•  About Me 
  •  FAQs 
  •  Mail Me 
  •  Links 
  •  Archives 

April 2, 2014 [Updated 9:30 AM to acknowledge Sweden.]

Carbon Tax Debunked

Even amongst climate change deniers it would be hard to find many who advocate increasing greenhouse gas emissions or burning more fossil fuels. Debate focuses on how rapidly fossil fuel consumption and greenhouse gas emissions can be reduced and what should be done to bring about the reductions. The most frequently discussed option is to put a price on carbon either through a carbon tax or through a cap-and-trade system. Less frequently discussed options include technological changes, such as more fuel efficient processes, and population control. Even with technological change, it is hard to argue that the world can sustain 7 billion people in 2014 as easily as it sustained 1 billion in 1804.

British Columbia is held out as a leading example in the use of a carbon tax. It started July 1, 2008 at $10/ton and increased by $5/ton each July 1st until it reached $30/ton in 2012, making it the highest price in the world on carbon. In the EU with its market determined price for cap-and-trade, carbon has been priced between 2.46 and 5.02 euros; it was trading below 4 euros ($6 Canadian) at the end of March 2014. California also has a cap-and-trade system with the trading price in early 2014 just above the state set floor price of $11.34. Sweden appears to have the world’s highest carbon tax, but BC’s is one of the highest. When BC went to $30/ton it was expected that North America would soon be adopting a continent-wide cap-and-trade system. That has not happened.

Two studies claim that BC's carbon tax has been successful in reducing greenhouse gases in BC. In "BC's Carbon Tax is Working Well after Four Years (Attention Ottawa)", Stewart Elgie and Jessica McClay (Canadian Public Policy, Vol. XXXIX, supplement 2013) compared BC and the rest of Canada's change in consumption of fuels that are subject to the tax. In "Salience of Carbon Taxes in the Gasoline Market", Nicholas Rivers and Brandon Schaufele (June 10, 2013, available at SSRN) argued that "the carbon tax imposed by the Canadian province of British Columbia caused a decline in short-run gasoline demand that is significantly greater than would be expected from an equivalent increase in the market price of gasoline". To the extent that they contribute to directing public policy to a carbon tax rather than to alternative methods for reducing greenhouse gases, I believe the studies require critical reviews.

Elgie worked with McClay and Rivers in 2012, producing Sustainable Prosperity's four year report on the carbon tax. His 2013 update produced with Jessica McClay included a table that looked separately at propane, motor gasoline, diesel, fuel oil, petroleum coke and aviation fuels; both his 2012 and 2013 reports added the fuels (measured by the cubic meter); I don't know what you get when you add a cubic meter of butane to a cubic meter of gasoline but that is the unusual concept used in the reports. In particular, the reports combined cubic meters of propane and propane mixes, butane and butane mixes, naphtha specialties, motor gasoline, stove oil/kerosene, diesel fuel oil, light and heavy fuel oils, and petroleum coke (aviation fuels were excluded because the carbon tax does not apply to flights outside of BC). Adding different fuels by volume is methodologically flawed. In order to add the different fuels they would have to be converted to identical things. They could have calculated the amount of greenhouse gases each produces or each fuel could have been converted to gigajoules of energy and then added. They even could have added the value of each fuel, but it does not make sense to add cubic meters of different substances.

When I replicated their method and extended the time period back to July 1, 2000, I discovered the downward trend in the consumption of fossil fuels preceded the introduction of the carbon tax. They make passing reference to the pre-tax period when discussing their strange mix of different fuels, but they did not report how consumption of each separate fuel changed since 2000.

Data used by Elgie and McClay for consumption of those fuels are available in Statistics Canada's CANSIM table 134-0004. Data for population is available in CANSIM table 051-0001. Like them I converted the data to annual data aggregating July 1 through June 30, so 2001-02 is July 1, 2001 through June 30, 2002. The annual data was converted to per capita data following their method of using those aged 15 and over. Those results going back to July 1, 2000 are shown in the graphs and table below which identify the periods before and after the introduction of the carbon tax.

Notice in the table that only butane and heavy fuel oil showed BC not leading the rest of Canada in reduced per capita consumption. By volume, motor gas, diesel, propane and heavy fuel oil account for 92% of the fuels used in Elgie's analysis. Claims that BC led the rest of Canada as a result of the carbon tax are equivalent to seeing a crowd go down the street and rushing to the front to claim leadership of the parade. The trend was underway before the carbon tax was introduced and nothing in Elgie and McClay's analysis shows the tax changed the already established trend.

carbon tax graphs

carbon tax table

Classical economists wouldn't be surprised that the carbon tax is relatively ineffective in reducing consumption because hundreds of studies have shown that fossil fuels, gasoline in particular, are price inelastic. Price elasticity is the percentage reduction in demand caused by a percentage increase in price. If demand goes down by more than 1% when price increases by 1%, demand is said to be elastic; it is said to be inelastic if demand goes down by less than 1% for a 1% price increase. Gasoline is price inelastic; economists would expect it to take a very big increase in price to cause a substantial decrease in consumption of gasoline. Notwithstanding that orthodoxy, defenders of the carbon tax have come to the rescue!

Nicholas Rivers and Brandon Schaufele argue:
"Our main result is that the BC carbon tax generated demand response that is 7.1 times larger than is attributable to an equivalent change in the carbon tax-exclusive price. In our preferred model, a five cent increase in the market price of gasoline yields a 1.8 reduction in the number of litres of gasoline consumed in the short-run while a five cent increase in the carbon tax, a level approximately equal to a carbon price of $25 per tonne, generates a 12.5% short-run reduction in gasoline demand."

If that is true, text books on economics need to be re-written! However, Rivers and Shaufele cite other studies that support their argument that the effect of taxes on prices can result in different consumer behaviour than the result of market forces on prices. Perhaps the text books really do need a re-write. When the HST was being debated in BC, tax experts came out of the woodwork to argue that all self-respecting economists support value added taxes. Orthodoxy around a price increase is a price increase regardless of the cause is no less rigorous, yet I can find no economists who specialize in tax policy who have refuted Rivers et al. That may be because there are only a small handful of articles that challenge conventional thinking that a dollar is a dollar regardless of the cause of a price increase.

Rivers and Schaufele applied several multiple regression techniques to a model that included BC's carbon tax. They then interpreted the coefficient on the carbon tax to be equivalent to a measurement of quasi-elasticity of demand. Many criticisms can be made of measurement without theory, but in this case it should be noted that notwithstanding their use of a monthly data set that contained 2580 province-month observations, only 42 of the 2580 observations had a non-zero value for BC's carbon tax, and of those 42, only 4 values were possible (i.e. it was a discrete variable which poses additional challenges in interpretation). The primary conclusion and argument in their paper focuses on an interpretation of the coefficient of that variable which had a value of zero for 2538 of the points in the data set.

Rivers and Schaufele might have a point, but without theory and founded on shaky empirical work, 42 observations, they have a long way to go to convince the majority of economists that it is time to re-write the textbooks. Environmental activists are quick to pick up their study and promote it as fact. The website Atlantic Cities spun their research as proving "BC's carbon tax totally works". The contrary argument is a handful of well-intentioned environmentalists are off on the wrong course and advocating policies that will take resources away from far better alternatives that could save our planet. The jury is out on whether BC's carbon tax significantly accelerated the downward trend in the consumption of fossil fuels that had started years before its July 1, 2008 introduction.