Mathias Cormann (WA, Liberal Party, Shadow Assistant Treasurer) Share this | Link to this | Hansard source

asked the Minister representing the Treasurer, upon notice, on 18 August 2011:

With reference to the Treasury Carbon Tax modelling, ‘Strong Growth, Low Pollution: Modelling a carbon price’:

(1) Is it correct that Treasury did not use Warwick McKibbin’s G-Cubed model on this occasion for any of its modelling; if so, why not, given the G-Cubed model is a highly regarded global model which has been widely used for precisely these sorts of exercises, and it was used for the Carbon Pollution Reduction Scheme modelling in 2008.

(2) Is it correct that, for the 2008 modelling (Australia’s Low Pollution Future), the G-Cubed results for the 5 per cent emissions reduction scenario showed a cost to the level of Gross Domestic Product in 2030 that was more than twice as large as in the Global Trade and Environment Model (GTEM).

(3) Is it correct that, for the 2008 modelling (Australia’s Low Pollution Future), the G-Cubed results for the 5 per cent emissions reduction scenario showed only just over half as much abatement being achieved globally via internationally-linked Emission Trading Schemes (ETS), in 2050, as the GTEM model.

(4) Is it good modelling practice in general – especially for complex exercises like Carbon Tax Modelling that are potentially so dependent on assumptions about factors such as technological progress – to run several different models so as to allow comparison of the results (as was done in 2008).

Penny Wong (SA, Australian Labor Party, Minister for Finance and Deregulation) Share this | Link to this | Hansard source

The Treasurer has provided the following answer to the honourable senator's question:

(1) Treasury used a broad suite of models that included the global computable general equilibrium (CGE) model GTEM, the domestic CGE model MMRF, the set of PRISMOD models for household price impacts, and detailed sector specific modelling of transport, electricity generation and land sector abatement by experts in those fields. The modelling exercise undertaken for the Australia’s Low Pollution Future (ALPF) report was predominantly focussed around providing information on the targets and trajectories for Australian emission reductions. In contrast, the modelling exercise for the Strong growth, low pollution (SGLP) report was predominantly focussed on domestic policy mechanisms. As such, more resources were devoted to the domestic implications of carbon pricing rather than the international dimension. This required a reordering of priorities given the time and budget constraints faced by the modelling exercise and resulted in the use of two sector specific electricity generation sector models rather than two international CGE models.

(2) Table 5.12 and 6.4 in the ALPF report presents a range of estimates for Gross Domestic Product in 2020 and 2050. Page 111 of the ALPF report outlines some of the reasons for the divergence of results between the G-Cubed and GTEM models.

(3) In the ALPF modelling exercise both GTEM and G-Cubed were required to meet the same environmental targets through time. It is not accurate to suggest that global emissions were significantly differently in the two models.

(4) See question 1.