Canadians have been buying more gasoline every year. Over the last 20 years, fuel demand has risen from about 36 billion liters to about 43 billion liters. For anyone concerned about climate change, this is the wrong direction. If Canada is serious about reducing its carbon dioxide emissions, the graph below should start showing a reversal. Why is fuel demand still rising even though many policies—fuel taxes, carbon taxes, feebates and subsidies for cleaner cars, fuel economy standards—incentivize lower fuel consumption? This article will provide some answers, and makes it clear that the solutions for the problem require a mix of different approaches.

click on image for high-resolution PDF version

In order to understand the upward trend in gasoline demand, it is necessary to disaggregate fuel demand into separate components and analyze what is going on with each of these components separately. The fuel demand identity equation below \[F = N\cdot\left[\frac{V}{N}\right]\cdot\left[\frac{M}{V}\right] \cdot\left[\frac{F}{M}\right] \] links total annual fuel demand (F) to population size (N), the number of vehicles (V), and the mileage driven (M). The ratios are the number of vehicles per person \(N/V\), the annual mileage per vehicle \(M/V\) in kilometers per year, and the average fuel economy \(F/M\) in Liters per kilometers. The above equation is an identity because the expressions on the right hand side cancel out to yield just F, so the left-hand side and right-hand side are indeed identical. Much research (including my own) has gone into studying the drivers of gasoline demand, and figuring out implications for public policy There are also important regional differences, which the net graphs will show.

Once gasoline demand is expressed in per-capita terms, it becomes apparent that these lines are pretty flat. In most provinces, per-capita fuel demand is quite stable. Only in British Columbia it has trended a bit more downwards, which research links to the beneficial effect of carbon pricing. Without such pricing, B.C.'s slight downward trend would look as flat as the other provinces. The graph suggests that in three of the four large provinces, rising gasoline demand is simply linked to population growth. More people—more demand for gasoline. However, this still isn't the full picture. As the fuel demand identity suggests, we need to understand three other components: vehicle ownership, how much people drive, and how fuel-efficient our cars are. The next graphs show the trends for two of the three components. Unfortunately, there are no continuous surveys about that could reveal how much annual mileage has changed over the years. What we do see are large differences across provinces in terms of fuel demand per capita. This reflects differences in the mix of urban and rural driving. Driving in urban areas is typically shorter, and this is part of the reason why British Columbia and Quebec have lower per-capita gasoline demand than Alberta and Ontario, where population is less clustered geographically.

click on image for high-resolution PDF version

The next element in the fuel demand identity looks at car ownership. As countries grow richer and people more affluent, there is a tendency to own more cars. Many households acquire a second car, and this is also associated with a net increase in mileage per household. Many countries, Canada included, have also experienced a decrease in household size, and today we see many more one-person households than two decades ago. All these extra households also have specific mobility needs, which tends to raise car ownership. The next graph shows how this trend plays out in four of Canada's large provinces. The trend is clearly upwards: there are more cars per person today than two decades ago. This is even true in British Columbia, where per-capita car ownership has been increasing by 1.8% per year, slightly more than Ontario's 1.6% per year, but much less than Alberta's 2.7% per year.

click on image for high-resolution PDF version

The last graph combines several stories. It shows fuel demand per registered vehicle. Thus it combines what it happening at the extensive margin of car ownership and the intensive margin of car use. As the vehicle fleet turns over, the average fuel economy improves as more modern cars tend to be more fuel-efficient. Public policy has steered people to owning more fuel-efficient vehicles, while at the same time motorists have flocked to owning bigger cars and light trucks (SUVs). The net result, surprisingly, is still positive. Technology has been moving faster than the trend towards heavier vehicles. Fuel economy has been improving in all provinces, but faster in those with active public policies (in particular, higher fuel taxes). Since 2014, fuel economy improvements have been stagnating, however. Differences across provinces are a mixture of fleet composition (average fuel economy of vehicles) and mileage per year (rural versus urban driving).

click on image for high-resolution PDF version

What are the policy conclusions that can be drawn from the above empirical observations? Population growth will continue, as well as the trend towards smaller households. Income increases will increase demand for mobility. The ways to change the effect of gasoline demand and carbon dioxide emissions are all well known. The first policy aims at reducing the volume of individual driving by expanding public transit. The second policy aims at making cars more fuel-efficient by setting fuel economy standards and incentivizing purchases of fuel-efficient cars. Ultimately, the large-scale transition to electric mobility offers the possibility of shrinking fossil fuel use to a fraction of what it is today. It is clear that there is no single policy that achieves both. One one hand, improving public transit requires large infrastructure investments combined with putting an appropriate price on road congestion, vehicle-miles traveled, and vehicle emissions. On the other hand, we need to invest in the technologies (namely: cheaper batteries) that will facilitate the large-scale transition to e-mobility that will happen through market forces when electric vehicles are cheaper than internal combustion engine vehicles. To make this happen, public policy needs to make the necessary investments that are conducive to public transit expansion and e-mobility expansion.

Sustainable mobility in the 21st century must become emission free and put appropriate prices on all negative externalities in order to achieve efficient mobility mode choices. Such policy changes require significant political determination and careful understanding of distributional outcomes. Policy changes that are considered "unfair" will likely be resisted strongly, as past experience shows. Technological progress will do its bit to make e-mobility happen. However, to expand public transit (an absolute necessity as our population continues to grow) we need forward-looking public infrastructure investments. If Canada wants to reverse the trend towards more and more fuel use depicted in the first graph above, public policy needs to step on the accelerator—the one that doesn't use gasoline.

Data sources: All data in this article were obtained from Statistics Canada CANSIM Table 23-10-0066-01 Sales of fuel used for road motor vehicles, annual, Table 23-10-0067-01 Vehicle registrations, by type of vehicle, and Table 17-10-0005-01 Population estimates on July 1st, by age and sex. Vehicle use (mileage) used to be reported in the Annual Canadian Vehicle Survey, but this survey was discontinued after 2009.