The role of price and nonprice factors in predicting Australia's trade performance

Abstract

This article investigates the role of price and nonprice factors in predicting Australia's trade performance. Results broadly suggest that Australia's trade performance is largely explained by the nonprice factors namely, R&D, reliability of domestic supply, aggregate world demand and Foreign Direct Investment (FDI) flows in long run. Price factors such as, relative price of Australian exports and domestic prices are also important predictors of trade competitiveness. The policy implications of these findings are that there are dividends in terms of improved trade performance by encouraging R&D expenditure, attracting FDI, improving domestic supply and implementing appropriate policies to improve price competitiveness.

Keywords:

• trade performance
• R&D
• openness
• FDI

JEL Classification:

• F49
• P33

Acknowledgements

The authors are grateful to Oilver Morrissey, Chris Milner and Richard Kneller for very useful suggestions in a seminar at the School of Economics, University of Nottingham. The authors also thank Premachandra Athukorala, Edward Oczkowski and Nada Kulendran for useful advice. An anonymous referee and the editor of this journal have also made useful suggestions. Financial assistance from a CSU Grant is gratefully acknowledged. All remaining errors are ours.

Notes

¹ Compared to other advanced countries, Australia invests a small fraction of Gross Domestic Product (GDP) on R&D. For instance, in 2002 Australia spent 1.64% of GDP on R&D as against 2.65% by USA and 3.12% by Japan (OECD, 2007).

² See, for example, Fagerberg (1988), Greenhalgh (1990) and Anderton (1999).

³ Literature is too many to cite. See Anderton (1999) and references cited therein.

⁴ This literature is now growing rapidly and is too large to cite. However, discussions about the role of nonprice factors in determining trade flows can be found in Owen and Wren-Lewis (1993), Greenhalgh (1990), Greenhalgh et al. (1994), Anderton (1999) and Salim and Bloch (2009).

⁵ Magnier and Toujas-Bernate (1994) argue that models that include both factors perform better in predicting trade flows.

⁶ In international trade theory this phenomenon is known as ‘product life cycle’ theory where it is argued that when a new product is innovated, innovative firms enjoy monopoly profit until it becomes standardized and reaches its maturity. After this point, product is produced using standard technology in developing countries and innovative firms lose its competitiveness in this particular product group.

⁷ Ioannidis and Schreyer (1997) provide classification of industries into high- and low-technology intensive based on the R&D intensity data. Accordingly chemicals, nonelectrical machinery, professional and scientific goods, and office and data processing machinery, electrical machinery and transport equipment are classified as high-technology-intensive industries; whereas food, textile, wood products, paper products, nonmetallic goods, basic metals and metal products are classified as low-technology intensive industries (see Table A.1 in Appendix A for the classification of industries included in this study).

⁸ Since most Australian industries are small to medium in size by international standards, they do not tend to invest much on R&D, instead they rely largely on Foreign Direct Investment (FDI) as a means of acquiring advanced technology to compete in both domestic and export markets. This is evidenced by massive inflows of FDI to Australia in recent years (Sharma and Bandara, 2010).

⁹ We are aware that there are several problems in using R&D data and they may not accurately capture the nature of technological advancement and innovation for the following reasons (Hughes, 1986). First, they are input measure, which do not tell us how much has been actually translated into product and process innovation. In other words, not all R&D is commercially successful. Second, since, the effects of R&D are cumulative; ideally a stock measure should be applied. However, because of the measurement problem (such as depreciation) the stock data are not reliable. Despite these limitations, our results will provide useful insights about the role of innovation in explaining trade performance.

¹⁰ The precise way of deriving the long-run effects include adding short-run and lagged effects, which is then divided by (1-lagged dependent variable). For details of the dynamic modelling approach, see Baltagi (2005) and Gujarati (2003).

¹¹ Note that we have 70 time series and 34 cross-sectional observations.

¹² Our cross-sectional results – which are not appended to this article due to space constraint – tend to suggest that R&D is important even for industries such as, leather and leather manufacturing. While this finding is somehow surprising, given the nature of leather goods – which are income elastic – firms tend to undertake R&D to develop new products to remain competitive.

¹³ However, we find that in intermediate term protection of domestic market retards trade performance as expected and this relationship is statistically significant.