Monday, 30 April 2018

Canadian Carbon Pricing Confusions

The federal government (Environment and Climate Change Canada - ECCC) released on April 30, 2018 its estimate of the incremental effect of its carbon pricing initiative relative to other policies in achieving Canada’s GHG reductions – Estimated Results of the Federal Carbon Pollution Pricing System. They estimate annual reductions in 2022 of 80-90 mega-tonnes (MT) of CO2. Their numbers are dramatically higher than estimates by my research team. Why?

The approach we took to estimate the incremental effect of federal emissions pricing

At about the time the Trudeau government announced that emissions pricing would be “central” to achieving its GHG reduction targets, it also announced specific regulatory policies, including nation-wide adoption of Alberta’s coal plant phase-out and methane regulations, tightening of vehicle regulations, and (a bit later) a clean fuel standard, which is similar to BC’s low carbon fuel standard but applied to all sectors not just transport. To assess the incremental effect of the federal carbon pricing policy, we created a reference scenario which included everything that should happen absent the federal pricing initiative. 

Thus, our reference scenario included all carbon pricing and regulatory policies of the provinces and simulated their effect on emissions between 2018 and 2030. BC and Alberta had both committed to $30 /tCO2 for carbon pricing while Ontario and Quebec had committed to a price that climbs toward $20 and surpasses that threshold well before 2030. All four provinces had existing and announced regulations, such as Alberta’s announced cap on oil sands emissions, methane regulations, and coal-plant phase-out by 2030, and BC’s clean electricity standard and low carbon fuel standard. We included regulations by other provinces too, such as the electricity decarbonization policy in Nova Scotia. To these we added the existing and announced federal regulations. 

We sustained all of these provincial and federal policies through to 2030, which gave us a forecast of the evolution of Canadian emissions if Trudeau’s government had avoided carbon pricing as a policy and instead relied on existing provincial policies (pricing and regulation) and its own announced regulations. This reference scenario (without the federal pricing initiative) sees Canadian emissions fall approximately 6% from their 2005 level.

To this reference scenario, developed in early 2016, we later added the federal carbon pricing policy, which reaches $50 /tCO2 by 2022. We assumed, in the absence of a schedule beyond 2022, that the carbon price would remain constant after that. Not surprisingly, the incremental effect of the federal carbon pricing policy is very small by 2030 and even smaller by 2022. I’m getting my research associate to dig out the exact numbers for these two dates. (She now does modeling for the International Energy Agency in Paris and is awfully busy!). But my eyeball guess looking at our graphs is that the incremental effect of the federal carbon pricing policy is to reduce emissions 1-2% from their 2005 level, a reduction of 10-15 MT in 2030. That number should be smaller in 2022, far below the claim of 80-90 MT in the latest ECCC report.

The federal approach to estimate 80-90 MT of reductions from pricing by 2022.

In reading the report, I see two possible causes for the discrepancy between our estimate of a low incremental contribution from the federal carbon pricing initiative and the high estimate in this latest federal report. The first cause is that the federal government estimate takes credit for all pre-existing carbon pricing initiatives of the provinces. The second cause, more speculative on my part, is that they may not have done proper incremental policy modeling, meaning that the federal carbon pricing got recognition for emission reductions that should be attributed to non-pricing policies.

Page 3 of the report helps explain the first cause of the discrepancy. The estimated 80-90 MT attributed to federal carbon pricing is based on the assumption that there never have been provincial emissions pricing policies in Alberta, BC, Ontario and Quebec. The carbon pricing policy is “compared to a hypothetical scenario in which they [provincial governments] did not have pricing systems in place.” (p.3) In other words, the reference scenario for estimating the federal carbon pricing initiative is a hypothetical world in which there is no carbon pricing anywhere in Canada. Which of course is not true.

This is clearly not an accurate way to represent the incremental effect of the carbon pricing initiative of the federal government. While it makes a lot of sense to have better federal coordination and consistency of climate policies across the country, and the federal backstop carbon price does that, it is nonetheless grossly misleading to suggest that current provincial pricing can be attributable to federal policy any more than that the phase-out of coal plants in Ontario in 2004-2014, the policy-driven cancellation of coal plants in BC in 2007, and Alberta’s announced phase-out of coal plants in 2015, can be attributed to the federal coal plant policy announced in 2016.

While this is likely to be the dominant cause of the discrepancy in our estimates, I also could not find an explanation in the report of the method the federal policy modelers used to estimate the incremental effect of the federal carbon pricing initiative and federal regulations. As I noted above, this entails first simulating all of the provincial and federal non-carbon pricing policies and the provincial pricing policies to 2022 and 2030. And then to run a second simulation with only the addition of the federal carbon pricing initiative. The change in emissions between these two simulations indicates the incremental contribution of that policy. This contribution of the federal carbon pricing policy will be very small, as we found with our modeling research. But you actually don’t need a model to see what is obvious by surveying the portfolio of provincial and federal regulatory policies, past and present.


In 2016 I and two research associates produced a report entitled Is Win-Win Possible? Can Canada’s Government Achieve Its Paris Commitment . . . and Get Re-Elected? In which we explained that a rapidly rising carbon emissions price was needed to achieve the Paris commitment. We noted that while all climate policies are politically difficult, there is considerable evidence from real-world GHG policy experience and political science surveys to suggest that carbon pricing is far more politically challenging than some regulatory policies. We also noted that flexible regulations can be designed to approach carbon pricing in economic efficiency, if designed with that purpose in mind.

Some economists, including some at Canada’s carbon pricing advocacy entity, the Ecofiscal Commission, dismissed our assessment. They presented studies constructed to show a deliberately big economic efficiency gap between regulations and carbon pricing, instead of testing the likely long-run cost of using flexible regulations like the low carbon fuel standard over several decades to decarbonize transport. And they dismissed as na├»ve any research showing the visceral antagonism to pricing policies by significant segments of the population – and therefore the risks to politicians of relying on such policies. 

None seemed willing to even discuss the importance of comparing GHG policies using a criterion such as political cost per tonne reduced in order to compare this to economic cost per tonne reduced. This is unfortunate, because research by ourselves and others shows that carbon pricing has an enormous political cost per tonne in comparison to flexible regulations. This helps explain why many of these flexible regulations have played a much bigger role in GHG emission reductions thus far in Canada, California and Europe, including Scandinavia where there has been some form of carbon pricing for years.

A decade ago, Canada had a federal election dominated by the issue of carbon pricing. Voter rejection of carbon pricing enabled Stephen Harper to defeat Stephan Dion and win power for a decade, a decade in which he deliberately stalled on implementing effective GHG reducing policies. That was not an economically efficient outcome.

Because carbon pricing advocates have convinced the Trudeau government to take a large political risk for only a small incremental GHG reduction, history may soon repeat. Studies that are distorted to show an artificially large reduction from the federal pricing initiative are not going to save the day. Trudeau may win re-election and sustain the federal carbon pricing. But if so, this will occur in spite of carbon pricing, not because of it. One must ask if the risk is worth it, especially when the impacts of coal-plant phase out, methane regulations, and a clean fuel standard (that fairly efficiently decarbonizes transport) dominate our GHG reductions and yet are much less difficult politically – as polling continuously shows.

Ironically, our incremental modeling of various GHG reduction policies in Alberta shows a similar outcome. We estimate the incremental effect of Alberta’s carbon pricing policy (at its stringency level of $30 and different application in various economic sectors) is less than 5% of the GHG reductions caused by the other regulatory policies (excluding subsidy policies) in its Climate Leadership Plan. The vast majority of reductions are caused, again, by the coal plant phase-out, the methane regulation, the oil sands emissions cap (which varies depending on forecasts of future oil sands output), and various efficiency regulations. Yet some polls suggest that while the Notley government’s popularity is little affected by its introduction of regulations (most Albertans support coal plant phase-out), it has been greatly affected by the strong and ongoing opposition to her carbon tax.