The comprehensive reforms initiated in the mid-1980s, and accelerated in the 1990s, are expected to bring significant gains to the Australian economy over the medium term. Empirical studies have estimated the benefits of the various reforms to be in the range of 5–10 percent of GDP, if fully implemented (see Box 8.1). Dividends from the reforms are already apparent. Australia is currently experiencing a sustained economic expansion with per capita real GDP growth having averaged 2¾ percent since 1992. The strong growth performance gives rise to two main questions. Have productivity growth and potential growth accelerated compared with the past? What is the medium.-term outlook for potential growth?
Acceleration of Productivity Growth
Productivity growth is difficult to assess over a short horizon because it depends critically on the cyclical position of the economy. Productivity is procyclical: it falls sharply in recessions when output growth declines, and rises strongly at the outset of a recovery. Therefore, one cannot simply compare productivity in a given year with the year before, but must compare productivity growth at present with the same stage of earlier economic cycles, or construct averages around entire economic cycles.
Staff estimates, based on analysis over the economic cycle, show a significant productivity improvement compared with the past. The production function analysis in Chapter 2 shows that average growth in total factor productivity has increased from 0.8 percent during the previous cycle (1982–88) to 1.3 percent during the latest cycle (1989–96). Estimates by the Australian Bureau of Statistics (ABS, 1997) confirm the pattern, with 1 percent and 1.2 percent average annual productivity growth in the two most recent cycles, respectively. Total factor productivity growth over the cycle, however, remains well below the growth rate of about 2½ percent experienced in the “golden age” (1967–73).
Modeling the Gains from Reform
A number of recent studies have estimated the economy-wide gains to Australia from tariff, labor, and product market reforms. It is widely accepted that these models have inherent limitations because of the difficulty in capturing real world adjustment processes. Significantly, though, all of the model estimates showed large positive returns to the reforms.
The Industry Commission (1990) modeled reform in transport, aviation, communications, water, and electricity, contracting out by governments, and the removal of rural manufacturing assistance. The results suggested a long-term increase in real GDP of 6.5 percent.
The Bureau of Industry Economics (1990) estimated the benefits from reform over the seven-year period from 1988/89 to 1994/95. In addition to reforms examined by the Industry Commission, the Bureau of Industry Economics included the impact of investment incentives and labor market reforms assumed to result in large increases in labor productivity. The study found that the reforms would increase GDP by 9.5 percent.
For the Economic Planning Advisory Council (EPAC), Filmer and Dao (1994) analyzed the effects of the wide set of reform measures undertaken in the late 1980s and early 1990s (prior to the introduction of the National Competition Policy). They estimated reform would eventually increase GDP by 7.5 percent, of which 3.5 percent stemmed from higher private sector productivity; 0.8 percent from the direct effect of tariff reductions (with indirect effects included in the increase in productivity); 1.6 percent from higher general government services productivity; and 1.5 percent from infrastructure reforms, including those to government business enterprises. The study also estimated that potential growth could be as high as 4½ percent over the period 1995/96 to 2000/01 before returning to the baseline of 3 percent thereafter. Moreover, the benefits of the reform would be spread widely throughout the economy. All of the model’s 25 industries experienced reform-induced increases in output.
The Industry Commission (1995) estimated the growth and revenue implications of the Hilmer and related reforms (the basis for the National Competition Policy reforms). Specifically, the Industry Commission modeled the impact of reforms in the transport, communications, and utilities sectors, and to statutory marketing arrangements, government services, unincorporated enterprises, and anticompetitive legislation. The Industry Commission results suggested that over time there would be a gain in real GDP of 5½ percent.
The improvement in productivity growth in the 1990s is due to a better labor productivity performance. Growth in labor productivity more than doubled to an estimated 1.7 percent per annum during the latest economic cycle, compared with 0.7 percent in the previous cycle (ABS, 1997). Capital productivity growth, however, slowed some what—to 0.3 percent during the latest cycle, from 1 percent in the 1980s—as a result of a rapid growth in the capital stock.1
Reforms of the public enterprise sector and tariff regime appear to have made a significant contribution to the increase in productivity. Since 1990, labor productivity levels in the public enterprise sector have risen by about two-thirds (Commonwealth Government of Australia, 1997, p. 19). Productivity improvements have been spurred by the increased openness of the economy, primarily as a result of the tariff reforms. Exports and imports as a share of GDP have risen from about 30 percent on average for the 1980s, to more than 36 percent in the two years to 1996/97.
Moreover, there are reasons to believe that the reforms have had further benefits on productivity that have not been reflected in the official data, owing to measurement problems. For example, the financial sector was excluded from the production function analysis in Chapter 2 and in the Australia Bureau of Statistics analysis of multifactor productivity because output in the financial sector was, until very recently, measured by the Australian Bureau of Statistics in terms of labor inputs, so that labor productivity has by definition been constant. Industry studies, however, show large increases in output per worker, partly as a result of deregulation (Oster and Antioch, 1995). Similarly, deregulation of trading hours in the retail and wholesale trade sector is likely to have improved the quality of service to customers and increased efficiency. But as the measure of output is indexed to sales (with no account taken of improved service) the extended shopping hours may have actually decreased measured labor productivity.
Acceleration of Potential Growth
Potential growth appears to have accelerated during the latest economic expansion. Estimates of the production function framework presented in Chapter 2 show that average potential growth has accelerated to around 3½ percent over the past three years from an average of 2½–3 percent in the late 1980s. This acceleration traces back to an increase in productivity, as outlined above, and an increase in production inputs, in particular capital inputs. The present expansion has been marked by a strong investment performance, with real investment growing on average by 4¼ percent per capita in the period 1993–96.
The strong investment performance has been partly in response to the structural reforms. The link is clearest in the communications sector, where, largely as a result of the reform, real output grew by over 85 percent in the six years to 1996, with a significant increase in investment (particularly in plant and equipment). This investment, combined with private sector investment in infrastructure (also spurred by reform of the utilities sector) and Olympics-related investment, has made a strong contribution to the recent growth in total investment.
What Is the Medium-Term Outlook for Potential Growth?
The medium-term outlook for potential growth hinges critically on whether the government succeeds in its efforts to raise national saving and on the productivity benefits of its past and current reforms. To truly sustain higher growth rates, national saving needs to be raised sufficiently to avoid excessive borrowing.
If the recent productivity gains are sustained, breaking the long-term downward trend in total factor productivity growth, the rate of total factor productivity growth over the medium term can be assumed to be about 1 percent per annum on average. This allows for a small decline from present rates of about 1¼ percent as the expansion matures further, but is still significantly higher than the rates of about ½ percent experienced in the late 1980s, Together with contributions from labor and capital inputs, this implies that potential growth can be expected to average around 3½ percent per annum over the medium term.2
This growth rate is more conservative than implied by the studies of the gains from reform discussed earlier. For instance, the Filmer and Dao (1994) study estimates potential growth rates as high as 4’/2 percent through 2000, more than 1 percent above that implied by the production function. Moreover, the most recent Australian Bureau of Statistics trend estimates suggest that the underlying rate of total factor productivity growth was running at around 2 percent per annum in the three years to 1995/96. This may imply some upside to the estimates of potential growth.
Conclusion
For a long time, Australia’s growth performance relative to other OECD countries has been disappointing. This has apparently changed, as the comprehensive structural reforms embarked on in the mid-1980s and accelerated in the 1990s have begun to bear fruit. The reforms have opened the economy and made factor and output markets more flexible and input use more efficient. The results so far have been an acceleration not only of actual growth, but also of productivity and potential growth. The long-term downward trend in total factor productivity seems to have been broken, and productivity growth (particularly for labor) is now relatively high, both compared with the past and with other industrial countries. The current economic expansion has been well balanced, unusually prolonged, and accompanied by only moderate price pressures. Nevertheless, despite strong output growth, the unemployment rate has remained stubbornly high. The challenge ahead will be to continue the current reform effort and implement additional reforms across a wide range of fronts, in order to sustain high potential growth rates and reduce unemployment significantly.
References
Australian Bureau of Statistics, 1997, Australian National Accounts Multifact Productivity, 1995–96, ABS catalog No. 5234.0 (Canberra).
Commonwealth Government of Australia, 1997, Budget Papers, Statement 3 (Canberra).
Bureau of Industry Economics, 1990, Microeconomic Reform and Structure of Industry, Discussion Paper No. 9 (Canberra).
Filmer, R., and D. Dao, 1994, Economic Effects of Microeconomic Reform, EPAC Background Paper No. 38 (Canberra: Economic Planning Advisory Council).
Industry Commission, 1990, Annual Report 1989–90 (Canberra).
Industry Commission, 1995, The Growth and Revenue Implications of Hilmer and Related Reforms (Canberra).
Oster, A., and L. Antioch, 1995, “Measuring Productivity in the Australian Banking Sector,” in Productivity and Growth, proceedings of a conference held at the Centre for Financial Studies, Kiribilli on July 10–11, 1995 (Sydney: Reserve Bank of Australia).
As discussed in Chapter 2, this slowdown is likely to be overstated because of the measurement bias due to the growing importance of investment in computer equipment. Consequently, capital productivity growth may be higher than estimated.
This assumes national saving is sufficient to permit a modest rise in the ratio of investment to GDP over the medium term without an excessive buildup in net external liabilities. The labor input is then projected by using information embodied in the historical pattern of the capital-labor ratio. In Australia, there has been a slight trend increase in the capital-labor ratio since the late 1960s, from less than 0.2 in 1968, to around 0.3 in the early 1990s, Under the assumption that over the next few years this ratio rises by about 1 percent per annum (to just less than 0.33 in 2000), labor inputs can be derived directly from the projections for the net capital stock.