One of the bigger areas of debate in emissions trading is how to allocate permits. The purists (eg, Garnaut) suggest they should all be auctioned while political realities suggest a large chunk will be given away to compensate the industries that will suffer most under the scheme (Garnaut suggested that you should give cash compensation so it's transparent).
An important thing to keep in mind is that this aspect of the debate doesn't change the total emissions - it's not an environmental debate; it's about fairness and economics. You get the same emissions whether you auction all the permits or give them all away. What changes is who pays and who gets the revenue. Under auctioning, the government gets the revenue and hopefully uses it to reduce other taxes. If you give permits away, the emitting industries get the money.
A new study has examined the economics of auctioning versus freely allocating permits under the US emissions trading scheme, with some possibly surprising results. They point very much to auctioning a vast majority of permits as a better way.
First, the unsurprising result: the more permits that you auction, if you use the revenue to reduce other taxes, the lower the economic cost of the trading scheme.
Second, the possibly surprising result: you only need to give away 15% of permits to completely compensate the industries most affected by the scheme. I think the proportion the government and opposition are debating freely allocating for the Australian scheme are substantially larger than that*, so think about what that means: these industries will actually profit from the introduction of the ETS. The study suggests that giving away 100% of permits leads to a doubling of profits for many industries.
*Our scheme and industries are slightly different so the numbers from the US won't directly translate, but they should be in the same ball park. Does anyone know if anyone's run the numbers here? Maybe in the Treasury modelling...