Deconstructing Jaccard and the Green Party

Last week, the Green Party issued a press release claiming that a “secret government study backs $50 carbon tax”, which is convenient since the Green Party recently endorsed a $50 carbon tax. My initial response to the Green’s carbon tax was one of skepticism, mostly in regards to the likely non-impact on driving, and the flawed emphasis on tax shifting.

I have now had the chance to look more clearly at the report, done for Natural Resources Canada, by SFU’s Mark Jaccard and his team of associates. The modeling involves estimating a reference, or business-as-usual, case then looking at the impact of a variety of carbon taxes from $10 to $250 per tonne of CO2 equivalent, and for years 2010, 2015, 2020 and 2030. The impact is specified as relative to the business-as-usual case.

The spin from the Greens was that “a $50/tonne tax on carbon would lead to reduced greenhouse gas emissions and have an insignificant impact on the Canadian economy.” This is not exactly true; my initial suspicions were correct about the efficacy of a $50 per tonne carbon tax. To get there, we have to read through the lines a bit in the report, but the bottom line is that there is little economic impact because there is little impact on CO2 emissions.

Under the business-as-usual case, Canada’s CO2 emissions continue to climb, from 624 megatonnes of CO2 equivalent (or MT CO2e) in 2005 to 659 in 2010, 744 in 2020 and 838 in 2030 (p. 15). Against this baseline we can evaluate the modeling outcomes for the $50 per tonne carbon tax (p. 47). The simulations find a reduction relative to business-as-usual of 36 MT CO2e in 2010, rising to 114 in 2030. Put into percentages, the carbon tax does not actually reduce aggregate CO2 emissions at all: they fall by 5.5% relative to the baseline in 2010, but the total flow of emissions (659 – 36 = 623) remains about the same as 2005 levels (624); and, while emissions fall by 13.6% relative to baseline by 2030, aggregate CO2 emissions actually rise by 2030 to 724 MT CO2e (838 – 114).

This is especially interesting given that the Tory plan for climate change got smacked by Jaccard and Nic Rivers in a recent publication for the CD Howe Institute. In this publication, published only five months after the carbon tax analysis, they conclude:

We estimate that these policies are likely to reduce emissions substantially compared to their business-as-usual evolution. By 2020, emissions would be 120 megatonnes below projected levels and by 2050 the reduction would be almost 400 megatonnes below the business-as-usual projection. However, the results also indicate that overall emissions in Canada are unlikely to fall below current levels. The government is likely to miss its 2020 emissions target by almost 200 megatonnes.

To be fair to the Greens, the other study was not specifically about the impacts of their plan, only on carbon tax models. And the Greens climate change plan has more elements to it, even if the carbon tax is the centrepiece. But the spin that the carbon tax would have negligible economic impact is only technically true because it does not actually reduce the annual flow of CO2 emissions.

What then would? The trouble with the report is that it assumes a certain amount of carbon tax would be implemented immediately (in fact, 2006) and stay unchanged over time. Based on the model, we do not even get meaningful reductions in CO2 emissions unless we are well over $100 per tonne. The maximum scenario of a $250 per tonne carbon tax leads us to a 43% drop in emissions relative to baseline by 2030, to a level of 478 MT CO2e, which is in turn 23% below 2005 CO2 emission levels.

It would, of course, be foolish to implement such a large tax all at once. More interesting and more politically palatable would be to introduce a lower level of the tax with a schedule to increase it over the course of two decades (perhaps starting with $50 but then scaling up to $500 over two decades; see this recent post). These numbers also suggest that a carbon tax alone may be insufficient to the task at hand; we will need a variety of other measures to get to reasonable targets.

Also in the report there are some sectoral and regional breakdowns. On the regional front (Alberta, you will want to sit down right now), any way you price it the impact of a carbon tax is going to have a disproportionate impact on Alberta, which I suppose should be obvious. Alberta bears 39% of the total regional adjustment with a $10 carbon tax, and just over half with a $100 carbon tax.

On the sectoral front, the impact on transportation is not huge for a $50 carbon tax. In 2010, CO2 emissions are reduced by 5 MT CO2e relative to baseline of 194, for a 2.6% reduction, but that is still higher in aggregate flows than what came out of tailpipes in 2005. By 2030, the reduction is 13 off a baseline of 253, a drop of 5%, but a rather large 31% increase in total emissions relative to 2005.

As was speculated, the impacts are much better for electricity generation. Emissions drop by 8 MT CO2e in 2010 from a baseline of 105, and a level of 111 in 2005 for an actual reduction in total annual emissions. And by 2030, the reduction is 24 off a baseline of 103 MT CO2e, which means an actual reduction relative to 2005 levels of 29%.

A final thought is that if revenues from the carbon tax were used to finance infrastructure, industrial policy, basic R&D, etc there would be a feedback that would improve both the economic and the emissions reduction impacts, and could offset the regressivity of the tax. This is another reason why the “tax shifting” promoted by the Greens is more electioneering than sensible public policy.