RDP 2004-11: Trade Openness: An Australian Perspective 5. Conclusions

Our openness regressions show that it is possible to explain openness well with only a few variables. The most significant variables are not unexpected – countries with small populations tend to trade relatively more and countries that are located closer to potential trade partners also trade more. The significance of these two variables would be consistent with several different models of trade (see Haveman and Hummels 2004), so our results do not really help to discriminate between different models. We find only modest support for some additional variables designed to capture resource differentials, but the significance of population and area in our basic regression can be viewed as partly capturing such effects.

After controlling for other variables related to openness, we find a number of interesting results regarding the relationship between economic development and openness. In our basic openness equation, after controlling for trade policy differences, we found that there is no evidence that richer countries trade relatively more: indeed richer countries appear to trade relatively less. However, after including a variable for national price levels to control for differences in the valuation of GDP, we find this negative relationship largely disappears.

Returning to the question that sparked our interest, the openness equation suggests that Australia's level of external trade is close to the level that would be predicted. This stands in contrast with the results from the gravity model which tended to suggest that Australia actually trades much more than expected. The results from the openness equation suggest that the factors that best explain Australia's low level of trade are its remoteness and large size. The first of these can be viewed as a natural disadvantage (see Redding and Venables 2004), while the second can be viewed as an advantage – because of the natural diversity of our large land mass, Australia is able to produce many goods internally and does not need to trade for them externally. However, the implications of our low openness for broader economic outcomes are unclear. While some researchers (for example, Frankel and Romer 1999) argue that economic growth is positively correlated with openness, others (for example, Rodriguez and Rodrik 2001) have questioned this conclusion. In particular, there are good reasons to think that there is no straightforward link, and indeed that natural barriers to trade and policy barriers might have different effects on growth. Finally, it is worth noting that remoteness and economic location are not necessarily static but dependent upon the growth rates of other countries. Australia already has strong trading links with China and India, and as these economies continue to grow relative to other countries, Australia's relative geographic location is likely to be less of a barrier to trade.