Unemployment in Australia since the Second World War
The unemployment rate is influenced by a range of factors. These include the prevailing economic environment, the institutional and policy framework for the labour market, and a range of social factors such as attitudes towards desired hours of work and willingness to move locations or undertake further training in order to secure employment.
Chart 1 illustrates Australia’s measured unemployment rate over the period from the end of the Second World War to 2003-04 (Box 1 briefly reviews the trends in unemployment in Australia between Federation and the Second World War). It is evident from the chart that unemployment rates have traditionally been subject to significant short-term fluctuations. It is also evident that there was a substantial increase in the average unemployment rate from the mid to late 1970s, compared with earlier decades.
In the period from the Second World War to the mid-1970s, the unemployment rate fluctuated around a longer-term average of 2áperácent, reaching as high as around 3áperácent, with a low point of around 1áperácent. In contrast, in the period from the mid-1970s to the early 2000s, the unemployment rate fluctuated around a longer-term average of 7Żáperácent, reaching as high as 10.9áperácent in 1992, with a low point currently below 6áperácent.
These longer-term average rates of unemployment are sometimes referred to as structural unemployment rates, on the assumption that they are more likely to reflect the longer-term economic, institutional and social factors driving the labour market. These factors can change over time, and when they change the structural rate of unemployment will change. However, most moves in the unemployment rate are not structural, but rather reflect only short run or cyclical swings in economic activity.
Chart 1: Australia’s unemployment rate, 1945-46 to 2003-04
Source: Data to 1965-66 from Preliminary Annual Database 1900-01 to 1973-74 (Reserve Bank of Australia Research Discussion Paper 7701). Data from 1966-67 to the present is from Labour Force, Australia (ABSáCat. No. 6202.0). 2003-04 figure is the average for the nine months to March 2004.
The increase in the structural rate of unemployment since the mid-1970s likely reflected a range of factors. In the 1970s, poor macroeconomic policy management (especially in the face of the first oil price shock), combined with rigid institutional arrangements in the labour market and strong union influence, resulted in much higher inflation and a sharp increase in real wages relative to underlying trends in productivity. The resultant real wage ‘overhang’ resulted in a reduction in employment opportunities and rising unemployment as businesses reacted to the increased cost of labour.
Box 1: Australia’s unemployment rate since Federation
Chart 2: Unemployment rate, Australia, 1900-01 to 2003-04
Source: Data to 1973-74: Preliminary Annual Database 1900-01 to 1973-74 (Reserve Bank of Australia Research Discussion Paper 7701). Data from 1974-75: Labour Force, Australia (ABS Cat. No. 6202.0). 2003-04 figure is the average for the nine months to March 2004.
Chart 2 shows the level and variability of Australia’s unemployment rate since Federation.1 Australia’s unemployment rate prior to the Great Depression in the early 1930s generally varied in a range between 3 and 6áperácent. The Great Depression was associated with a very large increase in unemployment, rising from just over 4áperácent in 1926-27 to over 19áperácent in 1931-32.
It then took a further 9 years, and the outbreak of the Second World War in 1939, for the unemployment rate to return to under 4áperácent. During the war, the labour market was distorted by the demand for military personnel, and so called ‘manpower planning’ in many key sectors of the economy.
Through the late 1970s productivity growth also slowed, prolonging and intensifying the effect the real wage overhang had on the unemployment rate. The existing centralised wage fixing system meant that wages were generally tied to inflation, so the real wage overhang could not be quickly removed over time. The early 1970s also saw a range of reforms to the social welfare system specifically aimed at providing a more generous and comprehensive welfare safety net. However, a consequence of some of these reforms is likely to have been a reduced incentive for some unemployed people to actively seek work.2
The pattern of a substantial increase in both unemployment and inflation (often referred to as ‘stagflation’) was seen in many countries during the period of the late 1970s, and Australia’s experience was not unique. Although starting from a higher base, the average unemployment rate jumped by around 2ápercentage points in the less regulated labour market of the United States, and by an average of over 4ápercentage points in the larger European countries with more regulated labour markets. The increase in average unemployment rates in Australia was one of the largest at around 5ápercentage points (comparing 1960-1973 with 1973-1990), with Australia moving from having one of the lowest unemployment rates in the world to a relatively high rate. Box 2 provides further detail on longer-term trends in unemployment and inflation for Australia and some other OECD countries.
Chart 1 also shows that, since the sharp increase in unemployment in the 1970s, previous episodes of low unemployment have generally not been sustainable. They were often associated with an overheating economy, with the unemployment rate rebounding to higher levels shortly after reaching historical lows. The factors which drove these rebounds in unemployment are complex and likely to have varied from episode to episode. In broad terms, emerging tightness in labour markets as the unemployment rate declined towards the lower point is likely to have put increased pressure on wages (especially in the early 1980s), at least in some industries and regions, particularly if output and employment had been increasing rapidly.
With centralised wage fixing, which Australia has had in some form over most of the period, wage pressures in one sector or region would quickly flow into other parts of the economy, even where labour demand pressures were less apparent. This, in turn, tended to raise concerns about increased inflationary pressure as higher wage costs were passed on in the form of higher prices, a process exacerbated by limited or restricted competition in the market for many goods and services. Put another way, whatever ‘impulse’ to the economy generated pressure for higher wages in one sector or region, the centralised wage system helped ‘propagate’ this to the broader economy. Limited competition in product markets then allowed this first-round inflationary effect to be embedded into the economy.
Box 2: Unemployment and inflation — the Phillips Curve
One way to measure the change in Australia’s economic performance over time is to trace out the combination of inflation rates and unemployment rates recorded over a period of years. The result is sometimes referred to as a ‘Phillips Curve’.3
Chart 3: The Australian Phillips Curve
Source: Labour Force, Australia (ABS Cat. No. 6202.0) and Consumer Price Index, Australia (ABSáCat.áNo. 6401.0), Treasury calculations.
Australia recorded a combination of low inflation and low unemployment in the 1960s. In the 1970s, in response to poor macroeconomic policy settings and rising pressure for higher wages, inflation increased sharply and there was some increase in unemployment. Inflation moderated somewhat in the 1980s, but unemployment continued to rise on average. Since the early to mid-1990s, inflation has been sustained at much lower levels, and unemployment has been trending down. This more recent experience reflects the benefits of more stable and coherent macroeconomic policy settings and the impact of the longer-term programme of microeconomic reforms, which has made the economy much more adaptable and flexible.
Chart 4 shows that the United Kingdom (UK) and the Netherlands’ Phillips Curves over a similar time period display many of the features revealed in the Australian curve. Like Australia, both these countries undertook broad structural reform programmes during the 1980s and 1990s which have recently enabled their economies to simultaneously demonstrate low inflation and moderately low unemployment rates.
Chart 4: Other nations’ Phillips Curves
Governments and monetary authorities tended to respond to these inflationary pressures by tightening fiscal and monetary policy (either through ‘credit rationing’ in earlier periods when financial markets were highly regulated, or higher interest rates in more recent decades) to slow the economy and reduce inflationary pressure, thereby exacerbating (or magnifying) the negative consequences for employment of the original shock.
Once inflationary pressures had emerged, it often required a period of much slower economic and employment growth, and hence a sustained period of higher unemployment, to reverse the process, particularly in the absence of a coherent and credible medium-term macroeconomic strategy and framework in earlier decades, and the absence of flexible labour markets. Further, it usually proved difficult for policy makers to judge the required extent of tightening of fiscal and monetary policy to quell the emerging inflationary pressure.
This created the risk that either too little would be done initially and inflationary pressure would continue to build, followed by an excessive tightening of policy and a very sharp slowdown in growth and rise in unemployment, or policy would be tightened excessively in the initial response.
As discussed in more detail in the next section, Australia’s unemployment experience since 2002 stands in contrast with these earlier outcomes. Unemployment has declined steadily to below 6áperácent and is expected to remain below 6áperácent on average over the forecast horizon, in the context of steady economic growth and minimal wage and price pressures.
The labour market in the 1980s, 1990s and 2000s
Chart 5 shows Australia’s unemployment rate for the past 20áyears in more detail, including the Budget forecasts out to mid-2005. Following the structural upward shift in the unemployment rate in the late 1970s Australia’s unemployment rate has declined to levels around 6áperácent on only three occasions: 1989 to 1990, late 2000, and late 2002 to the present.
Chart 5: Unemployment rate, Australia, 1984 to 2005
Source: Labour Force, Australia (ABS Cat. No. 6202.0), Treasury forecasts.
These three episodes have had quite different aftermaths. In the case of the 1989 episode, the unemployment rate subsequently rose sharply, reaching almost 11áperácent by 1992. Following the 2000 episode the unemployment rate rebounded modestly, but only to slightly over 7áperácent, before resuming a downward trend. In contrast, in the current episode, the unemployment rate has been around 6áperácent for an extended period since Octoberá2002, and has been below 6áperácent since Septemberá2003. The unemployment rate is expected to remain below 6áperácent on average over the forecast horizon. Statement 3 in this Budget Paper contains more detail about the forecasts for the labour market.
While the unemployment rate declined to around 6áperácent in each case, there are significant differences in the overall performance of the labour market in the three periods, particularly with respect to employment growth, wage pressures, hours worked and regional variability in unemployment. As discussed below, these various indicators point to the current decline in unemployment to below 6áperácent being much more sustainable than previously.
In the 1989 episode the overall annual rate of employment growth was very high, at around 4 perácent. In fact, the through-the-year growth of 5 perácent experienced in the year to May 1989 has only been exceeded during one three-month period (in 1986) since monthly data collection commenced in 1978. This period of strong growth in employment was largely a reflection of very strong growth in output. The economy grew by around 5Żáperácent in the year to the September quarter 1989 (when some of the lowest rates of unemployment for the episode were reached).
The period between May 1989 and November 1989 includes four of the ten highest through-the-year growth rates in total employment on record. Between Juneá1988 and Aprilá1989, when the unemployment rate declined from over 7áperácent to 6áperácent, employment grew at an annualised rate of 4Żáperácent.
In comparison, between September 1999 and September 2000, when the unemployment rate declined from 7 perácent to 6 perácent, employment growth was steadier but still above trend at just over 3áperácent (in annualised terms). The more subdued employment growth reflected the overall growth of the economy, which reached 4áperácent through-the-year to the June quarter 2000, and was 3Żáperácent through-the-year to the September quarter.
During the most recent episode, the unemployment rate declined from 7áperácent in Januaryá2002 to 6áperácent in Augustá2003. During that period, employment grew at an annualised rate of 2áperácent, broadly in line with the longer-term average growth rate of employment in Australia. Growth in the overall economy was also relatively steady, at 2.7áperácent through-the-year to September 2003 (although this growth rate was below trend due to the continued impact of drought on the agricultural sector and the global slowdown).
The annual growth rate of employment did reach 3áperácent briefly in this period, but returned to trend rates soon after. Employment growth has been maintained at around trend rates since then as the unemployment rate has trended down further to below 6áperácent in recent months. These sustainable rates of employment growth have helped to contain wage and price pressures as the unemployment rate has declined.
The steadier and more stable growth in employment in the two most recent episodes reflects the more stable growth in overall GDP.4 And this in turn reflects the improvements to overall macroeconomic policy frameworks, and the long-term programme of microeconomic reform, which have helped to provide greater stability to the overall economy. These issues are discussed in more detail in the second section of this paper.
Other indicators of labour market tightness
While the unemployment rate is the most commonly used summary measure of the state of the labour market, it does not capture the full story of what is happening within the market. Movements in the unemployment rate are, in fact, the result of large flows into and out of employment, and into and out of the labour force, and the unemployment rate captures only the net impact of these flows (Boxá3). Over time these flows will also have an impact on the industry distribution of employment (seeáBox 4).
Movements in other labour market indicators, discussed below, can also provide useful additional information about the possible emergence of excessive tightness or imbalances in the labour market, and hence how sustainable the unemployment rate may be over the medium term.
Growth in labour costs
One such indicator to assess the relative tightness of the labour market, and hence the sustainability of the unemployment rate, is the trend in the growth rate of labour costs. A very tight labour market is likely to be associated with accelerating wages growth relative to growth in productivity and hence rising labour costs. One measure which captures such changes in wage growth relative to productivity trends is Nominal Unit Labour Costs (NULCs).5
Chart 6 shows that NULCs grew by nearly 8 perácent in the year to the March quarter 1989. The average growth rate of NULCs was above 7áperácent for 7 of the 8 quarters between December 1988 and September 1990, which is the longest period above 7áperácent since the early 1980s (when NULCs growth reached as high as 18 perácent in 1982). This reflected both relatively high growth in wages, but also very poor growth in productivity, with labour productivity actually falling by 1Żáperácent in the year to the March quarter 1990.
Box 3: The dynamics of the labour market
An increase in unemployment is often considered to be a reflection simply of a loss of jobs, while a decrease in unemployment is interpreted simply as people finding jobs. However, the reality of the labour force is that jobs are both lost and gained in large numbers during any given month. There are also people moving in and out of the labour force, as well as growth in the potential labour force as the population increases. These movements in and out of employment and the labour force can occur for a variety of reasons, of which losing a job is only one. Other examples include people taking the opportunity for a holiday between contracts, occupations where employment is seasonal, or leaving the labour force to care for children.
During 2003 the number of unemployed persons in Australia fell from a seasonally adjusted level of 632,000 in January 2003 to 584,000 by the end of the year. Taken simply, this suggests that 50,000 people moved from unemployment to work. But breaking down these figures shows that in fact, in every month around 130,000 people move from unemployment to work, and around 92,000 move from employment to unemployment. All together there were 1.6ámillion movements from unemployment to work (either part-time or full-time), and 1.1ámillion movements from work to unemployment. The final change in unemployment further depends on flows between unemployment and those not in the labour force, and the growth in the population aged over 15.
It is clear that it is these flows, and the factors behind them, that are the most important force in determining the unemployment rate. In particular, small changes to these flows can have large impacts on the overall movement in the measured unemployment rate. Small disincentives to work or create jobs can have large consequences, as over time even small differences in flows can be reflected in large numbers of people being recorded as unemployed at given points in time.
While around 50áperácent of those currently unemployed were unemployed for less than 3ámonths in 2003, there was a significant group who had been unemployed for longer, with a fifth of people unemployed having been so for a year or more. So while these flows show that there are people constantly moving in and out of unemployment, there is also a group of long-term unemployed for whom re-entry into employment can be particularly difficult.
In contrast, the episode in 2000 saw a more modest increase in labour costs, and the period since 2002 has seen minimal pressure on labour costs to date. Between Juneá2000 and March 2001 NULCs growth increased from 2áperácent to 4Żáperácent. However, part of this acceleration was a return to more normal rates of growth, from the low and even negative rates experienced in 1998 through early 2000. Productivity growth also slowed somewhat, from the relatively high annual rates of 3átoá4áperácent through the late 1990s to an average annual growth rate of less than 1áperácent through 2000.
In 2002 and 2003 growth in NULCs was very modest at around 2áperácent peráannum, only slightly above the average for the 1990s, and trending downwards slightly in recent quarters. This is indicative of the combination of steady employment growth, a gradual downward movement in unemployment and a more flexible labour market helping to prevent excessive tightness in the labour market and the associated increases in labour costs. Another contributing factor was a strong productivity performance, with through-the-year growth rates averaging around 2áperácent over 2002 and 2003.
Chart 6: Growth in nominal unit labour costs
(through-the-year), 1984 to 2003
Source: Unit Labour Costs, Australian Government Treasury.
The lower and more stable growth in unit labour costs over recent years is another dividend from the improvement to the macroeconomic framework and labour markets, especially the lower inflation rates that have been sustained over the last decade or so compared to the inflation rates in the 1980s. Lower inflation, and lower inflation expectations, improves the wage bargaining process as both employers and employees can be more certain of the future path of prices.
Another indicator of emerging labour market tightness is rapid increases in the participation rate in the short term. The participation rate measures how many people in the working age population (defined in Australia as those 15 years of age or older) are looking for or have a job. When the participation rate rises rapidly in the short term, in the context of strong employment growth, it often indicates that the labour market is tightening.
Box 4: Industry distributions of employment
A decline in employment in one industry or sector, in relative or absolute terms, does not necessarily represent a decline in employment for the economy as a whole. In a flexible, adaptable economy, lower employment in one sector is generally offset by increasing employment in other sectors.
Chart 7: Change in share of total employment by industry
Source: Labour Force, Australia, Detailed (ABS Cat. No. 6291.0.55.001).
Chart 7 shows how the industry by industry shares of total employment have changed over the past 18 years. Some industries that have traditionally been large in Australia have declined relatively, such as manufacturing. And in their place employment has grown more strongly in the services sector.
One important note on these shifts is that they occur within the context of a growing labour market — between 1985 and 2003 the total level of employment grew by over 40áperácent. Most of the declines shown in Chart 7 are purely relative declines, and do not reflect levels of employment, only that the industry is not growing as fast as others. For example, while the share of total employment in wholesale trade has fallen by 1.6ápercentage points between 1985 and 2003, total employment in that sector has grown by around 22,000 jobs over the same period.
When the unemployment rate fell from 7áperácent to 6áperácent in 1989, the participation rate rose by around Żáof a percentageápoint. The episode in 2000 saw a similar increase in the participation rate over the period where the unemployment rate fell from 7áperácent to 6áperácent. But this fall in unemployment occurred over a longer period, suggesting that there may have been less tightness in the labour market.
In contrast, when the same decline in unemployment occurred in 2002 the participation rate was lower at the end of the period (although it did fluctuate upwards for 6 of the 18 months in question). This is partly a result of a steadier approach to low unemployment levels than in previous episodes.
The longer-term trends in Australia’s participation rates are largely driven by structural factors, rather than short-term labour market fluctuations. The most significant driver of the Australian participation rate over the past four decades or so has been the increasing participation of women in the labour force. Looking forward, participation rates will be strongly influenced by the ageing of the population, as discussed in the Intergenerational Report contained in the 2002-03 Budget Papers. 6
Other important structural factors include the incentive that individuals on welfare and other benefits have to enter employment, the education and skills attainment of those entering the labour force, and the options for retraining and further education available to those who have become unemployed. The influence of these factors is discussed in more detail later in this paper.
Another dimension to the issue of possible increasing tightness in the labour market, and hence the sustainability of the unemployment rate, is the impact of part-time work. The growing importance of part-time work reflects a significant structural change in the labour market in Australia. Australia has one of the highest rates of part-time work in the world, reflecting a relatively flexible labour market that allows many people to choose the number of hours of work that best suits their individual circumstances.
Since the mid-1960s the proportion of employees working part-time (that is, working less than 35áhours per week) has grown from around 9áperácent to around 30áperácent today. This increase reflects a long-term improvement in the ability of people to choose how they will participate in the labour force, and how many hours they will work.
Survey data indicate that around threeáquarters of part-time workers are not actively seeking more hours. The fact that the majority of part-time workers are satisfied with their existing hours of work is testament to the flexibility of Australia’s labour market, and its ability to deliver employment arrangements which are satisfactory for both employees and employers. Nevertheless, there is still a pool of part-time workers who would work more hours if such employment opportunities became available (about oneáquarter of part-time workers, and hence around 7áperácent of the total labour force). This could provide an additional ‘buffer’ against emerging wage and price pressures when there is rapid growth in employment.
Chart 8: Proportion of labour force working part-time
available for more hours of work, 1978-2003
Source: Treasury calculations based on Labour Force, Australia (ABS Cat. No. 6202.0) and Underemployed Workers, Australia (ABS Cat. No. 6265.0).
Chart 8 shows that the proportion of the total workforce that wish to work more hours has grown over the past 20áyears, but is a comparatively small proportion of the workforce.
Average hours worked
Chart 9 shows an alternative summary indicator of the overall level of tightness in the labour market, namely the average number of hours worked per week per person aged over 15 (including those not employed or not in the labour force). This measure captures the net effect of not just the unemployment rate per se, but also employment growth, the participation rate, average hours of work for both full-time and part-time employees, and the relative proportions of full-time and part-time employment. Although the series is much more volatile than the unemployment rate, the overall trends are similar to the unemployment rate (Chart 5). In particular, on this measure, there are peaks of greater than average tightness in the labour market in 1989 and 2000. In contrast, there is less evidence of tightness in the current episode of low unemployment, with the hours measure peaking at a lower level than in the earlier episodes and subsequently trending down.
Chart 9 also shows the movements in average hours per employed person. The broad trends are similar, although the peaks in 1989 and 2000 are more moderate relative to trend than in the overall population. During these peaks, the additional demand for labour was met more by additional employment than changes in the hours worked by existing employees.
Chart 9: Average hours worked, Australia
Source: Treasury calculations based on Labour Force, Australia (ABS Cat. No. 6202.0).
Regional unemployment rates
Labour market trends in particular capital cities might not reflect the labour market experience in other capitals or in more regional areas. For example, a period of labour market tightness, as measured by some of the overall national indicators, might reflect a very low unemployment rate in a few areas, such as Sydney and Melbourne, and higher unemployment elsewhere. In contrast, when the falls in the national unemployment rate reflect a relatively uniform fall in unemployment across most regions, there is less risk of bottlenecks and the associated price and wage pressures in particular areas or markets.
Historically, the capital cities have had a lower unemployment rate than regional areas. Chart 10 shows how these two unemployment rates have varied over the past decade and a half. The falling unemployment rate in 1989 and 2000 was relatively uniform across these two groups, and hence the gap between capital city and regional unemployment rates remained constant in both of these episodes. In contrast, during the most recent episode the most significant falls in unemployment have occurred in regional areas (notwithstanding the impact of the drought on farm employment), which has resulted in a closing gap between the capital city and regional areas. While unemployment rates in capital cities are currently somewhat above the lows of around 5╝áperácent reached in 1989, regional unemployment rates of just above 6áperácent are well below the rates of around 7áperácent reached in 1989 and 2000.
Chart 10: Unemployment rates for capital city and
Source: Treasury calculations based on Labour Force, Australia, Detailed — Electronic Delivery (ABSáCat.áNo. 6291.0.55.001). Capital cities includes all State and Territory capitals except for Darwin, Regional includes the balance of Australia.
International trends in unemployment rates
Australia’s current unemployment rate is not only low relative to the experience of recent decades, but also well below many other members of the OECD. The Euro area had an average unemployment rate of 8.8áperácent in 2003, with countries such as France and Germany experiencing unemployment rates above 9áperácent. Australia’s unemployment rate is currently over 1ápercentageápoint below the OECD average (in standardised terms), and has been below the OECD average since late 2001. This is a significant turn around since the mid-1990s, when Australia’s unemployment rate was as much as 3ápercentageápoints above the OECD average. Box 5 provides further detail on movements in unemployment rates in other key economies over recent decades.
That said, there are some countries with lower current unemployment rates than in Australia, such as New Zealand (averaging 4.7áperácent in 2003) and Ireland (averaging 4.6áperácent in 2003). The unemployment rate in the United States in 2003 was comparable to Australia, at around 6áperácent. However, this probably represented a cyclical peak for the unemployment rate in the United States reflecting the weak United States economy, while as discussed above, it represented a 23 year low in Australia. This suggests that the longer term or structural unemployment rates in the United States may still be below those for Australia.
Box 5: International trends in unemployment rates
Chart 11: Standardised unemployment rates in selected countries
Source: OECD Main Economic Indicators.
The extent to which unemployment rates fell in the 1990s differed across countries. While the UK unemployment rate fell from over 10áperácent in 1992 to less than 5áperácent in 2003, the German unemployment rate has stayed between 7áand 10áperácent since January 1993. Moreover, the long-term unemployed account for as much as oneáthird of the total unemployed in Germany, reflecting structural problems in the German labour market.
Japan has been an exception to the general trend of falling unemployment rates during the 1990s. Prolonged weaknesses in aggregate demand has resulted in a doubling of Japan’s unemployment rate from the 1980s average of 2Żáperácent to the current 5 per cent.
A low unemployment rate may not always present a true picture of the labour market. While the rise in the US unemployment rate during the recent downturn has been modest by historic standards, the US labour market has weakened significantly, shedding 1.95 million jobs since March 2001 with the participation rate falling markedly. Similarly, the sustained fall in the Dutch unemployment rate during the 1980s and the 1990s happened in conjunction with a falling participation rate, which suggests that there may have been an increase in the number of ‘discouraged’ workers.
More generally, the fact that unemployment rates are lower in some other developed economies than in Australia suggests that there is scope to lower unemployment further in Australia over the medium term, given appropriate policy settings. This is discussed in more detail in the following section.
Continued next page...
1 Care should be taken in interpreting this chart. The definition of unemployment has changed over time, and the estimates prior to the Second World War are relatively unreliable.
2 Further discussion of this structural increase can be found in J. Borland and S. Kennedy (1998), ‘Dimensions, Structure and History of Australian Unemployment’, in G. Debelle and J. Borland (eds), Unemployment and the Australian Labour Market, proceedings of a conference held by the Reserve Bank of Australia and the Australian National University, Sydney.
3 Further discussion of the ‘Phillips Curve’ in Australia is contained in D. Gruen, A. Pagan and C. Thompson (1999), ‘The Phillips Curve in Australia’, Journal of Monetary Economics 44(2), pp. 223-258.
4 See Budget Paper No. 1, Budget Strategy and Outlook 2002-03, Statement 4, pp. 4-10 to 4-11.
5 Real Unit Labour Costs can also be used to assess changes in labour costs taking into account the effects of inflation, although the trends are broadly similar to those for NULCs.
6 Australian Government (2002), Budget Paper No. 5, Intergenerational Report 2002-03.