Ken Boessenkool on Getting a Triple Dividend from a Carbon Tax

Using the carbon tax as the primary mechanism to reduce GHG emissions is an opportunity to eliminate the economic and regulatory distortions caused by other green programs.

Advocates of using a carbon tax to reduce GHG emissions sometimes refer to a double dividend from placing a price on carbon. The first dividend comes from appropriately pricing externalities through a carbon tax. The second dividend comes from using the revenues raised by that carbon tax to reduce other, more economically distorting taxes.

But there is a third dividend available when a carbon tax is implemented that has received much less attention. This third dividend is substantial and much easier to see than either of the first two dividends.

This third dividend is the simple result of using the carbon tax as the primary mechanism to reduce our GHG emissions, meaning that many existing green programs—along with the spending and regulatory distortion they cause—can be eliminated.

The First Dividend

Ken McKenzie from the University of Calgary points out that carbon emissions generate at least two different externalities—costs that are not appropriately included in the underlying production costs associated with those emissions, and thus are “un-priced” in the market place.

The first un-priced externality is the costs of climate change. This results in an overuse of these fuels, producing excess carbon and climate change. This problem can be curbed by putting a tax on these fuels to bring their market price more in line with the cost of those externalities. The level of taxation should be equal to the social costs of climate change. A tax up to that level will eventually reduce emissions to the socially efficient level.

There is little agreement on what the social cost of carbon is when it comes to climate change. The US Government Accountability Office estimates the domestic social cost of carbon for 2020 at $7 per tonne which compares to a previous global cost of carbon of $50 per tonne. Admittedly the former estimate was under an administration much more skeptical about the need to address climate change. Canada’s Ecofiscal Commission lays out the Government of Canada approach to calculating the social cost of carbon which had the “central” measure at $45.10 per tonne for 2020. These estimates also have large “tail risks,” going as high as $179. These are current estimates at a point in time, with a central estimate of $80 in 2050. But the social cost of carbon will also rise over time as the costs of climate change increase.

The second un-priced externalities associated with the use of fossil fuels and motor vehicles are unrelated to climate change. These include things such as particulate pollution and road congestion. Recent research by the IMF suggests that a nationally efficient carbon price for Canada, leaving aside externalities due to climate change, is $30 per tonne.

The central lesson here is that there are economic gains to be had (because we are correcting for an externality) from a carbon tax up to $75 per tonne in 2020. This is more than double any existing carbon tax in Canada. And these benefits will rise as the carbon tax rises, in line with the rising social cost of carbon.

The Second Dividend

The second dividend comes from using the revenues generated by a carbon tax to reduce other, more economically harmful taxes. That was the subject of my first paper in this MAX Policy series but I’ll come at it here from a more technical perspective.

Any increase in tax rates can cause changes in the allocation of resources in an economy resulting in efficiency—or “deadweight”— losses. According to Ergete Ferede and Bev Dalhby, the marginal cost of public funds (MCF) is “a measure of the loss incurred by society in raising an additional dollar of tax revenue.” MCFs can be estimated by examining the responsiveness, or elasticity, of a tax base—consumption, personal income or corporate income—to changes in tax rates. The greater the elasticity of the behavioural response, the higher the MCF of a tax.

It is possible for a tax to produce little or no change in behaviour when introduced, in which case we would say that the MCF is 1.0. This would be an efficient tax, with no deadweight loss. Other taxes might induce changes in behaviour—changes to how much one works in the case of personal income taxes or how much investment is made in the case of corporate income taxes—in which case the MCF rises above 1.0 and a tax is less efficient. Another way to say this is that these changes in taxes produce deadweight losses.

Ferede and Dalby point to three things that affect these elasticity estimates. First, inter-jurisdictional competition means that smaller jurisdictions should have more elastic tax bases. Second, variations in socio-cultural makeup of a jurisdiction, such as language and culture, should affect tax base elasticity. Finally, the industrial mix of a jurisdiction should affect tax base elasticity because some industries or employment types may be more mobile than others.

Ken McKenzie describes it this way (emphasis in original), “taxes result in costs over and above the tax revenue raised because they distort economic behavior and prevent buyers and sellers from realizing some of the gains from trade.” One way to measure this is the MCF, which measures the loss to the economy, over and above the revenue raised, of raising $1 in revenue from a particular tax.

When a jurisdiction uses multiple revenue sources, there can be individual effects of each tax on the economy as well as interaction effects between two types of taxes. For example, increasing the personal income tax can reduce the returns to labour causing corporations to raise their compensation levels and thereby reduce corporate profits and the size of the corporate tax base.

Ferede and Dahlby’s estimates of MCF for personal, corporate, and consumption taxes in Canadian provinces take into account these interaction effects. As a result of their calculations, they conclude there would have been welfare gains in 2013 from a reduction in the CIT tax rate with a revenue-neutral switch to higher sales tax rates in British Columbia, Alberta, Manitoba, Ontario, and Quebec.

Carbon taxes did not exist for the period Ferede and Dahlby use to estimate MCFs so their work does not demonstrate whether the introduction of a carbon tax and a concurrent reduction in high-MCF tax rates would result in welfare gains. Indeed McKenzie is skeptical of such welfare gains as he argues that interaction effects might well swamp the efficiency benefits. He argues (emphasis in original) that the

“rise of the prices of goods and services due to the carbon tax results in a reduction of real wages, as the amount of goods and services that a consumer can purchase declines as a result of higher prices. This reduction in real wages has a similar impact as an increase in the tax on labour income, which lowers after-tax wages directly, and therefore increases the efficiency costs associated with existing taxes on labour income.”

I question this conclusion.

First, Ferede and Dalbhy (emphasis added), “…estimate the long-run responsiveness of Canadian provincial tax bases to tax rate changes by explicitly controlling for the possible interactions between tax rates.” Ferede's and Dahlby’s results show that for every jurisdiction, the MCF for sales taxes is lower than that for personal and corporate income taxes. If anything, general sales taxes operate more like the real wages McKenzie cites than carbon taxes do since they tax all consumption.

If general sales taxes have lower MCFs than corporate or personal income taxes when these tax interaction effects are controlled for, it seems reasonable to expect that carbon taxes might also have lower MCFs than corporate or personal income taxes. This is not an argument that the MCF of the first dollar of carbon taxation is 1.0 and that there are no interaction effects. Instead, the argument is that even when these are taken into account, given Ferede and Dahlby’s calculations, it is reasonable to expect the carbon tax MCF to be lower than the MCFs for corporate and personal income taxes.

In fact, based on the first dividend argument above, a case could be made that carbon taxes have an MCF below 1.0—the introduction of a carbon tax increases economic efficiency because the costs imposed by emissions are appropriately priced. This boosts the argument for the existence of a second dividend.

The Third Dividend

There is understandably some debate about the second dividend, as it relies on the interactions between complex taxes within a complex overall tax system. But there should be no such dispute about the third dividend.

Redundant Policies

Canadian governments have a whole host of regulatory, spending, and tax measures in place to address climate change that, in the presence of a rising carbon price, may be redundant and unnecessary. These include renewable fuel mandates, electric vehicle subsidies, converting commercial fleets to natural gas, increasing the use of low-carbon buses and trucks, industrial demand-side management programs, increasing wood for construction, and many more. For an exhausting list of 71 policies of this nature see Canada’s Ecofiscal Commission, Appendix A.

Having said that, it is also true that a carbon price does not have 100 percent coverage of greenhouse gas producing processes, so policies such as separate regulations for methane, for example, are necessary.

A good way to think about the relative efficiency of these other policies in the presence of a carbon tax is to measure their “implicit” carbon price—that is, the level of the carbon price which, if applied, would be required to achieve the same behavioural response as the non-price regulation.

Canada’s Ecofiscal Commission did a deep dive evaluation of Quebec’s electric vehicle subsidies. It found that the implicit carbon price inherent was an astounding $395/tonne. What is the justification for using policies which involve costs so far in excess of the existing or proposed carbon price? This is a highly costly and inefficient way to reduce carbon emissions.

The Ecofiscal Commission also examined a policy many governments put in place over the last 20 years—the phase-out of coal-fired electricity generation. It found that the implicit carbon price of the Alberta version of this policy was somewhere between $42/tonne and $99/tonne. In a world with a $75/tonne carbon price the former would happen without a separate policy initiative, and the latter would be an inefficient way to reduce carbon versus all of the things that would happen under a $75/tonne carbon price.

There is an argument that some of these policies might be more politically palatable than a carbon tax. Coal phase-outs have become standard across most developed countries, carbon taxes less so. In this sense their introduction may complement a carbon price and spur action. But this is an argument in favour of these policies from a political perspective—one I am open to. However, the political perspective does not undermine the economic argument that these policies are in many cases either unnecessary (if the carbon tax is higher than the implicit carbon price of the policy) or inefficient (if the carbon tax is lower than the implicit carbon price of the policy).

Some governments are replacing carbon taxes with a broad collection of regulatory and spending programs. For example, the current government in Ontario eliminated the previous government’s carbon-pricing mechanism and put in place a series of new command-and-control policies—a Made-in-Ontario Environmental Plan—intended to help them reach their emissions targets, including:

  • Low carbon vehicles uptake
  • Industry performance standards
  • Clean fuels (ethanol gasoline, renewable natural gas)
  • Ontario Carbon Trust
  • Other policies (organic waste, transit)
  • Policies to drive innovation

Many of these policies impose an efficiency cost on the economy but for the most part generate little to no revenue by which to reduce inefficient taxes. Ross McKitrick is quite adamant on this point and worth quoting at length:

Standard assertions about the efficiency of carbon taxes assume the absence of pre-existing command-and-control CO2 emission regulations. This was a reasonable assumption for Canada 20 years ago but it is not today. In the presence of inefficient pre-existing regulations, a carbon tax is not guaranteed to improve efficiency and may in fact worsen the distortions of the regulatory system… A policy implementation scheme that leaves distorting regulations in place and simply adds a tax or tradable permits system on top has no claim to being economically rational or efficient.

Moving beyond economic efficiency, there is also the raw size of the public service required to administer a command-and-control regulatory and granting system versus the size required to administer even an economy-wide carbon tax. Canada has an incredibly efficient tax system, even though provinces have control over much of their own tax policy, because most of the Canadian tax system is administered centrally and provinces “subcontract” this to the federal government. Having a single carbon tax collection agency run by the federal government would almost certainly be less costly than having ten provincial (plus three territorial) bureaucracies designing regulatory and spending schemes to reduce emissions.

It is striking, though by no means conclusive, that the British Columbia government—the Canadian government with the longest-standing experience with carbon pricing—has a regulatory burden that is one third of that in Ontario according to a survey of small businesses. Ontario is the province whose last government is most guilty of stacking regulatory and spending programs on top of a carbon pricing mechanism.

The benefits of Ontario shifting from a spending and regulatory approach to a carbon tax approach appear sizable. Enviroeconomics evaluated the costs of the current Ontario plan both in terms of cost per tonne of carbon removed as well as the cost to the average Ontario household. They compared these costs to the Federal government backstop—a carbon tax that rebates almost all revenues to households in proportion to the amount collected in each province. From their executive summary:

We estimate the total cost of the Ontario Plan in 2022 is $334 million with an average cost of $62 per tonne removed. In 2022, the Plan could reduce Ontario’s GHGs by about 3.3 percent or 5.4 megatonnes (Mt). To achieve the 18 Mt reduction in 2030 sought by the Plan, we calculate the average cost per tonne of emissions removed is $69, with a total cost of $1.23 billion. The cost of the Federal approach for the same level of emission reductions is estimated to be $214 million in 2022 with an average cost of $40 per tonne removed. In 2030, total costs are $811 million with an average cost of $45 per tonne.

Ontario could, in other words, get a large dividend from introducing a carbon tax and eliminating this entire suite of programs.

A Special Case: The Excise Tax on Fuel

One of the most effective political tactics used to fight carbon taxes has been to point to the increase in gasoline prices that would result from the introduction or an increase in carbon prices. “Let’s tax polluters and not commuters” has been an effective political charge in this debate.

The trouble is that raising the price of gasoline is utterly unnecessary until and unless the carbon tax being proposed is multiples higher than any being proposed today.

The total excise tax currently placed on gasoline at the federal and provincial level is the equivalent of a carbon tax of well over $100 per tonne. And the excise tax currently placed on diesel fuel is the equivalent of a carbon tax of over $60 per tonne. Those differ across provinces. For example, the carbon price equivalent of current excise taxes on gasoline is $111.76 in Ontario and $104.07 in Alberta. And the carbon price equivalent of current excise taxes on diesel fuel is $68.16 in Ontario and $63.31 in Alberta.

These taxes can be viewed as a user fee for drivers using public roads, and a popular one at that because taxing gasoline seems much easier than taxing—or tolling—roads. Having said that, some or most of the existing excise taxes on gasoline and diesel fuel are part of our first dividend—capturing the costs of emissions in terms of climate change and non-climate change externalities.

In practice this means that some portion of existing excise taxes on gasoline and diesel fuel are capturing the externalities associated with climate change and increases in the economy-wide carbon price may not require concurrent increases in gasoline and diesel fuel prices. That would certainly temper an attractive political argument against carbon taxes—the increase in pump prices.

The Third Dividend - Conclusion

An over-arching carbon price should be enough to provide all the incentives necessary to reduce GHGs. Yet layered on top of the carbon price in Canada are all sorts of costly, unnecessary, and economically damaging policies. And to the extent that eliminating these programs can save money or reduce the size of government, taxes could be lowered further.

Eliminating redundant or overlapping green regulations, spending, and taxes in favour of a carbon tax produces the third dividend.

The New Federal Climate Plan

The headlines following the newest federal government climate plan were all about the carbon tax rising $15/year starting in 2023 and reaching $170/tonne by 2030. This plan was widely praised as bold, aggressive and positive. And it is certainly all those things.

Yet, despite being bold, the plan demonstrates a surprising lack of confidence in carbon pricing. And that lack of confidence will be expensive for government and hamper economic growth for the economy.

The middle pillar of the government’s five-pillar plan lays out the trajectory for carbon taxes in Canada. The plan has, as we’ve said, been correctly lauded as positive and aggressive. The plan’s rationale for rapidly rising carbon pricing is very good and worth quoting: “A carbon price establishes how much business and households need to pay for their pollution. The higher the price, the greater the incentive to pollute less, conserve energy and invest in low carbon solutions.”

The federal government is right: a rising carbon price is precisely what we need to pollute less. It is what we need to conserve energy. And it is what we need to drive investments in low-carbon solutions.

Yet if the government really believed its own logic, many of the items in the rest of its climate plan would not have made an appearance.

There are all kinds of new spending and/or new regulations designed to encourage us to pollute less. Things like cash incentives for electric vehicles and new fuel mandates. Why do we need these if the rising price of gas will drive consumers to buy cleaner vehicles and the energy industry to produce less-carbon-intensive fuels?

There is new spending and/or new regulations designed to encourage us to conserve energy. Things like home retrofit programs and new regulations for new buildings. Why do we need these if the rising costs of heating is all the incentive people and businesses need to retrofit their homes and office towers?

And there’s new spending to spur investment in low-carbon solutions like programs for clean energy, cleaner electricity grids, and billions for the Canada Infrastructure Bank to partner with the private sector. But why do we need these things if appropriately pricing pollution through a higher carbon price is all the price signal needed to drive investments in clean energy and clean electricity—with or without the infrastructure bank?

There are one hundred and four initiatives in the federal government’s new plan. Sixty-three of these are things the federal government should be doing—paying for basic research; funding transition programs from old jobs (coal plants) to new jobs (biofuel plants); retrofitting the government’s stock of buildings and vehicles. There’s also a lot of stuff that the government was doing anyway that has been rebranded as climate policy. Finally, and importantly, there are five proposals to deal with emissions that a carbon tax doesn’t capture—mostly around regulating and incenting lower methane use.

These are all good and defensible policies. Government needs to reduce its own carbon footprint and should spend money to do so. It should fund research and fund transition programs for hard hit industries. And this is not the first time a government takes credit multiple times for the same policy.

But the plan also has 26 unnecessary spending items totalling $41 billion spread over ten years. These include building energy retrofit programs, numerous subsidies to encourage electric vehicles and charging stations, and cutting corporate taxes for companies making zero-emission products. These are front-end weighted with half being spent in the first three years. That’s $20 billion over the next three years in mostly unnecessary spending on things that the carbon tax is designed to do anyway.

That $20 billion amounts to an expensive lack of confidence in what a carbon tax can do.

The plan also contains ten new or stiffened regulatory regimes creating additional red tape—largely in the transportation sector. These include renewable fuel mandates, new clean fuel standards, fertilizer emission standards, greening building codes, and various vehicle emission regulations These are unnecessary regulations to change behaviour that should change anyway in the presence of a rising carbon tax.

That’s a lot of red tape caused by a lack of confidence in what a carbon tax can do.

A carbon tax should be the centerpiece of Canada’s plan to address climate change. It harnesses the power of markets and private behaviour to reduce emissions in the most efficient way. The federal government was right to put forward an aggressive plan to increase the carbon tax. There are good reasons for policies that fill in gaps that the carbon tax doesn’t cover, and of course the government must do its part to lower its own emissions.

But this five-pillar plan would have been much better and indeed much bolder—more efficient, less expensive and with less red tape—if it was a one-pillar plan that let the carbon tax do its work.


Canadian governments have the opportunity to generate a triple economic dividend if they implement a carbon price in the right way.

The first dividend is the result of properly pricing carbon to account for climate change as well as other costs that the burning of fossil fuels produces.

The second dividend is the result of using the revenues from a carbon tax to cut more economically harmful taxes like personal or corporate income taxes.

The third dividend comes from relying primarily on carbon pricing to reduce emissions which means governments can reap the economic benefits of eliminating all of the costly and unnecessary programs current in place to address climate change.

The federal government’s recent climate plan aggressively increases carbon taxes in Canada through to 2030. But it also contains numerous spending, tax, and regulatory measures that are redundant at best and very expensive at worst. The federal government should put more confidence in its carbon tax and eliminate these other measures.

About the Author

Ken Boessenkool is the J.W. McConnell Professor of Practice at the Max Bell School of Public Policy at McGill University.

He is also a Research Fellow at the CD Howe Institute, contributor to The Line and President and founder of Sidicus Consulting Ltd.

Ken was a founding partner of Kool Topp & Guy Public Affairs with Don Guy and Brian Topp. In the course of his career, he has served as Senior Counsel at GCI Canada, was Senior Vice President and National Practice Director for Public Affairs at Hill & Knowlton Canada, where he was also General Manager for Alberta and Manager of Business Development. Ken was a senior regulatory economist with two electricity firms. He once worked at a bank.

Ken has played senior strategic and policy roles in four national election campaigns for the Conservative Party of Canada under the leadership of Prime Minister Stephen Harper. He has been Chief of Staff to former Premier of British Columbia Christy Clark, a senior policy advisor to three national Conservative leaders, two Alberta Finance Ministers, and an Ontario Progressive Conservative Leader. He has played senior policy or campaign manager roles in three national leadership campaigns and three provincial leadership campaigns.

Ken is the Board Chair of Sonshine Community Foundation – a Calgary based women and children’s shelter. He has served on the following boards: The Canada-Israel Committee; The Cantos Music Foundation (now the National Music Centre); The Canadian Friends of Hebrew University, Civitas and the Forum of Federations.

Ken has published numerous policy and academic papers on a range of key national issues and is a frequent contributor to numerous online and print publications.


Back to top