The economic cost of carbon policy

The chair of Alberta’s climate change advisory panel on what he wishes its final report had said about the economic impact of carbon pricing

Content image
Premier Rachel Notley, right, and Environment and Parks Minister Shannon Phillips after unveiling Alberta's climate strategy in Edmonton, Alberta, on Sunday, November 22, 2015. THE CANADIAN PRESS/Amber Bracken
Premier Rachel Notley, right, and Environment and Parks Minister Shannon Phillips after unveiling Alberta’s climate strategy in Edmonton, Alberta, on Sunday, November 22, 2015. THE CANADIAN PRESS/Amber Bracken

A year ago this week, I began work on Alberta Environment Minister Shannon Phillips’ Climate Leadership Panel. Our process ended on Sunday, November 22, 2015 when Premier Notley released our report and announced that the province would be following many of our recommendations.

Our process had been quite unique in that, over the course of five months or so preceding the announcement, we’d been in regular contact with government officials, both on the political side and in the public service, so our final report was really more of a report to Albertans than to the government.  Our recommendations had been presented to cabinet and to senior officials over the weeks and months prior to the announcement, as we sought to refine our recommendations based on input and questions from stakeholders and government officials.  We’d run a months-long process to bring together the best analysis we could from both inside and outside government, to analyze it, process it, and to bring forward the best ideas for Alberta. The report was intended to explain it all to Albertans—we hoped it would bring them inside our minds, inside the analysis and into the process we’d been through. While we knew that many would not agree with the recommendations, we hoped that they’d see the answers to their questions in our report. On the Saturday afternoon before the announcement, with all of the documents out of our hands, I headed off for a run and I felt for the first time a feeling I’ve felt almost every day since: a nagging feeling that we’d missed something, that we could have done something differently, could have explained it better, or could have provided more evidence.

Of all the times where I’ve second-guessed the final edits on our report, this past week was one of the worst. Waiting for the shoe to drop on Chris Varcoe’s article in the Calgary Herald on a leaked memo on carbon pricing (I spoke to him Wednesday morning), I’ve run over in my mind all the analysis we did and all the things that I wish we’d included in the report, some of which were never in it, and some which ended up on the proverbial cutting room floor. I feel very confident in the analysis we did and in the evidence supporting our policy recommendations, but over the past seven months since the announcement, I’ve seen a few places where we didn’t include enough to explain to Albertans what many economists take for granted about greenhouse gas policy in general and carbon prices in particular, and a more than a few places where we could have added more to the report, including more discussion with respect to the potential economic impacts and the tradeoffs involved in our proposed policies.

Hopefully I can set a few of those things straight here.

First, I think we need to start with the core of our recommendations—a price on carbon. Why do economists like carbon pricing? Economists from Mankiw to Krugman and from Mintz to Stanford generally like the concept of a price on carbon because it is known (and has been since Pigou in the 1920s) to reduce emissions at the lowest economic cost to society. Governments can do (and have done) a lot with regulations, but they can’t customize regulations to the particular circumstances of individuals and firms without incurring significant costs and inefficiencies, and even if they could, regulations are well-known to generate fewer incentives to innovate than an equivalent price. Simply put, the reason our policy recommendations included a broader and better carbon pricing system for Alberta was to minimize the economic cost of action. If you want to read more on this, start with this report (PDF) from Canada’s EcoFiscal Commission.

However, there are many issues with how carbon pricing is presented as a policy tool. Most importantly, people need to understand that carbon pricing in and of itself is not a magic bullet. Below are some of the reasons why.

First, stringency matters: there’s no reason to expect that a carbon price will do more than a stringent set of regulations—economics simply tells us that in most cases, a carbon price applied broadly could likely accomplish the same outcome at a lower total cost to the economy than would be possible through a suite of regulations.  As Simon Fraser’s Mark Jaccard has pointed out extensively of late, many of our most effective policies have been regulations, but this is because our governments have been willing to impose stringent policies in the form of regulations, not because regulations are inherently more or less effective. Some have argued that the plan we proposed for Alberta doesn’t accomplish enough while arguing in the same breath that it’s too stringent. Economics tells us that, if you want to accomplish more, you can either increase the price or impose more stringent and likely more costly regulations. It also tells us that either option will require individuals and firms to take more costly actions to reduce emissions. After all, while Friedman made the quote famous, economists have known since well before Adam Smith that there is no free lunch.

Second, interaction with other policies and constraints matters.  Since Lipsey and Lancaster in the 1950s, economists have also known that the presence of other market imperfections can mean that the textbook optimal solution to a problem (in this case, a pure carbon price) might no longer be optimal if it exacerbates other problems—this is known as theory of the second best. Here, Alberta’s electricity market provides a perfect example—our market places zero value on long-term reliability, even though we know that reliability is a valuable thing to electricity consumers. Advocates of textbook solutions can argue that we should price carbon and let the market sort it out rather than relying on complementary regulations. However, if that carbon price leads too many generating asset owners to implement early shut downs, and the market has no means to send a signal of the value of reliable supply, problems may ensue, the costs of which vastly exceed the gains of pricing versus regulation.

Third, there are distributive impacts of policies which a textbook analysis of the first best can often ignore. On their own, carbon prices can be regressive, although they need not be as Nic Rivers, Randy Wigle, and others have shown. Of course, regulations have regressive redistributive impacts as well—they impose costs on firms which will be passed through supply chains and affect consumers. The potential regressive nature of emissions reduction policies is why first B.C. and now Alberta have opted for rebates to shield lower-income consumers from the reduced purchasing power which would otherwise disproportionately affect them under a carbon price. That consumers will still respond to higher relative prices for carbon-intensive goods (substitution effects) while not being made poorer overall (income effects) is also taken as given by most economists, but evidently not something we did a sufficient job explaining in our report. More on this is a subject for another day.

Fourth, and related to each of the points above, political constraints matter.  Jesse Jenkins has done some great work on this topic, and as economists we ignore these issues at our peril.  Just as we’ve studied coalition formation for international climate change agreements, policy formation at national and subnational levels must be one of consensus building to some degree.  The range of stakeholders who have supported the policy proposed by the government on November 22nd is a product just that type of process—a process which involves some give and take, and will not necessarily lead to the lowest cost theoretical policy, but can deliver the lowest-cost, achievable policy. But, we’ve also seen many examples of policy proposals for which proponents were not able to build a strong, political consensus. Those proposals remain proposals.

Next, at the risk of validating a straw man, we know that pricing carbon in Alberta isn’t going to solve climate change. We know that climate change is a global collective action problem—when we burn fossil fuels, we impose costs (quantified as the social costs of carbon) on others—a burden for which we do not have to pay compensation. This is what economists know as an externality, and it leads to a less efficient overall outcome than you’d have with perfect markets. The Alberta economy, with some of the highest emissions per capita in the world, has benefited significantly from the fact that current and future generations of global citizens can’t easily send us a bill (or sue us costlessly for damages, if you prefer a Coasian view of the world).  Economic activity in Alberta emits, on average, 62 tonnes of greenhouse gas emissions per person per year, which implies a cost imposed on others of $,2800 to $4,500 annually per Albertan using the U.S. government’s central estimates of the social costs of carbon emissions, converted to Canadian dollars.  And, before you point out that action in Alberta won’t matter if China and India don’t act as well, allow me to point out that the social costs of carbon emissions today increase if you assume inaction on the part of other nations. In other words, if you’re going to argue that no one else is going to act on this problem, you’re arguing that the damages we impose on others through our actions are higher than the numbers I’ve listed above.

Finally, and most related to this week’s news, while correcting an externality leads to more globally efficient economic choices, it doesn’t mean that those who benefit from being able to pass these costs on to others will be made better off by having to pay for these costs. If government policy is going to reduce the ability of economic actors to impose costs on others, there will be impacts on those who would otherwise benefit from doing so. Those impacts will be larger in provinces with more emissions-intensive economies—a list on which Alberta sits near the top. So while we as economists know that carbon pricing is, under most circumstances, the policy with the least economic costs of emissions reductions, that doesn’t mean it comes without potential economic costs. However, economic costs must be judged compared to something. If you compare real-world policies against unrealistic alternatives, they are always going to perform poorly.

Which brings me to where we are this week, and to Chris Varcoe’s write-up in the Calgary Herald on the basis of a leaked memo based on preliminary analysis of the climate leadership plan. The economic impact numbers—projections of 1 to 1.5 per cent lower GDP by 2022, a drop in oil exports, and employment 15,000 below what would otherwise occur—and the fact that the numbers weren’t released sooner have ruffled feathers on all sides of the spectrum. With all of the above in mind, I’ve given a lot of thought over the last few days to what we might have included in our report to better address the economic costs of climate change policy in Alberta. Hindsight is, of course, 20:20, but we could have written something like this:

Alberta’s economy runs on resources and the production, processing, and combustion of those resources generates greenhouse gas emissions. Our emissions do not simply come from our large industries—almost everything else we do has greenhouse gas emissions impacts, whether it’s driving our cars and trucks, heating our homes, or purchasing goods delivered here by plane, truck or train. Reducing emissions in Alberta will not be easy—we don’t have a magic wand, and if cost-effective, lower-emissions substitutes were available in all cases today, we wouldn’t be facing this problem. But, not reducing emissions in Alberta is also potentially very costly. We’ve already seen policies and actions aimed at our resource sector whether through the rejection of pipelines, the application of low carbon fuel standards, or challenges to companies investing here from their shareholders or from sustainable investors. If Alberta chooses not to act, those costs won’t go away. And, we’re part of a federation, and our federation has committed to an ambitious target to reduce national greenhouse gas emissions. There will be costs if Alberta is not a constructive partner in those efforts—continued market access challenges or unfavourable policy design. Those costs may be difficult to quantify, but that doesn’t mean they’re not real. Finally, of course, we know that emissions impose costs on others around the world—if we use the most recent estimates, we are each imposing on average $2,800 to $4,500 worth of costs on current and future global citizens every year with our emissions, and many of these impacts affect some of the poorest countries on earth.

The policy package we are proposing to reduce emissions has the potential to impose some costs on the Alberta economy—there will likely be a real and measurable impact on GDP relative to a business as usual outcome, if we assume that Alberta would face no material economic costs from inaction on climate change. Based on internal and external economic modeling results, this impact will likely be on the order of 0.25 to 0.5 per cent cumulatively by 2022. In other words, the impact of the actions we propose could amount to approximately one day’s lost economic output in 2022.  The impacts will depend on the decisions the government makes with respect to the use of revenues beyond those which we’ve suggested be used for renewable energy and energy efficiency programs—the more cost-effectively these revenues are used, the more modest the economic impacts will be.

There are some impacts we were not able to quantify. While we were able access both private-sector and government economic models, we were not able to complete analysis of the impacts of reduced air pollution from our proposed policies on labour productivity and health care expenditures. Based on the latest economic evidence, we expect these impacts to be significant and positive and thus they would offset some of the negative impacts on GDP. We also have, as mentioned earlier in the report, limited quantitative evidence on the effects on emissions-intensive small and medium sized business.  Unlike large emitters, these SMEs do not report their emissions annually, and so our ability to estimate effects is much more limited. We have, through the application of output-based allocations in large industrial sectors, sought to reduce the competitiveness implications for our emissions-intensive and trade-exposed industries far beyond what would be the case with a BC-style revenue neutral carbon tax. However, these measures will not be perfect and some industries and firms—those with high emissions and low value-added per unit emissions in particular—will suffer disproportionate effects, as will some individuals with particularly high emissions lifestyles. Based on modelling presented to us, and given the emissions-intensive nature of Alberta’s economy, there will likely be reductions in output, exports, and potentially some impact on total employment, relative to a classic business-as-usual case.

However, it is our view that the business-as-usual case that a modeller might use to assess the costs of these actions does not exist for Alberta. We see today increasing pressure on firms to mitigate carbon risk, increasing pressure on governments to achieve their Paris commitments, and increasing focus on Alberta as a symbol of inaction on climate change. We do not see how a comparison to a case where Alberta can continue to find viable markets for its products and see investment return to the oil sands in the absence of credible action on greenhouse gases exists. What would likely exist as an alternative is a world where Alberta faces increasingly discriminatory and punitive policies and barriers to trade both from within and from outside Canada, and where firms face mounting shareholder and institutional investor pressure not to invest in Alberta. The costs of these are speculative, and more difficult to quantify, but we are confident that they far and away exceed the cumulative costs of the actions we’ve recommend. Simply put, if you assume there is no value in reducing emissions, you’ll find that there’s no value in reducing emissions. We believe that our policy recommendations will be of net benefit to Alberta, yes in terms of the avoided costs of greenhouse gas emissions and air pollution, but also in terms of the avoided costs of discriminatory and punitive policies imposed upon Alberta,

Our policy package relies on economic evidence tailored to the specific, current circumstances of Alberta’s economy and proposes what we believe is a package which allows Alberta to go from defence to offence with respect to climate change policies, and do this in the economically smartest manner given the nature of the Alberta economy. It is our hope that this will allow Alberta to prevent discriminatory policies applied against our resources, to compete effectively for global market share, and to be a constructive partner as our federation moves to address climate change and to reduce the costs which our actions impose on others. Of course, we also expect that it will reduce Alberta’s emissions in a way consistent with the goals of the government and with the lowest aggregate cost to our economy.

Of course, writing reports is much easier in hindsight.