Governments across Canada—including Quebec, Ontario, Alberta and British Columbia—have all implemented forms of carbon pricing. The federal government has also announced it will implement a carbon price “floor,” a minimum value that each province must put on carbon emissions. That price is set to start at $10 per tonne of greenhouse gas emissions in 2018 rising to $50 per tonne by 2022.

Most often, these carbon-pricing systems are justified because some economists and groups such as Canada’s own Ecofiscal Commission portray them as the most efficient way to reduce greenhouse gas emissions, while being better than the regulatory approaches that have heretofore been the norm.

The problem is this—for carbon pricing to be efficient, and economically benign, there are a set of conditions that must be met, but are not. Carbon pricing must replace regulations, not simply layered on top. Revenue should be rebated as broadly as possible to lower other distortionary taxes such as the personal income tax or corporate income taxes. And lastly, the revenues should not be used to further distort the energy economy by having governments “invest” in pet projects or selective forms of energy production. As a new study from the Fraser Institute shows, none of the jurisdictions that have enacted carbon taxes in Canada come anywhere close to meeting these conditions.

Consider Ontario’s cap-and-trade system instituted by Premier Kathleen Wynne, which her government estimated would bring in $2 billion in revenue per year. According to the Ontario auditor general, out of the $8 billion to be collected in four years, $1.32 billion will be earmarked to help with residential and business electricity bills. The rest will be spent on the usual governmental preferences—transit, subsidies to renewable energy, dubious efficiency programs, etc.

Or consider Alberta, with its new carbon tax of $30 per tonne. Phased in by 2018, it’s expected to generate almost $3.9 billion from 2017 to 2020. Part of that will be used to subsidize Alberta’s emitters (granting a windfall to the very people putting out most of the emissions). A small portion will be given to low-income Albertans, ostensibly to ease the pain of higher power bills. The rest will be spent on government pet projects.

And then there’s Quebec, which has a cap-and-trade system that has brought in $330 million but is expected to bring in $2.5 billion by 2020 (and probably more, as it will have to match the escalating national price floor established by Ottawa). Where does the revenue go? Free permits are given out to emitters while the remaining revenue will be spent on “programs to fight climate change.”

Finally, consider the much vaunted B.C. carbon tax. Another study by the Fraser Institute verified that in this tax’s early years, it was truly revenue neutral. Personal and corporate taxes were reduced and additional tax reductions were introduced to ensure revenue neutrality. But by 2014/2015, only five years into the tax system, the government had taken to shaky bookkeeping to preserve the appearance, but not the reality, of revenue neutrality.

Subsequent to the Fraser study, the B.C. government has restored revenue neutrality in the overall economic sense. But a closer look at the details shows that rather than purely rebating revenues to the general population, the government turned to special-interest tax credits (Industrial Property Tax Credits for Major Industry, School Property Tax Reduction for Farm, etc.) until $148 million (12 per cent) of actual offsetting tax measures went to those special interests. Again, the B.C. government has since restored the tax to overall revenue neutrality, but the set-asides are likely to continue.

Simply put, Canada’s experience with carbon taxes reveals an unpleasant reality. While there’s a textbook carbon tax that’s efficient and economically benign, governments have no interest in establishing or maintaining it. Carbon tax revenues are too tempting for governments to simply rebate—they often would rather use them to advance their political agendas.