest payments (RPIX). | The Treasury announces the inflation target for the year in its annual budget with the expectation that it will be kept stable. The Bank of England publishes the Inflation Report quarterly (including its inflation forecasts). It also publishes minutes of the Monetary Policy Committee’s monthly meetings after a six-week interval. | In May 1997, the Chancellor gave the Bank of England operational responsibility for setting interest rates, and achieving the inflation target set by government. |
| Sweden | January 1993 | Annual rate of CPI inflation of 2 percent from 1995, within a range of ±1 percent. | None. | Publication of inflation reports three times a year, but reports lack explicit inflation forecasts. | There is no legislated price stability goal; in fact, the Riksbank can change the target at any time. |
| Finland | February 1993 | Underlying CPI inflation rate to be stabilized permanently at about 2 percent by the end of 1995. | In practice, an underlying inflation measure is used which excludes the effects of indirect taxes, government subsidies, house prices, and mortgage interest payments. | Quarterly publications that cover price developments in a broad sense and assess whether the target will remain within reach. | The inflation target was endorsed by the government in March 1993. |
| Australia | 1993 | Underlying inflation rate of 2-3 percent, on average, over the economic cycle. | The underlying inflation measure excludes volatile items such as interest rate charges, fuel, and food. | Monetary policy report every six months. Broad statements on the inflation outlook, but no detailed publication of inflation forecasts by the Reserve Bank. The Governor reports to a parliamentary committee. | In a joint policy statement by the Treasurer and Governor in 1996, the government endorsed the target and acknowledged the Reserve Bank’s independence. |
| Spain | 1994 | Less than 3 percent by early 1997; remain below 3 percent for 1997. Below 2 percent for 1998 and beyond. | None. | Monetary policy report every six months, but no publication of inflation forecasts. The Governor reports to a parliamentary committee. | Target set by the Bank of Spain, which gained independence in 1994 with the primary goal of maintaining price stability. |
The public announcement of inflation targets in these countries, along with measures to improve the transparency and accountability of monetary policy, has been undertaken in an effort to improve the credibility of the authorities’ commitment to price stability, and thereby lower the economic costs of achieving and maintaining low inflation. Countries with established records of low inflation, such as Germany and Switzerland, have not seen the need to adopt such a framework, given the already strong credibility of their commitment to price stability.
Key Features of Australia’s Inflation-Targeting Framework
The adoption of the inflation-targeting framework required the specification of: (1) the level of the target; (2) bounds, if any, of the target; (3) which measure of inflation to target; and (4) the policy horizon over which the target applies.
The Level of the Inflation Target
The level of the inflation target should be consistent with the optimal rate of inflation, which is a function of the costs of inflation. In Australia, the authorities consider that the optimal target for inflation is likely to be above zero given the downward rigidity of nominal wages and upward bias in the CPI. Hence the adoption of the 2–3 percent target, which is somewhat higher than in several other inflation-targeting countries.9
Downward rigidity in nominal wages arises from the system of awards in Australia that provide a floor for the terms and conditions of employment (see Chapter 6). Wages could become more flexible because of the recent reforms to industrial relations, although a degree of downward rigidity in nominal wages is likely to remain. Furthermore, a longer period of price stability would likely alter inflation expectations and wage-setting habits.
The upward price index measurement bias creates a wedge between the optimal inflation target in theory and in practice. No in-depth study of the measurement bias has been undertaken in Australia, but the bias is unlikely to be large given the regular rebasing of the CPI. In Canada and New Zealand, which have comparable statistical systems, the bias in the CPI is estimated at about ½ to 1 percent per annum.10
Target Bounds
In Australia, the inflation target of 2–3 percent is thought of as a desired “central tendency” for inflation, or “thick point,” rather than a narrow “target band” with upper and lower bounds. The authorities, however, have more recently begun to emphasize the midpoint of the target range as the medium-term target.11 This new emphasis on the midpoint, rather than the 2–3 percent range, is motivated by the wish to avoid the public perception that 2 and 3 percent are hard bounds.
Several considerations led the authorities to adopt the current specification. Inflation forecasts involve considerable difficulties, so much so that the margin of uncertainty surrounding a central inflation forecast over a year has been estimated by Debelle and Stevens (1995) to be well over ±1 percentage point.12 The uncertainty stems from both the variation of inflation over the economic cycle and from a range of exogenous shocks. Debelle and Stevens (1995) also present a model of the Australian economy showing that short-term fluctuations in prices and output may be substantial in Australia, because of a range of exogenous shocks reflecting the open nature of the economy and the sizable impact of commodity price swings. They note that, even with policy responding appropriately to shocks, lags mean that short-term inflation control may not be very precise. Therefore, the monetary authorities consider that an inflation-target band wide enough to absorb shocks could be too wide to provide a useful anchor for expectations. Moreover, considering that no country has maintained inflation consistently within a very narrow band (for example, 2 percentage points wide),13 the authorities regard a narrow band (with hard bounds) to be unrealistic and potentially damaging to their credibility.
The lack of specific bounds is an innovative aspect of the Australian inflation-targeting framework. Several other inflation-targeting countries specify upper bounds, and sometimes lower bounds, for the inflation target. The Australian authorities consider that their framework reflects the lessons learned from the experience of other countries with inflation targeting. Moreover, they consider that their concerns about the credibility of specifying upper bounds has been reinforced by the experience in New Zealand, where the upper bound of 2 percent was increased to 3 percent in late 1996, after it had been exceeded by a small amount in 1995 and 1996.
The difference between the Australian and New Zealand approach to bounds is not as significant in practice as implied by the formulation of the targets. In New Zealand, temporary and limited deviations from the target band are tolerated. The target band is seen as the objective that monetary policy aims at achieving, but the target will, at times, be missed due to the “inevitable uncertainties in forecasting and lags in the effectiveness of monetary policy” (Reserve Bank of New Zealand, 1996b). In Australia, the authorities took vigorous action to limit the increase in inflation above the target in 1995 and early 1996, during a period of strong economic growth, even though the extent and duration of deviations are not precisely defined in the framework.14 This suggests that although there is no hard-edged band, in practice, the authorities have recognized the risks to the credibility of the medium-term inflation target that can arise from sustained periods of inflation significantly above the target.
Inflation Measures
Inflation targeting, by definition, requires the choice of an inflation measure. Given that the ultimate objective of monetary policy is to achieve aggregate price stability for the economy as a whole, a broad index of prices, such as the Gross National Expenditure (GNE) or Gross Domestic Product (GDP) deflator may be the most appropriate index. In Australia, the measure used for inflation targeting is the underlying CPI, which is more familiar to the public than the GDP deflator, more timely, and, unlike the GDP deflator, not subject to revision. Moreover, the underlying CPI is more closely linked to inflationary expectations and wage demands.
The use of an underlying measure of inflation in Australia, although differing in definition from other countries (Box 5.2), has proven effective in removing a degree of variability from the target variable.15 In turn, this has meant that monetary policy has not been required to react to factors that did not impact on the trend rate of inflation, but may have involved a once-and-for-all change in the price level. In this way, the targeting of underlying inflation rather than headline inflation has likely reduced the extent of output variability that would have resulted if monetary policy had reacted instead to developments in the headline rate. Furthermore, by targeting underlying inflation, policymakers avoid “chasing their tail” by reacting to past changes in interest rates that are reflected in the headline rate of inflation.
Comparison of Underlying Inflation Measures in Australia and New Zealand
The approach to measuring the underlying inflation rate in Australia differs from that in other inflation-targeting countries, particularly New Zealand. In Now Zealand, a policy-targets agreement is signed by the Treasurer and the Governor of the Reserve Bank, which includes caveats to allow deviations from the targets prompted by shocks that should not be offset by monetary policy. These shocks include significant changes in foreign trade prices, indirect taxes, government charges, and a crisis such as a natural disaster or a major disease-induced fall in livestock numbers. Caveats are activated by the Reserve Bank only when they have a large and identifiable impact on inflation. Therefore, the underlying inflation rate in New Zealand is the consumer price index excluding credit services less the caveats that are activated (which may differ from quarter to quarter).
The Australian authorities did not follow this approach for two reasons. First, judging the direct impact of the shocks may be difficult. Second, tolerating temporary deviations from the inflation objective based on such judgments may have negative effects on the credibility of monetary policy.
Although Australia did not adopt caveats as in New Zealand, the exclusion of a wide range of items from the underlying measure of inflation in Australia, more than in most inflation-targeting countries, means that the two approaches treat some shocks in a similar way.1 For instance, prices for items such as automotive fuel and local government rates and charges that are covered by the caveats in the New Zealand approach are excluded from the Australian underlying CPI.
Indirect taxes are included in the underlying CPI in Australia. This is at odds with the practice in most other inflation-targeting countries (including New Zealand), where indirect taxes are generally excluded from the underlying measure. However, the Reserve Bank calculated a “true” underlying inflation index that excludes the impact of indirect taxes for two quarters in 1995 that were heavily affected by indirect tax changes, but this has not been emphasized in assessing performance.
While the current measure of the underlying CPI in Australia is not fully comparable with the practice in other countries (because of the inclusion of indirect taxes and the exclusion of a wider range of other items), it has been retained as the target index because it has become a widely accepted measure of underlying inflation. Therefore, the authorities have been reluctant to use an alternative measure for the inflation target because of concerns that this may be perceived by the public as manipulating the target variable for their own gain.
1Some items, such as travel, accommodation, and clothing, are excluded from underlying inflation in Australia because of the seasonality of their price movements, and it is debatable whether these items should be excluded.The Policy Horizon
In most inflation-targeting countries, the policy horizon over which the inflation target is to be achieved is well defined so that the central bank can be held accountable for the inflation outcome in that period. In Australia, however, the policy horizon and the extent and duration of deviations from the “central tendency” are not precisely specified in the framework. The authorities’ intention in this regard is to allow monetary policy some latitude in responding to fluctuations in output without prejudicing the medium-term inflation objective. In the authorities’ view, the formulation recognizes that in the fastest-growing phases of the business cycle there will be a tendency for inflation to exceed the medium-term objective, and at other times to fall below it. In the case of a large shock, the Australian formulation recognizes explicitly that inflation should be brought back on track gradually to avoid causing undue real costs.
The Australian approach, however, has the disadvantage of lacking a clear policy horizon. This makes it difficult to assess the success of the Reserve Bank in achieving the inflation objective. While the Reserve Bank can only be held ultimately accountable over the economic cycle, there is an obligation for the bank to explain the inflation outlook and what is being done to ensure that it is consistent with the medium-term target. Nevertheless, over a shorter period, it is more difficult to assess whether inflation is consistent with the medium-term objective, and therefore the target may not provide a firm anchor for inflationary expectations.
Institutional Support for the Inflation-Targeting Framework
In Australia, the adoption of the target did not require any change to the Reserve Bank Act of 1959, which provides the framework for the operation of monetary policy, because inflation targeting is consistent with the objective of price (or currency) stability in the Act. The inflation target was unilaterally adopted by the Reserve Bank in 1993, without a joint commitment with the government. Subsequently, the government informally endorsed the target. The target was also supported by the main labor union body, the Australian Council of Trade Unions (ACTU), which agreed in the June 1995 Accord VIII to keep wage claims in line with the Reserve Bank’s inflation objective, but this support was withdrawn when the Accord lapsed in March 1996.
It was not until August 1996, on the appointment of a new Governor, that the target was formally endorsed by the government in a joint monetary policy statement by the Governor and the Treasurer. The statement emphasized that the price stability objective was not in conflict with the Reserve Bank’s employment objective but was supportive of it. It also included a formal commitment by the Governor to the target.
The extent of legislative and institutional support for inflation targeting has varied widely across countries. While no inflation-targeting countries included inflation targets explicitly in their legislation, several countries supported the approach by legislating greater central bank independence.16 In part, this was based on research showing that a high degree of independence is associated with low inflation (Grilli and others, 1991). This is because an independent central bank is better insulated from political pressures that take the focus away from the medium-term objective of price stability. The Reserve Bank is not one of the most independent central banks in the world, but its relatively high degree of independence provides important support for inflation targeting (Box 5.3).
Transparency and Accountability
Specific targets, a well-defined policy horizon, and a transparent central bank provide the public with a yardstick with which to measure a central bank’s performance and to hold policymakers accountable. Some caution is needed, however. The long and variable lags between monetary policy action and the final inflation objective make it hard to assess whether a breach of the target was caused by policy errors or by shocks outside the control of the bank.
The need for accountability and transparency in Australia to support achievement of the inflation target was highlighted in the joint policy statement of August 1996. In recent years the Reserve Bank has taken steps to increase the transparency of its monetary policy operations, although to a lesser extent than in some other countries (see Box 5.4). Most important, it began to change official interest rates in discrete steps and announce and explain these policy changes. Some further steps to improve transparency and accountability were announced in the joint policy statement of August 1996. First, the Reserve Bank will release regular (roughly every six months) statements on monetary policy and progress toward the Reserve Bank’s objectives. Second, the Governor will report twice a year to a Parliamentary select committee on the conduct of monetary policy (this is a formalization of the previous practice).
Although the Reserve Bank will be held more accountable for inflation outcomes than in the past, the government has recognized that the Reserve Bank has no control over other policies that affect inflation. In this regard, the August 1996 statement emphasized the role that disciplined fiscal policy must play in achieving the inflation target.
Multiple Versus Sole Objectives for Monetary Policy
The multiple objectives of monetary policy in Australia’s Reserve Bank Act (i.e., currency stability, employment, and welfare) contrast most sharply with the focus on price stability in New Zealand’s 1989 legislation. In Australia, the concern about a strict inflation target is that an unfavorable supply shock may trigger unacceptable adverse consequences in the real economy in the short run.17
At the same time, the Australian framework essentially acknowledges that the employment and welfare objectives in the Reserve Bank Act are best pursued in the long run by targeting price stability. The Governor of the Reserve Bank has noted that “the most useful thing that monetary policy can do in the long run is to encourage sustainable low-inflation expansions” (Macfarlane, 1996). Similarly, New Zealand’s Policy Targets Agreement between the Governor and the Treasurer requires monetary policy to maintain price stability and thereby make its maximum contribution to sustainable economic growth, employment, and development opportunities within New Zealand.
International Comparison of Central Bank Independence
Some international comparisons of central bank independence suggest that although the Reserve Bank does not rank as one of the most independent central banks (such as in Germany, Switzerland, and the United States), it has a relatively high degree of independence.
Two of the more widely quoted studies (Grilli and others, 1991; and Cukierman, 1992) rank Australia’s central bank independence among the top half of the OECD countries covered. Grilli and others (1991) rank Australia relatively high on the basis that (1) the Governor is appointed for seven years, longer than the three-year Commonwealth electoral cycle; (2) the Reserve Bank Act requires the bank to pursue currency stability; (3) the Act strengthens the position of the bank in the case of a difference of views with the government, as any disagreements on policy action must be tabled in parliament if the bank is to be overridden; and (4) the bank has control over the extent of monetary financing of the budget deficit and the setting of the official cash interest rate.
However, Grilli and others (1991) do not rank the Reserve Bank’s independence as high as some other central banks because (1) the board is appointed for only five years, compared with eight years in Germany and 14 years in the United States; (2) the Secretary to the Treasury has a voting right on the Reserve Bank Board, whereas in Germany, Italy, the Netherlands, Portugal, Spain, Switzerland, and the United States there are no government officials on the board; and (3) the Reserve Bank is responsible for bank supervision, the objectives of which could conflict with the monetary policy objectives. (In September 1997, however, the government announced that the Reserve Bank would no longer be responsible for bank supervision; see Chapter 7.)
Another study, by Nolan and Schaling (1996), ranks Australia relatively low in its scale of central bank independence. This assessment is based primarily on the grounds that final authority does nor rest with the Reserve Bank, and that there is a government official—the Secretary to the Treasury—on the Reserve Bank Board. However, the independence of the Reserve Bank of Australia is greater than suggested by the Nolan and Schaling (1996) study, and is more accurately assessed by Grilli and others (1991).1 Nolan and Schaling (1996) do not take account of the effective degree of authority that the Reserve Bank has over decisions. While it is true that final authority for monetary policy decisions rests with the government and not with the Reserve Bank, as the government has the power to override the bank’s decisions, the override procedures are highly public and politically demanding. These procedures have never been used, which reinforces the Reserve Bank’s independence. The presence of the Secretary to the Treasury on the Reserve Bank Board (with a voting right), however, may have meant that disagreements between the bank and the government were resolved within the board and, therefore, the override procedures were not needed.
In assessing the independence of a central bank it is useful to distinguish between two forms of independence: goal independence and instrument independence. (A distinction first introduced by Debelle and Fischer, 1994.) Goal independence allows the central bank to choose the goals of monetary policy, while instrument independence lets the central bank pursue its goals in the manner it deems most appropriate. Viewed from this perspective, the Reserve Bank has a relatively high degree of overall independence. While the Reserve Bank Act establishes the broad objectives of monetary policy, the Bank has the independence to establish the specific goals, thereby giving it a degree of goal independence. Furthermore, it has the power to change the instrument, albeit with a government override.
1It should be noted that all three studies of independence are based on somewhat dated information. They do not take account of the changes in New Zealand in 1989; France and Spain in 1994; and the United Kingdom in 1997.The Australian framework is in line with the argument presented by Green (1996) that targeting both inflation and output consistent with nonaccelerating inflation may be preferable to a single price stability objective. He notes that a credible commitment to an inflation target by the monetary authorities and the abandoning of any effort to stabilize output can attenuate the extent of an inflationary bias. The single price stability objective may be at the expense of an increase in output volatility, however. He argues that the credibility of inflation targets may be enhanced by targeting output at a level consistent with nonaccelerating inflation. If the targets for inflation and output lie on the long-run supply curve so that the output objective corresponds to nonaccelerating inflation, no trade-off is needed to balance the two objectives. In this way, the framework addresses the underlying cause of the “inflation bias” in monetary policy by eliminating the monetary authorities’ ability to engineer surprise inflation without violating its output target. Thus, in contrast to the “conventional wisdom” that a sole objective of price stability for monetary policy is required to strengthen the credibility of the inflation targets, a dual monetary policy objective may enhance the prospects for low inflation. Green (1996) notes this approach is likely to be strengthened by the publication of the inflation projections along with projections of potential output, so that both targets are transparent.
Accountability and Transparency in Other Inflation-Targeting Countries
Inflation-targeting countries have generally attempted to build the credibility of the inflation targets by adopting measures to hold the central bank accountable for the inflation outcomes. In many cases, the accountability has provided an appropriate offset to the increased instrument independence that many central banks have obtained under this framework. In other words, the increased responsibility chat the central bank obtained for achieving desired inflation outcomes came with greater accountability for those outcomes. At the same time, inflation-targeting countries have required central banks to be more transparent about the conduct or monetary policy, so that both the government and the public can monitor their performance and more readily hold them accountable.
Some countries have gone further than Australia in emphasizing accountability and transparency. In particular, New Zealand and the United Kingdom have published regular reports on monetary policy that include official forecasts of quarterly inflation over a two- to three-year horizon. In both cases, the forecasts have helped to signal the path of projected inflation consistent with achieving the objective. The Reserve Rank does not publish detailed forecasts of inflation or output because the authorities are concerned that the lack of precision in the forecasts and the unpredictable nature of prices could lead to large forecasting errors and thereby undermine the credibility of the bank. (In part, the United Kingdom has attempted to address similar concerns by publishing the forecasts complete with confidence ranges.) Instead, the current practice is for the bank to make statements about the outlook for inflation in broad terms without providing details of the quarterly track.
The New Zealand framework incorporates a more transparent and immediate degree of accountability than in Australia and other inflation-targeting countries. The Governor of the Reserve Bank of New Zealand is held accountable for achieving inflation targets spelled out in a public Policy Targets Agreement signed by the Governor and Minister of Finance. If the Governor fails to achieve the targets, he risks losing tenure of his position. Walsh (1995) argues that this may be close to the optimal contract for a central bank governor. McCallum (1995), however, is skeptical that the contract eliminates the inflation bias, because the government has the same incentive not to enforce the contract as identified for the central bank in the inflation bias literature discussed earlier.
In Australia, the tenure of the Governor is not explicitly tied to inflation performance. Nevertheless, performance on the inflation front would no doubt be taken into consideration in any decision to reappoint the Governor once the seven-year term is complete. Therefore, a degree of accountability through the Governor’s tenure exists, albeit less explicit and over a longer period than in New Zealand. Furthermore, given the more medium-term focus of the inflation target, the Governor could not be held accountable for inflation outcomes over a shorter period, unless inflation was clearly inconsistent with the medium-term target.
The Australian authorities consider that the Governor’s accountability does not require strengthening, as the financial market plays an important role in disciplining monetary policy. Australian financial markets have been quick to react to news concerning the outlook for inflation, which has put public pressure on the bank to take appropriate action.
Operational Aspects of Inflation Targeting
Since monetary policy affects economic activity and inflation with long lags, and knowledge of the transmission mechanism is imperfect, the Reserve Bank takes a forward-looking approach and relies on multiple indicators to evaluate the inflation outlook. The bank continually assesses monetary conditions by tracking indicators of future trends in inflation and economic activity, such as wages, capacity utilization, the yield curve, movements in asset prices, market or survey-based expectations of inflation, indicators of the fiscal stance, and financial aggregates. No explicit intermediate target or forecasting rules are employed—the Reserve Bank simply makes “informed judgements” (Fraser, 1994).
Based on its evaluation of the inflation outlook, the bank can choose a time path for the official cash interest rate target (the main monetary policy instrument in Australia), which results in a forecast of future inflation (conditional on the policy setting, and based on an informed judgment) that is consistent with achieving the inflation target. Given that the horizon for achieving the target is not precisely defined, and given that the decision-making process involves assessing current and future inflationary trends and comparing this with the target, the process leaves room for discretion. In turn, such discretionary action will affect the credibility of the target, therefore underlying the importance of transparency and openness in the conduct of monetary policy.
The Experience in Securing Low Inflation and Its Benefits
The major aim of Australia’s inflation-targeting framework was to lock in low inflation, and there by build the credibility of the authorities’ commitment to achieving price stability. Therefore, success of the framework can be assessed in large part by analyzing inflation performance and the indicators of credibility, such as inflationary expectations.
Initially, the adoption of inflation targeting did not appear to enhance the credibility of the Reserve Bank, as inflationary expectations, which were at historically low levels, did not fall further in 1993 and 1994 (Box 5.5). This may have been because the inflation-targeting framework had not yet been tested in Australia (or in other countries) over a full business cycle. Moreover, concerns about the flexible nature of the target, particularly the absence of limits on the extent or duration of deviations from the target range, may have made the target less credible. More important, given the long history of high inflation in Australia, it is likely that inflationary expectations would only be reduced to levels consistent with the inflation target once a sustained track record of low inflation was achieved.
Recent actions by the Reserve Bank have helped build the credibility of the inflation target. In late 1994, the bank reacted to strong economic growth that threatened achievement of the inflation target by tightening monetary policy (the official cash rate was raised by 275 basis points). As a result, the bank was successful in limiting the extent and duration of the deviation of underlying inflation above 3 percent—the upper end of the target range. By December 1997, annual underlying inflation had fallen to 1.4 percent, the lowest rate since records for this measure began in 1971, and well below the midpoint of the inflation target.18
An important benefit of inflation targeting is that the financial markets have effectively re-rated Australian assets, partly on the strength of the inflation performance and the joint statement on monetary policy in August 1996, which formalized the government’s endorsement of the monetary policy framework. By the end of 1997, real long-term interest rates had fallen to just over 4½ percent, their lowest level since the current economic expansion gained pace in 1993/94.19 This contrasts with real long-term interest rates of about 7½ percent in 1989, at a similar stage of the previous economic expansion. Furthermore, the differential between 10-year government bond yields in Australia and the United States had fallen to about 30 basis points by the end of 1997, below the average of about 250 basis points in 1994/95 and 1995/96, and considerably less than the 350-600 basis point differential in the late 1980s. The re-rating by the financial markets also reflected other developments, such as the government’s fiscal policy and reform agenda, which demonstrated the benefits of the inflation target being supported by sound fiscal and structural policies.
Inflationary Expectations and Inflation Targeting
Inflationary expectations in Australia have fallen markedly over the past 10 years. Consumers’ inflation expectations fell sharply in the early 1990s, from about 10–11 percent in the late 1980s to average about 4½ percent over 1991–96 measured by the Melbourne Institute survey median expected inflation rate for the year ahead. The fall in expectations occurred in line with the fall in actual inflation, and there is no clear evidence to suggest that the adoption of inflation targeting in 1993 had a perceptible impact on consumers’ inflationary expectations. By the end of 1997, consumers’ expectations had fallen to about 3 percent.
Financial-market expectations of inflation are more consistent with the Reserve Bank’s target. The differential between the CPI indexed and non-indexed 10-year bonds (a proxy for financial-market inflation expectations over the next decade) fell from about 8 percent in the late 1980s to about 3-5 percent in 1991–95. By the end of 1997, the differential stood at slightly more than 2 percent. Financial-market economists surveyed by the Reserve Bank just after the release of the 1997 December quarter CPI figure had a median forecast of underlying inflation for the year to June 1998 of 1.8 percent, with a pickup expected to 2.7 percent for the subsequent year.
Business-sector inflationary expectations have been somewhat higher than financial-market expectations. According to the December 1997 National Australia Bank survey, the most common expectation continues to be in the 3 to 4 percent range, implying a “credibility gap” between the level of business sector expectations and the midpoint of the Reserve Bank’s target range of about 1 percentage point.
Inflationary Expectations
(In percent)
Sources: Australian Bureau of Statistics; and Reserve Bank of Australia.1Calculated as the 10-year Treasury bond yield less the comparable indexed bond yield.Inflationary Expectations
(In percent)
Sources: Australian Bureau of Statistics; and Reserve Bank of Australia.1Calculated as the 10-year Treasury bond yield less the comparable indexed bond yield.Inflationary Expectations
(In percent)
Sources: Australian Bureau of Statistics; and Reserve Bank of Australia.1Calculated as the 10-year Treasury bond yield less the comparable indexed bond yield.The maintenance of low inflation in the past four years has reduced the costs associated with inflation. One of the most widely recognized costs of inflation in Australia is the incentive that inflation and the tax and accounting systems create for investment in property (as discussed in Chapter 2). This is evident from the strong growth of private investment in dwellings and other construction in the 1980s, a period when inflation averaged about 8 percent (Figure 5.4). In the current economic expansion, however, private investment has been more concentrated on plant and equipment rather than dwelling and other construction. The shift toward more productive plant and equipment investment may have been due, in part at least, to the lower level of inflation in the 1990s and is likely to be beneficial for medium-term economic growth.
Assessment of Australia’s Inflation-Targeting Framework
The experience so far with Australia’s inflation-targeting framework is short-lived. It is not yet clear whether the relatively flexible approach to inflation targeting is preferable to the more rigid approach taken in other countries, particularly New Zealand. The rationale for adopting inflation targeting discussed earlier would argue that regimes with a well-defined and transparent target, and a central bank that is held accountable for inflation outcomes, would gain in terms of enhanced credibility and lower costs of maintaining inflation, when compared with the Australian approach. A comparison of Australia’s inflation and output performance with other inflation-targeting countries, however, suggests that Australia’s approach has produced superior results thus far. Since the adoption of the target in 1993, Australia has achieved inflation outcomes consistent with the target and slightly below the average for other inflation-targeting countries, while real GDP growth has been stronger and its variability lower than in other inflation-targeting countries (Table 5.2).
Inflation-Targeting Countries: Inflation and Real GDP Growth
(In percent)
Data for the period 1980 to the adoption of targets and from that point to 1997 are presented.
Year-on-year consumer price inflation, measured quarterly, based on headline indices for Canada, Finland, Spain, and Sweden; but underlying indices for Australia (Treasury’s underlying measure), New Zealand (CPI excluding credit services), and the United Kingdom (Retail Trade Price Index, excluding mortgage interest rates).
Year-on-year real GDP growth, measured quarterly.
Inflation-Targeting Countries: Inflation and Real GDP Growth
(In percent)
| Annual Inflation2 | Real GDP Growth3 | ||||
|---|---|---|---|---|---|
| Country and Period1 | Mean | Standard Deviation | Mean | Standard Deviation | |
| Australia | |||||
| 1980–92 | 7.2 | 2.4 | 2.8 | 2.8 | |
| 1993–97 | 2.2 | 0.6 | 3.9 | 1.1 | |
| Canada | |||||
| 1980–90 | 6.4 | 3.1 | 2.9 | 2.8 | |
| 1991–97 | 2.1 | 1.7 | 1.7 | 2.0 | |
| Finland | |||||
| 1980–92 | 6.6 | 3.0 | 2.0 | 3.8 | |
| 1993–97 | 1.2 | 0.7 | 3.4 | 3.0 | |
| New Zealand | |||||
| 1980–89 | 9.9 | 3.9 | 1.9 | 2.6 | |
| 1990–97 | 2.2 | 1.2 | 2.3 | 3.3 | |
| Spain | |||||
| 1980–93 | 9.0 | 3.8 | 2.4 | 2.0 | |
| 1994–97 | 3.7 | 1.2 | 2.7 | 0.6 | |
| Sweden | |||||
| 1980–92 | 7.8 | 3.2 | 1.5 | 1.8 | |
| 1993–97 | 2.1 | 1.7 | 1.6 | 2.5 | |
| United Kingdom | |||||
| 1980–92 | 6.9 | 3.8 | 1.8 | 2.7 | |
| 1993–97 | 2.8 | 0.3 | 2.8 | 1.2 | |
| Average for inflation-targeting countries | |||||
| 1980 to adoption of targets | 7.7 | 3.3 | 2.2 | 2.6 | |
| Adoption of targets to 1997 | 2.3 | 1.0 | 2.6 | 2.0 | |
Data for the period 1980 to the adoption of targets and from that point to 1997 are presented.
Year-on-year consumer price inflation, measured quarterly, based on headline indices for Canada, Finland, Spain, and Sweden; but underlying indices for Australia (Treasury’s underlying measure), New Zealand (CPI excluding credit services), and the United Kingdom (Retail Trade Price Index, excluding mortgage interest rates).
Year-on-year real GDP growth, measured quarterly.
Inflation-Targeting Countries: Inflation and Real GDP Growth
(In percent)
| Annual Inflation2 | Real GDP Growth3 | ||||
|---|---|---|---|---|---|
| Country and Period1 | Mean | Standard Deviation | Mean | Standard Deviation | |
| Australia | |||||
| 1980–92 | 7.2 | 2.4 | 2.8 | 2.8 | |
| 1993–97 | 2.2 | 0.6 | 3.9 | 1.1 | |
| Canada | |||||
| 1980–90 | 6.4 | 3.1 | 2.9 | 2.8 | |
| 1991–97 | 2.1 | 1.7 | 1.7 | 2.0 | |
| Finland | |||||
| 1980–92 | 6.6 | 3.0 | 2.0 | 3.8 | |
| 1993–97 | 1.2 | 0.7 | 3.4 | 3.0 | |
| New Zealand | |||||
| 1980–89 | 9.9 | 3.9 | 1.9 | 2.6 | |
| 1990–97 | 2.2 | 1.2 | 2.3 | 3.3 | |
| Spain | |||||
| 1980–93 | 9.0 | 3.8 | 2.4 | 2.0 | |
| 1994–97 | 3.7 | 1.2 | 2.7 | 0.6 | |
| Sweden | |||||
| 1980–92 | 7.8 | 3.2 | 1.5 | 1.8 | |
| 1993–97 | 2.1 | 1.7 | 1.6 | 2.5 | |
| United Kingdom | |||||
| 1980–92 | 6.9 | 3.8 | 1.8 | 2.7 | |
| 1993–97 | 2.8 | 0.3 | 2.8 | 1.2 | |
| Average for inflation-targeting countries | |||||
| 1980 to adoption of targets | 7.7 | 3.3 | 2.2 | 2.6 | |
| Adoption of targets to 1997 | 2.3 | 1.0 | 2.6 | 2.0 | |
Data for the period 1980 to the adoption of targets and from that point to 1997 are presented.
Year-on-year consumer price inflation, measured quarterly, based on headline indices for Canada, Finland, Spain, and Sweden; but underlying indices for Australia (Treasury’s underlying measure), New Zealand (CPI excluding credit services), and the United Kingdom (Retail Trade Price Index, excluding mortgage interest rates).
Year-on-year real GDP growth, measured quarterly.
Have Inflation Targets Made a Difference?
Australia has maintained inflation at low levels in recent years, but was this because of inflation targeting per se or would an alternative monetary policy approach have worked just as well? This question is difficult to answer, as it is not possible to assess the counterfactual. Nonetheless, a comparison of Australia (and other inflation-targeting countries) with non-inflation-targeting countries can help assess the performance of inflation targeting relative to other approaches. In a recent study, Haldane (1996) presents some evidence of a regime shift in the inflation performance of inflation-targeting countries because of the adoption of the targets. This is based on the observations that mean inflation fell significantly in the inflation targeting countries following the adoption of the targets, and that the fall was greater than in a control group of large non-inflation-targeting countries (France, Germany, Italy, Japan, and the United States; see Table 5.3). In fact, inflation was lower on average in the 1990s in the inflation-targeting countries than for this control group.
Inflation and Growth
(In percent)
Headline consumer price inflation for all countries except Australia (the underlying CPI), New Zealand (the CPI excluding credit services) and the United Kingdom (the Retail Price Index, excluding mortgage interest rates). Inflation rates are calculated as the year-on-year change in the quarterly index.
Dates used for adoption of targets are: Canada, 1991; Finland, 1993; New Zealand, 1990; Spain, 1994; Sweden, 1993, and the United Kingdom, 1992.
France, Germany, Italy, Japan, and the United States.
Belgium, Denmark, Greece, Iceland, Ireland, Norway, and Luxemburg and Portugal.
Inflation and Growth
(In percent)
| Annual Inflation1 | Real GDP Growth | ||||
|---|---|---|---|---|---|
| Country and Period | Mean | Standard Deviation | Mean | Standard Deviation | |
| Australia | |||||
| 1980–92 | 7.2 | 2.4 | 2.8 | 2.8 | |
| 1993–97 | 2.2 | 0.6 | 3.9 | 1.1 | |
| Other inflation-targeting countries | |||||
| 1980 to adoption of targets2 | 7.8 | 3.5 | 2.1 | 2.6 | |
| Adoption of targets to 1997 | 2.3 | 1.1 | 2.5 | 2.1 | |
| Large non-inflation-targeting countries3 | |||||
| 1980–89 | 6.0 | 3.7 | 2.5 | 1.8 | |
| 1990–97 | 2.9 | 1.2 | 2.1 | 2.2 | |
| Small non-inflation-targeting countries4 | |||||
| 1980–89 | 13.9 | 6.5 | 2.7 | 2.7 | |
| 1990–97 | 4.6 | 2.3 | 2.8 | 1.9 | |
Headline consumer price inflation for all countries except Australia (the underlying CPI), New Zealand (the CPI excluding credit services) and the United Kingdom (the Retail Price Index, excluding mortgage interest rates). Inflation rates are calculated as the year-on-year change in the quarterly index.
Dates used for adoption of targets are: Canada, 1991; Finland, 1993; New Zealand, 1990; Spain, 1994; Sweden, 1993, and the United Kingdom, 1992.
France, Germany, Italy, Japan, and the United States.
Belgium, Denmark, Greece, Iceland, Ireland, Norway, and Luxemburg and Portugal.
Inflation and Growth
(In percent)
| Annual Inflation1 | Real GDP Growth | ||||
|---|---|---|---|---|---|
| Country and Period | Mean | Standard Deviation | Mean | Standard Deviation | |
| Australia | |||||
| 1980–92 | 7.2 | 2.4 | 2.8 | 2.8 | |
| 1993–97 | 2.2 | 0.6 | 3.9 | 1.1 | |
| Other inflation-targeting countries | |||||
| 1980 to adoption of targets2 | 7.8 | 3.5 | 2.1 | 2.6 | |
| Adoption of targets to 1997 | 2.3 | 1.1 | 2.5 | 2.1 | |
| Large non-inflation-targeting countries3 | |||||
| 1980–89 | 6.0 | 3.7 | 2.5 | 1.8 | |
| 1990–97 | 2.9 | 1.2 | 2.1 | 2.2 | |
| Small non-inflation-targeting countries4 | |||||
| 1980–89 | 13.9 | 6.5 | 2.7 | 2.7 | |
| 1990–97 | 4.6 | 2.3 | 2.8 | 1.9 | |
Headline consumer price inflation for all countries except Australia (the underlying CPI), New Zealand (the CPI excluding credit services) and the United Kingdom (the Retail Price Index, excluding mortgage interest rates). Inflation rates are calculated as the year-on-year change in the quarterly index.
Dates used for adoption of targets are: Canada, 1991; Finland, 1993; New Zealand, 1990; Spain, 1994; Sweden, 1993, and the United Kingdom, 1992.
France, Germany, Italy, Japan, and the United States.
Belgium, Denmark, Greece, Iceland, Ireland, Norway, and Luxemburg and Portugal.
Haldane also observes that there is little evidence that the mean or variability of output was adversely affected by disinflation in the inflation-targeting countries. In fact, it is quite the contrary for the inflation-targeting countries, as there is evidence that output variability decreased more recently for this group, whereas variability increased for the large non-inflation-targeting countries.
A comparison with an alternative control group of smaller non-inflation-targeting countries (Belgium, Denmark, Greece, Iceland, Ireland, Norway, Luxembourg, and Portugal) also suggests that the inflation-targeting countries were more successful in achieving low inflation.20 While this control group reduced annual inflation from relatively high levels in the 1980s to about 4½ percent in the 1990s, inflation remained about twice as high as in the inflation-targeting countries.
A comparison of Australia with the other country groups suggests that inflation targeting helped achieve a regime shift in inflation in Australia, without costs in terms of greater output variability. Both the mean and variability of underlying CPI inflation in Australia were significantly lower in the period following the introduction of inflation targeting in 1993. Furthermore, the mean and variability of inflation following the adoption of the target was lower than that for all other country groups. At the same time, Australia’s real GDP growth following adoption of the target was higher than for the earlier period, and was well above the average for the other country groups. Moreover, the variability of Australia’s real GDP growth was lower in the period after the adoption of the target, and was lower than for all other country groups in this period.
The cross-country evidence tends to suggest that the inflation-targeting approach has made a difference, particularly in Australia, without involving excessive real costs. The experience with the approach, however, has been relatively short. None of the inflation-targeting countries has experienced a full business cycle or a range of shocks, and the true test of the approach will come over time.
Outstanding Issues
Considerable progress has been made in refining Australia’s forward-looking monetary framework, which has allowed greater emphasis on and success in targeting the middle of the band for the inflation target. But the new framework itself will not guarantee maintenance of low inflation. The strength of the authorities’ commitment to firm monetary and fiscal policies will be a more important factor in this regard. In particular, the authorities will need to make careful use of the flexibility to take countercyclical monetary and fiscal policy measures built into the macroeconomic policy framework, and to ensure that the focus on the medium-term objectives is not lost and that the credibility of the framework is not undermined.
The framework has been successful in maintaining low inflation during an expansionary phase of the cycle, but has not yet faced a contractionary phase. Such a contraction could lead to a potential conflict between the medium-term price stability and full employment objectives of the Reserve Bank, as political pressures to loosen monetary policy for short-term gains come to bear. An inappropriately loose stance of monetary policy in such circumstances may pose a risk for the medium-term inflation objective, given the long lags in the transmission mechanism. Furthermore, the framework has not yet been tested by a supply-shock with world inflation having been benign in recent years.
The framework has gained credibility with the financial markets but inflationary expectations have lagged the actual reduction in inflation, particularly among consumer and business groups, with continuing costs in terms of wage claims and unemployment. These considerations argue for continued vigilance to build a strong track record of low inflation so as to bring expectations into line with the actual low rates. Greater public disclosure of the Reserve Bank’s inflation and output forecasts may help the process of building confidence in the inflation-targeting framework.
References
Barro, Robert J., and David B. Gordon, 1983, “Rules, Discretion, and Reputation in a Model of Monetary Policy,” Journal of Monetary Economics, Vol. 12, pp. 101–21.
Calvo, G.A., 1981, “On Time Inconsistency of Optimal Policies in a Monetary Economy,” in Rational Expectations, ed. by R.E. Lucas and T.J. Sargent (Allen and Urwin).
Costello, The Rt. Hon. P., and Ian Macfarlane, 1996, “Statement on the Conduct of Monetary Policy,” August 14, reprinted in Reserve Bank Bulletin (Sydney: Reserve Bank of Australia, September).
Cukierman, Alex, 1992, Central Bank Strategy, Credibility and Independence: Theory and Evidence (Cambridge, Massachusetts: MIT Press).
Crawford, A., 1993, “Measurement Biases in the Canadian CPI,” Bank of Canada Technical Report No. 64.
Debelle, Guy, and Stanley Fischer, 1994, “How Independent Should a Central Bank Be?” in Goals, Guide lines and Constraints Facing Monetary Policymakers, Conference Series No. 38, ed. by Jeffrey C. Fuhrer (Boston, Massachusetts: Federal Reserve Bank of Boston).
Debelle, Guy, and Glenn Stevens, 1995, “Monetary Policy Goals for Inflation in Australia,” Research Discussion Paper 9503 (Sydney: Reserve Bank of Australia, March).
Fraser, Bernie, 1994, “The Art of Monetary Policy,” speech delivered at Twenty-Third Conference of Economists, September 26, Surfer’s Paradise, Australia.
Green, John, 1996, “Inflation Targeting: Theory and Policy Implications,” Staff Papers, International Monetary Fund, Vol. 43 (December).
Grilli, Vittario, Donato Masciandaro, and Guido Tabellini, 1991, “Political and Monetary Institutions and Public Financial Policies in the Industrial Countries,” Economic Policy, Vol. 6 (October), pp. 342–91.
Haldane, Andrew, 1996, “Some Thoughts on Inflation Targeting,” paper prepared for the Kinstanz Seminar (London: Bank of England, April).
Kydland, Finn, and Edward Prescott, 1977, “Rules Rather than Discretion: The Inconsistency of Optimal Plans,” Journal of Political Economy, Vol. 12, No. 3.
Macfarlane, Ian, 1992, “The Lessons for Monetary Policy,” in Proceeding of a Reserve Bank of Australia Conference on the Deregulation of Financial Intermediaries (Sydney: Reserve Bank of Australia).
Macfarlane, Ian, 1996, “The Task of Monetary Policy,” speech delivered to CEDA Annual General Meeting Dinner, November 28, Melbourne, Australia.
McCallum, Bennett, 1995, “New Zealand’s Monetary Policy Arrangements: Some Critical Issues,” Reserve Bank of New Zealand Discussion Paper G95/4 (Sydney).
Nolan, Charles, and Eric Schaling, 1996, Monetary Policy Uncertainty and Central Bank Accountability (London: Bank of England).
Rae, David, Michelle Lloyd, and A. Fung, 1992, “Defining Price Stability: What Should We Aim For?” (unpublished; Wellington: Reserve Bank of New Zealand).
Reserve Bank of Australia, 1996, Bulletin (Sydney, September).
Reserve Bank of New Zealand, 1996a, Briefing on the Reserve Bank (Wellington, October).
Reserve Bank of New Zealand, 1996b, Monetary Policy Statement (Wellington, June)
Walsh, Carl, 1995, “Optimal Contracts for Central Bankers,” American Economic Review, Vol. 85 (March), pp. 150–67.
Although the Australian currency was fixed to the United Kingdom pound sterling in the 1960s, it was not devalued with the sterling in 1967. From December 1972, the exchange rate was fixed against the U.S. dollar, and from September 1974 it was fixed against a trade-weighted exchange rate index.
Similarly, in the early 1980s, a high wage settlement in the booming metal trade industry set the pace for other wage settlements, and average nonfarm wages increased by almost 50 percent in the three years to end-1982.
The M3 target was introduced in 1976 and was the Treasurer’s target. It was announced in the budget speech with major decisions on monetary policy taken by the Monetary Policy Committee of the Cabinet.
The authorities saw the greatly increased bank advances in the first half of the 1980s as a reflection of “re-intermediation” resulting from deregulation (i.e., a shifting of financial intermediation back to the banks from the fringe financial institutions that had not been subject to the same extent of regulation as the banks).
The underlying consumer price index (CPI) is defined as the headline CPI adjusted for special factors that are judged to be either highly volatile or not directly related to excess demand pressures (such as changes in mortgage and consumer interest charges and in the prices of publicly provided items). It was defined by the Australian Treasury and is derived by excluding the following items from the CPI basket: meat and seafood, fresh fruit and vegetables, clothing, government-owned dwelling rents, mortgage interest charges, consumer credit charges, local government rates and charges, household fuel and light, automotive fuel, postal and telephone services, urban transportation fares, tobacco and alcohol, health services, pharmaceuticals, holiday travel and accommodation, and education and child care.
Headline CPI inflation fell from almost 10 percent in 1986 to about ¼ percent by the end of 1992. The larger fall in the headline rate was due to a reduction in interest rates. The Australian Bureau of Statistics includes mortgage rates directly in the headline CPI, with a weight of about 6 percent. Mortgage rates are closely related to short-term interest rates, as most mortgages in Australia are of the floating rate type. As a result, changes in monetary policy tend to have a pronounced effect on the headline CPI.
Private sector credit grew by more than 150 percent in the five years to 1989, and this growth occurred despite the relatively firm monetary policy stance. Macfarlane (1992) notes that the credit growth was associated with a fall in credit standards in the late 1980s, on the pan: of both lenders and borrowers. On the part of lenders, credit standards fell because of strong competition among financial institutions as a result of the removal of interest rate ceilings and the entry of new players, particularly foreign banks. Macfarlane argues that this explanation is too simplistic and that a major factor in the credit growth was pent-up demand for credit from businesses responding to the incentive to borrow provided by inflation and the tax deducibility of interest payments. “A wide cross-section of Australian business was convinced that the road to increased wealth was through geared asset acquisition, and some who were in doubt felt compelled to engage in it anyhow as a defense against takeover” (Macfarlane, 1992).
Nonetheless, monetary policy can have a significant impact on the real economy in the short run.
Alternative approaches to monetary policy, such as monetary or exchange rate targeting, aim to achieve the ultimate objective of price stability by targeting an intermediate variable. A disadvantage of the alternative approaches is that the ultimate objective of price stability is less transparent than with inflation targeting, making it more difficult to assess the performance of monetary policy and there by reducing the accountability for achieving the ultimate objective of price stability.
The midpoint of the target range for Australia 2½ percent) is one percentage point above the midpoint of the target range for New Zealand (revised upward recently from 1 to 1½ percent), and half a percentage point above the midpoint for Canada, Finland, and Sweden. The target for Spain (less than 3 percent by 1997 and less than 2 percent from 1998 onwards) is more comparable with Australia, but given that this is an upper bound (with no lower bounds specified), it implies a lower level of targeted inflation than in Australia. Only in the United Kingdom does the inflation target (2’ percent) match the midpoint of the target in Australia.
See Crawford (1993) for Canada; and Rae, Lloyd, and Fung (1992) and Reserve Bank of New Zealand (1996a) for New Zealand.
The Governor has described the inflation target as essentially requiring that monetary policy be conducted so that, in the medium term, inflation averages around 2½ percent per annum (Macfarlane, 1996).
This result is based on an analysis of forecasts of inflation for the full period 1985–94. Given that the variance of inflation has fallen as the mean inflation rate has fallen over this period, forecasting errors are now likely to be lower than in the late 1980s. This is reflected in somewhat lower forecasting errors reported by Debelle and Stevens (1995) for the period 1990–94 than for the late 1980s.
Debelle and Stevens (1995) note that the longest period of relatively stable low inflation in recent history occurred in Germany from 1954–71, when average inflation was 2.3 percent with a standard deviation of just under 1 percent.
Year-on-year underlying inflation peaked in this period at 3.3 percent in March 1996 and exceeded 3 percent for only four quarters (September 1995-June 1996). However, year-on-year headline inflation reached a peak of 5.1 percent in December 1995, which may have pushed up inflationary expectations. By the end of 1997, underlying inflation had been reduced to 1.4 percent, below the target range.
While it is much less volatile than the headline rate, the underlying measure has followed the same medium-term trend over the cycle as the headline index. For instance, over the 10 years to the end of 1997, the underlying measure increased by 40.4 percent, while the headline measure rose by 40.1 percent.
In New Zealand, the adoption of inflation targeting was accompanied by legislating the goal of price stability as the prime objective of an independent and accountable central bank. In Spain, the adoption of targets was also accompanied by a greater degree of independence for the central bank. In the United Kingdom, greater central bank independence was adopted in 1997, several years after the adoption of inflation targeting.
In the case of a demand shock—for example, an unanticipated increase in aggregate demand—the monetary policy response of raising interest rates would serve the dual purposes of slowing activity and lowering inflation. However, in the case of a supply shock—for example, an unanticipated fall in aggregate supply—monetary policy aimed at dampening inflation would tend to exacerbate the shortfall in output.
The headline CPI fell by 0.3 percent from September 1996 to September 1997 and by 0.2 percent from December 1996 to December 1997. These were the lowest headline CPI inflation rates since June 1962, and were lower than the underlying CPI inflation mainly because of a fall in mortgage interest rates (which are included in the headline measure but excluded from the underlying measure).
The Australian real long-term interest rates are calculated here using the 10-year government bond yield less the annual underlying CPI inflation rate; however, use of the current annual inflation rate may mean the measure of real interest rates is inappropriate as it does not necessarily reflect inflationary expectations. Yields on CPI-indexed bonds (which by definition need no adjustment for expected inflation) suggest somewhat lower real interest rates of about 4 percent at the end of 1997.
These smaller countries may be a better control group than the large non-inflation-targeting countries, because they are small and experienced high inflation in the 1980s, similar to many of the inflation-targeting countries.
