Richard Tol's latest discussion paper looks at carbon taxes and wonders just how much you get by way of emissions reduction if you stipulate that you are not going to increase the overall tax take. In some countries the answer is not very much - if your current tax take is low then you can only impose a small carbon tax. If you try to impose a higher tax in these countries, they will (theoretically at least) drop out and their emissions will remain "out of reach".
Different levels of carbon tax are discussed -
a stabilization target of 650 ppm CO2 equivalent would require a carbon tax of $6/tCO2e, in all countries, on all emissions, of all gases. A target of 550 ppm CO2e would require a tax of $29/tCO2e, and a 450 ppm target would need a $143/tCO2e tax. These are the three price levels shown in Figure 3
When you plug these tax levels into the equation carbon tax take = current tax take, you get the following answer for the 450ppm target, which Tol tells us is equivalent to the EU's 2°C target:
For $143/tCO2e, the carbon tax revenue is greater than 100% of tax revenue for more than 10% of emissions; and greater than 10% for all countries. Such a carbon tax would not be fiscally prudent in many economies. $143/tCO2e is the tax needed to meet the 450 ppm CO2e target, which roughly corresponds with the 2°C target of the EU and UN (den Elzen et al. 2007).
As Pielke Jr notes, this appears to mean any proposed carbon taxes must be low.