Revised: December 2003 Sources of Australia’s Productivity Revival



Download 238.16 Kb.
Page1/3
Date conversion04.05.2016
Size238.16 Kb.
  1   2   3
Revised: December 2003

Sources of Australia’s Productivity Revival

DEAN PARHAM*

Productivity Commission

Canberra, ACT, Australia





  1. Introduction

After languishing in the late 1970s and 1980s, Australia’s productivity growth surged in the 1990s. Labour and multifactor productivity growth reached record highs, surpassing the ‘golden age’ rates in the 1950s to early 1970s. The productivity revival was significant. It shifted the economy away from its traditional reliance on factor accumulation as a source of growth (Parham 2000). It underpinned strong growth in GDP at an average of nearly 4 per cent a year. And it fully accounted for an uplift in per capita income growth from 1.6 to 2.6 per cent a year (Parham et al 2000).
The objective of this survey is to identify the key sources of Australia’s productivity revival. First, the survey establishes a set of ‘stylised facts’ about Australia’s productivity performance. Second, the survey draws on available evidence — as far as possible from formal analytical studies — to explain the stylised facts.
Since the survey focuses on the 1990s productivity revival, it does not attempt to do justice to contributions in the broader field of theoretical and applied productivity and growth analysis. Readers with a broader interest are referred to Rogers (2003b) for a survey of economic growth and to Dawkins and Rogers (1998) for a survey of Australian productivity analyses.
Before proceeding, a clarification of productivity variables of interest is in order. Most interest usually lies in labour productivity (LP) and multifactor productivity (MFP). Growth in LP (output per hour worked) holds interest because, at the aggregate level, it is highly correlated with average income growth. It also affects growth in unit labour costs. LP growth is influenced by capital deepening (increases in the capital-labour ratio) and MFP growth. MFP growth is of interest because it reflects improvements in efficiency, broadly defined. It is sometimes thought of as the unexplained growth residual or (inappropriately) as ‘manna from heaven’. But it occurs for a reason — investments in R&D, improved management techniques, improved skills and so on. In principle, where such influences are embodied in inputs, their effects can be factored out from the growth residual by specifying and allowing for growth in the relevant input. Some studies have factored out a few of these influences, in which case they continue to be reflected in LP growth, but not MFP growth. Consequently, the interpretation of MFP growth depends on the input specification.


  1. What is different about the 1990s productivity experience?

This section puts Australia’s recent productivity trends in historical and international perspective and identifies stylised facts that distinguish the 1990s from earlier decades.
The principal source of productivity estimates is the official Australian Bureau of Statistics (ABS) series of labour, capital and multifactor productivity for the market sector — the part of the economy for which output and therefore productivity can be relatively well measured.1 In 1999, the ABS introduced changes to its methodology that included a switch in the capital measure from the traditional wealth-based measure of net capital stock to a capital services measure.2 The capital services measure takes account of the loss of economic efficiency of assets over time and the relative marginal products of different asset types.3 Changes in the composition of capital, for example a switch to short-lived assets with high marginal products, means that growth in capital services departs from growth in the traditional capital stock measure. Use of the capital services measure factors out efficiency–related composition effects from MFP growth. To illustrate its practical significance, the switch to a capital services measure has reduced estimated MFP growth by probably half a percentage point or so in the 1990s.4
i) Around one percentage point more productivity growth

The ABS estimates indicate that productivity accelerated by just over one percentage point during the 1990s. Table 1 shows the underlying rates of LP growth and their decomposition into a capital deepening contribution and an MFP growth contribution. The ABS determines these underlying rates by calculating average growth between the same point in successive productivity cycles — from productivity peak to productivity peak. Comparison of the mid-1990s cycle and the previous cycle reveals an acceleration of around 1.1 percentage points in both LP and MFP growth.


Estimates for the most recent years (up to 2002-03) do not form a complete cycle and therefore have not been included in Table 1. The ABS has not determined a productivity peak beyond 1998-99. The average rate of productivity growth has slowed since 1998-99, but it is not yet clear whether, or to what extent, this represents a slowing in the underlying rate rather than a number of short-term factors such the post-GST pause in housing construction in 2000-01, the drought and the global downturn.

Table 1 Growth in productivity and contributing factors (% per year)



LP

MFP


Capital deepening

Output


Capital services

Hours worked

1964-65 to 1968-69

2.5

1.2

1.3

5.1

6.0

2.5

1968-69 to 1973-74

2.9

1.6

1.3

4.6

5.1

1.6

1973-74 to 1981-82

2.4

1.1

1.4

2.1

3.6

-0.3

1981-82 to 1984-85

2.2

0.8

1.4

1.8

3.6

-0.4

1984-85 to 1988-89

0.8

0.4

0.4

4.1

4.3

3.2

1988-89 to 1993-94

2.0

0.7

1.3

1.8

3.1

-0.2

1993-94 to 1998-99

3.2

1.8

1.3

4.6

4.7

1.3

1964-65 to 1998-99

2.4

1.1

1.2

3.4

4.3

1.0

Source: ABS Cat. no. 5204.0, except the capital deepening column, which contains Productivity Commission estimates.
A number of studies have used other methods to determine the existence and magnitude of an MFP acceleration (Appendix 1). They suggest an acceleration in the range of 0.5 to 1.4 percentage points, although the low end estimate is likely to be biased downward. Overall, the studies provide strong support for an acceleration of around 1 percentage point or so.
ii) Improved efficiency rather than additional capital deepening

Table 1 also shows that the LP acceleration was due to stronger MFP growth or improved efficiency, rather than additional capital deepening. It is of interest to note that the rate of capital deepening has been stable over nearly all productivity cycles so that, historically, variations in LP growth have been due to variations in MFP growth.


iii) The productivity shift commenced in the early 1990s

The ABS estimates imply that the productivity revival got underway from 1993-94. It is difficult to disentangle the start of the structural shift from the effects of recovery from the 1990-91 recession. Some studies suggest that the shift may have started earlier, but the general view is that it was established by 1993-94 (Appendix 1).


iv) From international productivity growth laggard to frontrunner

The 1990s also brought a change in a long-term pattern in which Australia performed poorly compared with other high-income countries (Thimann 1998). Australia’s international productivity ranking slipped markedly over the 20 century. It actuallyslipped most during the ‘golden age’ of the 1950s to early 1970s, when productivity growth was historically strong in Australia, but not as strong as it was in other countries.th Australia’s ranking among OECD countries on GDP per hour worked slipped from 4 in 1950 to 12 in 1973, and to 16 by 1990.5 In the 1990s, however, Australia posted strong growth in the midst of mixed performance among OECD countries (OECD 2001b; Gust and Marquez 2001). Productivity growth accelerated in the US, Canada and a number of smaller economies including Australia, but slowed in France, Germany and Japan. Although Australia did not improve its productivity ranking, it did outperform the resurgent US to lift GDP per hour worked from 76 per cent of the US level in 1990 to 83 per cent in 2002. Finland and Australia had the strongest acceleration in trend MFP in the 1990s in the OECD area (OECD 2001b).


v) New service industry contributors

Industry productivity trends provide some insight into the sources of the aggregate productivity surge. However, the ABS publishes estimates of industry LP but not MFP.6 The Productivity Commission has published industry MFP estimates according to a commonly-used method, in which industry value added is related to inputs of labour and capital.7 These estimates may be subject to some error, but can be relied upon for broad trends.


The industry estimates indicate that the productivity surge has been broadly based. Table 2 shows the underlying growth rates in industry MFP. The raw series have been smoothed to circumvent industry-specific differences in productivity cycles. A new set of service industries — wholesale trade and finance & insurance in particular, but also retail trade, construction and transport & storage — contributed to the acceleration. Agriculture also contributed. But other traditional contributors — mining and manufacturing — did not.

Table 2 Growth in trend MFPa, by industry, over aggregate productivity cyclesb

Per cent per year




1974-1975 to 1981-1982

1981-1982 to 1984-1985

1984-1985 to 1988-1989

1988-1989 to 1993-1994

1993-1994 to 1998-1999

Agriculture

1.6

1.1

1.4

2.6

4.3

Mining

-1.7

0.5

2.6

2.5

1.2

Manufacturing

2.1

1.8

1.7

1.6

1.3

Electricity, gas & water

2.0

3.2

4.2

3.7

1.8

Construction

1.4

0.4

-0.3

-0.2

0.4

Wholesale trade

-0.7

-0.9

-0.5

1.2

3.2

Retail trade

1.0

0.6

-0.2

0.1

1.0

Accom. cafes & restaurants

-0.9

-1.3

-1.9

-1.6

-0.3

Transport & storage

2.2

1.2

1.0

1.4

1.9

Communication services

6.5

4.9

4.8

4.9

3.7

Finance & insurance

-2.0

-1.0

0.2

0.7

0.8

Commun. Rec. services

-1.4

-2.2

-2.9

-3.1

-3.3
  1   2   3


The database is protected by copyright ©essaydocs.org 2016
send message

    Main page