RDP 2006-01: Modelling Manufactured Exports: Evidence from Australian States 7. Conclusion

This paper examines the determinants of manufactured exports through the use of a panel of five Australian states, taking advantage of the cross-state variation in manufactured exports, real exchange rates and trading partner GDP. This approach can potentially provide more robust estimates of the determinants of manufactured exports than direct estimation of a national model.

The results indicate that this estimation approach provides reasonably robust estimates of the price elasticity of manufactured exports, using both a mean-group and a fixed-effects panel and various specifications of export demand. In contrast, income elasticity estimates are sensitive to the inclusion of other trending variables. These results are then compared with the robustness of direct estimates. It is found that direct estimates of the national price elasticity are similarly robust, and direct estimates of the income elasticity are similarly unstable. These results indicate that the direct approach to modelling manufactured exports is appropriate, despite the instability of parameter estimates. Section 6 of the paper extends the analysis to consider the role that domestic conditions play in determining exports. The results are consistent with the theoretical considerations discussed in the paper, which posit an inverse relationship between the strength of domestic demand and manufactured exports, controlling for trading partner GDP.