RDP 2005-01: Long-Term Patterns in Australia's Terms of Trade 1. Introduction

The terms of trade are an important determinant of economic welfare since they govern the quantity of foreign goods that can be purchased with a given amount of domestic output.[1] Changes in this key relative price can have large dynamic consequences for macro variables – notably consumption, savings and investment.[2] Shocks to the terms of trade will also have substantial distributional implications within a country as relative incomes and prices change.

This paper considers two important properties of the Australian terms of trade using 135 years of data: its level and its volatility. Australia largely exports commodities and imports manufactured goods. Over the past century commodities have averaged over four-fifths of Australia's goods exports, while manufactures have constituted a similar proportion of Australia's imports. The Prebisch-Singer hypothesis, that commodity prices decline relative to manufactures prices, suggests that Australia will experience a long-run decline in its terms of trade. Indeed, over the past 135 years, the terms of trade for goods has declined 12 per cent, as seen in Figure 1. From the mid 1950s there was seemingly a stronger negative trend, at least until the mid 1980s. However, since then the terms of trade have improved quite substantially. Indeed, based on projections for the terms of trade using known commodity contract prices, the terms of trade are likely to have increased by around 50 per cent over the period 1987–2006, unwinding the decline over the preceding 30 years.

Figure 1: Australia's Terms of Trade

The other notable aspect of the terms of trade is their changing volatility, as seen in Figure 1. From the end of World War I until the mid 1950s the terms of trade experienced many large shocks, frequently doubling or halving within a matter of years. Such abrupt relative price changes have the potential to cause severe economic dislocation. It is important to know why volatility was so much higher in this period and why it has declined. Understanding this can help us to consider whether a return to such volatility is likely.

The remainder of the paper is structured in two parts, considering first the level of the terms of trade, then changes in the volatility of the terms of trade. In the first part, Section 2, we test whether there is a trend in Australia's terms of trade and assess its magnitude. We then consider export and import developments separately, in particular their changing composition, and how these have influenced the terms of trade. In the second part, Section 3, we address the volatility of the terms of trade by testing whether it has changed in a significant way. To understand the changes we also separately consider the volatility of export and import prices. Because it has experienced greater change, we focus on the export price series and examine whether the main contributor to the return to low volatility in recent decades has been diversification or a decline in price shocks to individual commodities. Finally, Section 4 concludes.


Though beyond the scope of this paper, trends in productivity are another important consideration for any welfare analysis. For example, a trend decline in the terms of trade may also be associated with stronger productivity growth in the export sector, which would have an offsetting welfare impact. [1]

For example, the traditional Harberger-Laursen-Metzler effect states that domestic saving declines in response to a temporary negative terms of trade shock due to intertemporal consumption smoothing. But a more sustained shock may reduce investment and potentially increase saving in anticipation of lower future output, as Kent and Cashin (2003) demonstrate empirically. [2]