RDP 2006-01: Modelling Manufactured Exports: Evidence from Australian States 2. Previous Research
April 2006
- Download the Paper 484KB
Recent published attempts to model Australian exports have tended to form part of multi-country studies, and have generally focused on modelling either goods and services exports or merchandise exports as a whole, rather than its components. The results of these studies have been quite diverse. With regards to the price elasticity of Australian exports, Wu's (2005) model of merchandise exports finds the smallest (and only insignificant) elasticity of −0.3. In contrast, Caporale and Chui's (1999) estimate of the price elasticity of goods and services exports is around −0.8, and Senhadji and Montenegro (1999) find an implausibly large elasticity of −2.2. Similarly, income elasticities also vary; ranging from 0.8 (Senhadji and Montenegro) to 1.3 (Caporale and Chui). This variation comes despite these studies all being estimated over similar samples (starting in 1960 and ending around the mid 1990s), using similar price variables (export unit values) and similar estimation techniques. This variation in results highlights the difficulty in finding robust estimates of aggregate export equations.
These results using aggregate exports are also likely to hide significant variation in the elasticities of export components, and are therefore not directly comparable to this study of manufactured exports. In particular, it is possible that the price elasticity of aggregate exports is somewhat lower than that for manufactured products, given the apparent insensitivity of the supply of resource and rural exports to changes in prices. Despite this, the only known study that separately models manufactured exports is that by Dvornak, Kohler and Menzies (2005), who find a price elasticity for manufactured exports of −0.8.
Finally, the only study that looks at manufactured exports at a state level is Neri and Jayanthakumaran (2005). Using annual data over the period 1989/90 to 2000/01 and a descriptive approach, they find that there is considerable diversity in the performance of manufactured exports across states which cannot be explained by differences in industrial composition.