CO₂ emissions and energy prices in the Asia-Pacific region

CO₂ emissions for a country depend directly on the size of the economy or the size of the population. In the following analysis, population has been chosen as the normalizing factor, so an attempt is made to explain the behaviour of CO₂ emissions per capita. If CO₂ emissions can be explained in terms of economic criteria then it will be possible to calculate the marginal impact of a change in CO₂ emissions for a change in the independent variables. The independent (explanatory) variables are simply the real GDP per capita and real price of energy.

Why look at the whole Asia-Pacific region rather than just Canada? Canada is one of the few industrialized Annex I countries in the region. The countries in the Asia-Pacific region are at different stages of economic development, so it is relevant to see whether the Canadian experience is typical or atypical of developing countries.

The Importance of Real GDP per Capita and Energy Prices for CO₂ Emissions

Since CO₂ emissions are so closely tied to energy use, it seems natural to attempt to explain CO₂ emissions per capita as a function of the same sort of economic variables used in explaining energy demand, namely real GDP per capita and real energy prices. It is hypothesized that the marginal response to real GDP per capita and to real energy prices are the same for each country, but the intercept term could be different to account for country-specific CO₂ emission infrastructure.

The model consists of twenty countries in the Asia-Pacific and South Asia region. The countries are Bangladesh, Hong Kong, India, Indonesia, Japan, New Zealand, South Korea, Malaysia, Pakistan, Philippines, Sri Lanka, Taiwan, and Thailand in the Asia region and Canada, Chile, Colombia, Ecuador, Mexico, Peru, and United States in the Americas region.

It would be ideal to have an aggregate price index for all energy types, but such an index is not always available for developing countries. Since it is necessary to have the same measure of real prices for all countries and since oil has the largest share amongst energy fuels in most countries, the price of fuel oil in the industrial sector is chosen as the unit of measure.

Data for Asia-Pacific Region and the Economic Model

Data for CO₂ emissions, real fuel oil prices in the industrial sector, real gross domestic product (GDP) per capita, and exchange rates have been collected from 1977 to 1992 for twenty countries in the Asia-Pacific and South Asia region. The data were obtained from a number of different sources.

The CO₂ emissions per capita data show an obvious pattern of increasing CO₂ emissions per capita with real GDP per capita. It is less obvious that CO₂ emissions per capita decline as real fuel oil prices increase.

Econometric Results for Asia-Pacific Region

Those countries with either large populations or large economies have high CO₂ emissions. In order to compare CO₂ emissions between countries, they are measured in per capita terms. Generally, the industrialized countries (Australia, Canada, Japan, New Zealand, and USA) have the largest CO₂ emissions per capita. Some developing countries with relatively high levels of real GDP per capita (Hong Kong, South Korea, Singapore, and Taiwan) also have relatively high levels of CO₂ emissions per capita. Despite its low level of economic development, China has relatively large levels of CO₂ emissions because of the very significant share of energy consumption met by coal. Canada has the second highest level of CO₂ emissions per capita behind the USA.

The data support the hypothesis. CO₂ emissions per capita are positively related to real GDP per capita and negatively related to the real price of fuel oil. When the countries are grouped together by income class, the real price of fuel oil is only statistically significant for high-income countries. The coefficient for real GDP per capita in low-income countries is lower than it would have been had China been included in the sample.

The GDP elasticities show an almost unit elasticity for low-income and middle-income countries and an elasticity of only 0.36 for high-income countries. The price elasticities are all very inelastic, although they show a more inelastic response in low and middle income countries.

The smallness of the price elasticities is surprising. To test the robustness of the econometric results, the cross-section of 20 countries is estimated for 1992 only. The elasticities are larger – and more realistic. For instance, the elasticity of CO₂ emissions per capita with respect to real GDP per capita is 1.39 (more than double the previous elasticity) and with respect to real price of oil is –0.54.

The smallness of the price elasticities for aggregate CO₂ emissions could be related to the fact that we used the fuel oil price in the manufacturing sector as a proxy for an aggregate energy price index. The IEA provides a disaggregation of CO₂ emissions by fuel and by end-use sector but only for 1990 and 1996. Therefore, it is possible to test the size of the estimated price elasticity using the 1990 cross-section data of 20 countries for all CO₂ emissions, for CO₂ emissions emanating from oil use, for CO₂ emissions emanating from oil use in manufacturing, and finally for CO₂ emissions emanating from oil use in road transportation. Using the same variable for oil price, the price elasticity is larger for oil-initiated CO₂ emissions than for aggregate CO₂ emissions (-0.50 vs. –0.36) and larger for oil-initiated CO₂ emissions in the manufacturing sector than aggregate CO₂ emissions (-0.45 vs. –0.36). Finally, if a new oil price variable is created that is a quantity-weighted average of the price of motor gasoline and diesel oil then the price elasticity for oil-initiated CO₂ emissions in the road transportation sector is smaller than for aggregate CO₂ emissions (-0.16 vs. –0.36).

In conclusion, it appears as if the combination of using a pooled model with dummy variables and using the price of oil in the manufacturing sector as a proxy for an economy-wide energy price index leads to relatively low price elasticities for CO₂ emissions. Nevertheless, the inclusion of dummy variables is an important element of the model specification, and the proxy energy price index is necessary because of unavailable data.

Implications for the Canadian Economy

Reducing CO₂ emissions is difficult. The last section demonstrated that most of the variation in CO₂ emissions per capita can be explained by real GDP per capita and real energy prices. Therefore, to reduce CO₂ emissions per capita one must either reduce real GDP per capita or increase real energy prices, assuming no change in technology. The relative magnitude of the elasticities in the various models means it is much easier to reduce CO₂ emissions by lowering real GDP per capita than by increasing energy prices. Of course, any government policy that proposed to meet the Kyoto target for reducing greenhouse gas emissions by reducing real GDP per capita would likely not be endorsed by the population.

If the government chose to reduce CO₂ emissions by increasing the real price of energy then it would have to impose some sort of energy-specific tax. Governments in Canada had been involved in setting domestic prices for oil and natural gas from 1973 until about 1985. Since 1986, energy markets (not governments) have determined oil and natural gas prices in Canada. The federal government has repeatedly stated recently that it would not impose any sort of energy-specific carbon tax.

The government of Canada expects to be able to meet its Kyoto target for reducing greenhouse gases through technological improvements that reduce energy intensity in economic activity. Therefore, the government is primarily focusing on the energy intensity component. Overall energy intensity has been generally declining in Canada since about 1973. Canadian energy consumers have already made significant shifts away from oil into natural gas and electricity. Unless government gets involved in setting fuel prices there is no obvious method to direct energy consumers to shift away from carbon-intensive fuels.

If the Canadian government were to use energy-specific taxes to induce consumers to reduce CO₂ emissions then what impact would it have on the economy? To fully assess the impact it is necessary to use an econometric model that integrates the energy sector and the macro-economy. Such a model shows that relatively large changes in oil prices would probably not lead to huge changes in key variables of the macro-economy. Consequently, the government could probably impose an energy-specific carbon tax without severely impacting performance of the macro-economy.

Canada’s CO₂ Emission Record from 1990 to 1996

CO₂ emissions in Canada and all other industrialized Annex I countries of the Asia-Pacific region (Australia, Japan, New Zealand and USA) have increased significantly between 1990 and 1996. Compared to 1990 levels, the countries agreed to reduce their greenhouse gas emission levels by 2008 to 2012 by 6% for Canada and Japan, 8% for New Zealand, and 7% for the USA; and Australia agreed to an 8% increase (Kyoto target). None of the countries are close to meeting the target.

The growth of CO₂ emissions from China, India, Indonesia and South Korea is much larger than it is for the industrialized Annex I countries. The growth of CO₂ emissions range from a low of 33% in China to an astounding 75% for South Korea. It appears as if none of the countries in the Asia-Pacific region will meet the Kyoto target.

Future Levels of CO₂ Emissions from Canada?

Several organizations have provided a forecast of future levels of Canadian CO₂ emissions in 2010, the approximate year by which countries have committed to meet the Kyoto targets. The forecasts do not allow for unknown policies that might be introduced in the future to encourage consumers to reduce CO₂ emissions. The main sectors to contribute greenhouse gases are transportation (the largest), fossil fuel production, electricity generation, and industrial (each with roughly equal shares).

The US Energy Information Administration predicts that Canadian CO₂ emissions from all energy fuels will grow by about 1 percent per year from 1996 to 2010 with most of the incremental emissions arising from oil and natural gas and none from coal. On the other hand, APERC predicts that CO₂ emissions from Canada will increase by about 2 percent per year from 1995 to 2010, with substantial incremental emissions coming from the use of coal to produce electric power.

None of these organizations are predicting a decline in CO₂ emissions for Canada, which is necessary if Canada is to meet its Kyoto target.

Summary

The levels of CO₂ emissions from Canada are viewed from the context of other countries in the Asia-Pacific region. A simple econometric model identified energy prices and especially real GDP per capita as the primary explanatory variables for CO₂ emissions. The potentially small value for the price elasticity and the declaration by the Canadian government that carbon taxes would not be used to encourage consumers to reduce CO₂ emissions leaves the policy-makers with limited options in attempting to meet the Kyoto target levels of emissions. Judging by the growth of CO₂ emissions from 1990 to 1996 in other Asia-Pacific countries, policy-makers everywhere in the region are having difficulty creating an environment that would convince consumers to change their consumption habits.