• No results found



Academic year: 2023



Loading.... (view fulltext now)

Full text

The correlation of Australian production with that of the OECD and particularly the United States is well documented. This article explores the foreign linkages by looking at the production side of the national accounts for Australia and the United States, which are often characterized as technological frontier countries. The industrial structures in the two countries are broadly similar, and about two-thirds of Australian manufacturing is linked to the US.

It is now a well-known feature of the Australian economy that the domestic business cycle is highly correlated with that of the OECD, and that of the United States in particular (McTaggart and Hall 1993; Gruen and Shuetrim 1994). Gruen and Shuetrim (1994), Debelle and Preston (1995) and de Roos and Russell (1996) attempted to explain this relationship using data from the expenditure side of the national accounts. The strength and nature of the relationship between domestic and US output is explored using cointegration analysis in Section 3.

A number of papers have pursued an explanation of the strong contemporary and long-run relationship between Australian and OECD/US manufacturing. Gruen and Shuetrim (1994) attempted to explain the correlation of Australian and foreign output in terms of the strength of foreign demand for Australian goods and services. They report that the structures of the manufacturing sectors in Australia and the United States are similar, although Australian manufacturing uses more intermediate inputs in production and thus has a lower value added.

Table 1: Summary Statistics on Australian and US Sectoral Output
Table 1: Summary Statistics on Australian and US Sectoral Output

A Closer Look at Sectoral Links

Overall, developments in the United States are relevant to assessing the outlook for Australian sectoral output. In manufacturing, the long-term sectoral linkages arise in the production of wood products, paper-related products, chemicals, non-metallic minerals, basic metals, fabricated metals, other machinery and miscellaneous manufacturing. The sectors for which developments in the United States are not important in the long run are usually those where domestic aggregate demand is important.

For the production of textiles, clothing, wood, paper and fabricated metals, domestic demand increases sector output in the short run, and the domestic aggregate effect multipliers are relatively large. Output in most manufacturing subsectors can be characterized as being tied to the corresponding US sector in the long run, but significantly affected by aggregate domestic demand in the short run. But the claim that foreign connections are related to the sector's foreign penetration does not seem to be supported by these data.6 Chemicals and paper production, for example, have strong ties to US manufacturing, but their foreign ownership was very different in the mid-1980s.

Similarly, foreign ownership in the transport equipment sub-sector is high, but there is no obvious correlation with US manufacturing, despite the US being the largest foreign investor in the transport equipment sector in Australia. Air transport, for example, is the only service sector that is sensitive to the corresponding US sector in the long run. The latter, like services, appear to be produced for the domestic market and largely unrelated to sectoral conditions in the US (although food and textiles are affected by aggregate US demand).

The transport connection also only occurs in the air sub-sector, which is the transport sub-sector where change and technology diffusion have been the fastest. As the production frontier moves out and innovation in goods takes place, changes in the US and elsewhere are transferred to Australia. The communications sector in Australia, for example, has grown much faster than that of the United States since the mid-1980s, likely due to catch-up following liberalization.

This raises the question of why there is a persistent gap in productivity levels between the United States and elsewhere in the face of continued technology transfer. It is not the case that productivity is higher in industries where the respective sectoral output in the United States is the main driving force. The oddity is perhaps explained by the nature of the production processes in both countries (Ergas and Wright 1994).

Capital stock in Australia may also be older than that in the United States, as plant and equipment are less frequently updated.

Table 2: Australian and US Sectoral Output Error-Corrections (1977-1993)
Table 2: Australian and US Sectoral Output Error-Corrections (1977-1993)

Some Implications

  • New Information?
  • The Relative Importance of Foreign and Domestic Influences
  • Controlling for Other Influences
  • Aggregate or Compositional Effects?

As the above analysis showed, while external output is important in some sectors, so are internal impacts, especially in the short term. It is a plot of the contemporaneous correlation of Australian single-digit sectoral growth rates with growth in total Australian output (excluding the specified sector) and with growth in the corresponding US sector. When the result is above the 45 degree line, Australian sectoral output is more affected by contemporaneous events in the domestic economy than in the corresponding US sector.

Growth in Australian sectoral output appears to be more correlated with growth in the rest of Australia than with growth in the corresponding industrial sector in the United States, at least on a contemporaneous basis. Stockman (1988) reported that changes in industrial production in European countries appear to be related to what is happening in the home country rather than what is happening to industry in a range of other countries. More formal tests fail to reject the hypothesis that Australia exhibits the same relationship as the other countries in the sample.

The time series analysis in Section 3 indicated that the United States is central in the long run, either because of aggregate demand effects or because of direct sectoral linkages. Unfortunately, these other variables could not be included in the analysis in Section 3, because there were too few degrees of freedom. When the manufacturing equations in Table 2 are estimated with US total output in place of US sectoral output, the explanatory power of the equations usually drops significantly, the dynamics terms all become insignificant, and the long-run coefficients become less significant, and in the case of wood products and furniture, insignificant.9 Shifting the specification to total US output itself significantly weakens ties to the United States.

Lags of domestic aggregate production in the equations are relatively short, while lags of foreign production are relatively long. In the simple modeling framework used in this paper, there are three possible transmission pathways through which policy affects sectoral outcomes. In many manufacturing industries, the policy affects output in the same quarter or with only one lag, unlike total output, where the effect appears to last up to two quarters.

Direct effects are not found in wood and furniture, paper, printing and publishing, chemicals, or miscellaneous manufacturing, as the policy effect in these industries is fully captured in the exchange rate and income channels.10 The indirect effect of income is the strongest. in food and metal products, while the exchange rate effect is strongest in non-metallic mineral products and machinery. This suggests that shocks to manufacturing from abroad can have multiplier effects on aggregate output, at least in the short run.

Figure 3: Correlation of Australian Industrial Output Growth
Figure 3: Correlation of Australian Industrial Output Growth


If sectoral linkages with foreign production explain movements in aggregate output, then we expect a positive coefficient on the shock variable: an increase in output due to a foreign shock increases aggregate output. A positive coefficient is found, and the magnitude of the coefficient is greater than the share of manufacturing activities in total production, although this difference is not statistically significant. This evidence supports our hypothesis that sectoral linkages play a role in explaining movements in total Australian GDP.

But the foreign connections for the bulk of manufacturing are not in the general economic development of the United States, but with what is happening in the corresponding American sector. The implication, then, is that these links with the US, which is the world leader in productivity and innovation, are driven by the supply side. This result appears to be robust to alternative specifications and is unexplained by institutional features such as foreign ownership or trade intensity.

These shocks also affect aggregate output, and so the analysis is identifying something more fundamental than just compositional changes in aggregate output. The importance of the US economy should not overshadow the result, however, that politics and events in the domestic economy are crucial, especially in the short and medium term. Indeed, even in simple modeling exercises, monetary policy can have large effects on various industries.

Interest rate changes have both direct and indirect effects on sector output, the latter being due to the effect of an interest rate change on the exchange rate and on the aggregate demand state. These particular effects vary across different manufacturing industries, with the manufacturing sectors with higher import quotas more affected by the exchange rate and the more domestically oriented sectors more affected by domestic demand. While developments in the United States have persistent, long-term effects on local manufacturing output, domestic policies affect local aggregate demand, which in turn can have a large impact on output in the short term.

Appendix A: Description and Sources of Data

Production Data

Capital Stock Data

Trade Concentration Data

Wood and furniture Wood and wood products; furniture and equipment Paper, printing and publishing Paper and related products; printing and publishing Chemicals, petroleum and coal Chemicals and related products; oil and coal.

Table A.1: Reconciliation of Australian and US Data
Table A.1: Reconciliation of Australian and US Data

Appendix B: Estimation Methodology for the Error-Correction Equations

Appendix C: Panel Data Estimation Method

To test whether Australia showed a significantly different pattern of shocks to other countries, a set of dummy variables was inserted to select each of the Australian observations. Numbers in parentheses ( ) are standard errors; the number in brackets { } is the p-value of the joint significance of the real exchange rates.

Table C.3: Impact of Manufacturing Industry Shocks on Aggregate Output Dependent variable:  ∆ y t aggregate , Sample period: 1980:Q3 to 1994:Q4
Table C.3: Impact of Manufacturing Industry Shocks on Aggregate Output Dependent variable: ∆ y t aggregate , Sample period: 1980:Q3 to 1994:Q4

Appendix D: Alternative Models Of Manufacturing Output


Table 1: Summary Statistics on Australian and US Sectoral Output
Figure 1: Real Sectoral Output Index: 1977-1993 = 100
Figure 2: Real Manufacturing Output Index: 1977-1993 = 100
Table 2: Australian and US Sectoral Output Error-Corrections (1977-1993)


Related documents

When they 3 test their solution with a searchString and dataString of different lengths the following output is produced: searchString dataString Actual output Fish Fishcake

† these percentages are the readings for recycling in general †† these percentes are the readings for public halls only ** these percentages are the averaged readings for sportsfields