Consumption-related greenhouse gas emissions in Canada, the United States and China
Archived Content
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
Craig Gaston, Environment Accounts and Statistics Division
Although the location of greenhouse gas (GHG) emissions is not important as far as their contribution to global warming is concerned, it can be useful to know how final domestic expenditures on products and services in Canada cause emissions in other countries and conversely, how final expenditures elsewhere cause emissions in Canada. This is a "consumption perspective" on GHG emissions as opposed to the production perspective by which countries normally present their GHG emissions.
GHG emissions statistics are generally compiled according to the various sources of emissions within the geographical boundaries of a country. Emissions in other countries related to Canadian expenditures are not directly observable but can be approximated using input-output models that describe the flows of goods between industries and countries. 11 Input-output models have a long tradition at Statistics Canada and have been used with environmental extensions to estimate the energy and GHG effects of expenditures by Canadian households. 12
This article uses a novel multi-regional input-output (MRIO) model to trace the connections between domestic final expenditures on goods and services in one country and the resulting GHG emissions in another. The model represents the economies of Canada, the United States and China. The rest of the world is not specified explicitly; only the trade flows with the rest of the world are articulated.
The model was built for the year 2002 because that is the most recent year for which detailed input-output tables are available for all three countries. There has been rapid growth of China's economy since then and additional analysis has been done here to provide some insight into the effect of increased Canadian expenditures on Chinese goods since 2002. A more recent MRIO model would be necessary to capture the fine-grained changes in the world economy over the last decade.
When using a single-country input-output model (unlike the MRIO model that has been used here), the simplifying assumption that imports have the same embodied emissions as similar goods produced in Canada is required. The Canada-U.S.-China MRIO model addresses this shortcoming. While a substantial improvement on single-country models for this reason, constructing the MRIO model requires a number of assumptions and considerable manipulation of the individual countries' input-output tables. The results presented here should, therefore, be considered experimental and taken as illustrative rather than final (see the textbox below for further details). 13
Some notes on terminology and data quality
"Direct" emissions refer in this study to the quantity of GHGs required to produce a good or service purchased by a final consumer and to GHGs emitted by household fuel consumption. "Indirect" emissions are those resulting from the production of intermediate goods and services required to make the final products.
For convenience, we refer to "emissions embodied in imports" as GHGs generated in a foreign country in order to produce goods and services imported into Canada. The purpose of a MRIO model is to translate final expenditures in one country into production in the countries distinguished in the model. Emissions can then be calculated by applying country-specific intensities (emissions per unit of production) to the outputs of each industry in each region.
Multi-regional models present difficult conceptual and practical problems. To keep the model manageable, we have focused only on the United States and China as they are Canada's two most important suppliers. China's volume of trade with Canada has been increasing steadily over the last decade.
The United States and China together accounted for about two-thirds of Canada's imports in 2002. China's share of Canada's imports almost doubled from 4.6% in 2002 to 8.8% in 2006, partly at the expense of the United States. It cannot be assumed that by modelling two-thirds of Canada's trade, two-thirds of our imported emissions have been modelled, but a crude estimate of the missing emissions can be made by using the U.S. model as a proxy for the rest of the world.
Even combining only three countries in a single model requires a considerable amount of data. Input-output tables of a sufficient size and quality are not available for each country every year. Although annual Canadian tables exist up to 2008, the most recent U.S. and Chinese benchmark input-output tables are for 2002. There are more recent tables available but they are smaller and of lower data quality. The MRIO model is thus based on 2002 data.
A uniform valuation of goods and services that maintains a consistent relationship between currencies and physical production is ideal, but this becomes difficult with respect to China because of the uncertainty of the exchange rates. Although market exchange rates are appropriate to convert dollars into Chinese currency, these are not necessarily compatible with the relationship between production and emissions in China. Some adjustments for purchasing power parity to take account of this problem have been made, with the assumption that the problem is inversely proportional to the export intensity of each industry.
This paper provides a consumption perspective on GHG emissions, or a consumption "footprint." From this point of view, the focus is on emissions related to final domestic expenditures.
It is important to note how emissions associated with trade between countries are treated in this study. The foreign emissions attributed to Canadian imports are those emissions generated in producing the goods and services purchased by final Canadian consumers (for example, imported food sold in supermarkets). The foreign emissions associated with intermediate imports 14 are excluded from the analysis except insofar as those imports end up incorporated into goods and services ultimately sold to final Canadian consumers. Similarly, the foreign emissions associated with intermediate imports that are incorporated into goods and services ultimately sold to final consumers in other countries are attributed to those other countries. This means, for example, that the emissions in the United States from producing auto parts that are imported into Canada to make a vehicle that is then exported to the United States belong to the United States in this study; these emissions appear in the U.S.-U.S. cell of Table 3. For this reason, one cannot determine the emissions related to a country's total imports from the results in this study.
It is also important to note that the MRIO used in this study is not the only model possible for estimating the GHG emissions footprint of consumption. There are a number of different approaches that can be used in doing so. Another approach that has been taken is to create an input-output model using data from the Global Trade Analysis Project (GTAP) database of the University of Purdue. 15 , 16 Depending on what input-output model is used and the data underlying the model, the results obtained will differ. This is unavoidable due to differences in the degree of detail with which national economies are described in the models.
The geography of greenhouse gas (GHG) emissions
Table 3 shows the geographical distribution of 2002 emissions for Canada and its two largest suppliers of goods and services. The first row shows emissions resulting from the domestic final expenditures in each region on Canadian goods and services. In total, 689 million tonnes (Mt) of carbon dioxide equivalent (CO2 eq) were emitted in Canada in 2002 in the process of satisfying Canadian domestic demand and demand from other countries. 17 , 18 Of this total, 217 Mt resulted from the production of Canadian goods and services destined for final consumers in the United States. Only 3 Mt resulted from Chinese demand for Canadian products.
The first column of Table 3 shows emissions in each region resulting from domestic final expenditures. In total, 530 Mt of GHG were emitted globally to satisfy Canadian domestic final expenditures. Of this, 401 Mt were emitted in Canada 19 while 58 Mt and 14 Mt were embodied in Canadian imports from the United States and China respectively. Only emissions embodied in imports used to satisfy final domestic expenditures are shown explicitly in Table 3. There are also foreign emissions related to Canadian exports, which are not shown explicitly, as these do not result from Canadian final expenditures. For example, the production in Canada of motor vehicles destined for U.S. consumers requires parts that are produced in the United States and then imported into Canada. The emissions resulting from producing these motor vehicle parts are included in the 6,232 Mt shown in the U.S.-U.S. cell of the table. This is appropriate because the United States is the ultimate consumer of the parts, which return in the completed vehicles, whereas Canada is only the proximate consumer. 20
The first row of Table 3 can be broken out by the industries in Canada that contributed the most to GHG emissions and assigned to the origin of the final expenditure (Chart 6). Of note is the fact that the U.S. purchases resulted in a relatively large quantity of GHG emissions from coal-generated electricity in Canada (23.8 Mt of CO2 eq). These are mostly indirect emissions resulting from the production in Canada of goods and services destined for export to the United States. Direct U.S. purchases of Canadian electricity generated from coal account for less than 10% of these emissions. In contrast, the real estate industry's emissions, most caused by providing heat and electricity to buildings, are dominated by Canadian purchases. (Canada's exports to China were small compared to Chinese imports so China barely appears in Chart 6.)
Greenhouse gas (GHG) emissions related to motor vehicle production
The purchase of motor vehicles provides a good example of how consumption in one country leads to GHG emissions in other countries. In 2002, Canadians spent $12.8 billion 21 to purchase motor vehicles produced in Canada (Table 4). This resulted in emissions of 1.7 Mt of CO2 eq in Canada, and 1.7 Mt of CO2 eq in the United States. In contrast, we see that U.S. purchases of their own vehicles had a relatively small impact in Canada (8.4 Mt) compared to the United States (102.7 Mt). The explanation for this is that the Canadian auto industry relied much more heavily on U.S. producers for inputs than the converse. Amongst these U.S. inputs to Canadian auto manufacturing are metal stampings, most of which were imported. The emissions from the U.S. primary metal sector required to produce those stampings are considerable. In neither Canada nor the United States did Chinese production contribute more than 10% of total emissions associated with motor vehicle purchases.
Greenhouse gas (GHG) emissions intensities
Table 5 shows selected direct and total emissions intensities underlying the GHG calculations in the MRIO model (total emissions are defined as direct plus indirect). Not all of these industries' products are significant in Canada's imports but they all contribute, directly or indirectly, to the emissions resulting from Canadian expenditures on goods and services.
Differences in emissions intensity of electric power generation largely reflect the mix of energy sources in each country. Hydro electricity accounted for about 60% of Canada's electricity generation whereas fossil fuels accounted for only 25%. 22 In China, coal and other fossil fuels accounted for 80% of electricity generation, with coal being the dominant energy source. 23 Fossil fuels accounted for 70% of U.S. generation, but one quarter of this was from natural gas, which has much lower GHG emissions than coal per unit of electricity produced. 24
China's relatively higher emissions intensities in textiles, motor vehicles and computer manufacturing reflect the higher emissions intensity of its largely coal-based electric power industry. Also, its relatively higher intensity for iron and steel production contributes to higher emissions in manufactured products.
Canada's emissions intensity for oil and gas extraction reflects the different mix of crude oil grades extracted in this country compared to the United States and China. Crude oil accounted for 20% of the rest of world's Canadian imports of GHG emissions in 2002.
The total emissions intensities for each county reported in Table 5 reflect similar relative levels as the direct intensities. One striking exception is the ratio of Canada's total-to-direct intensity for computer and electronic products, which is considerably higher than that of the United States and China. The Canadian industry is more oriented towards assembly of components than are its two trading partners, resulting in a lower direct intensity.
A more recent perspective
A more recent perspective on Canada's imported GHG emissions resulting from domestic final expenditures can be gained by running the 2002 MRIO model using Canada's 2006 import data and domestic final expenditures. The validity of this rests on the simplifying assumption that 2002 GHG emissions intensities and industrial technologies for all regions prevailed in 2006. Of course, updating the data for all variables in the model would yield more accurate results but it is the large relative shift in the source of Canada's imports from other countries to China that stands out during this period. In the absence of more recent data, the estimates produced using this simplified approach give a good indication of the trends in Canada's embodied emissions.
Canada's imports of embodied GHG emissions from each region increased by about the same amount, 17 Mt, between 2002 and 2006 for a total of 50 Mt. This was an increase of 39% over 2002 levels (129 Mt). China's share of the total increased from 11% to 17% at the expense of the United States and the rest of the world (Chart 7). China's share of the dollar value of Canada's merchandise imports almost doubled over the same period. The U.S. share declined by 8% due to the relative increase from other countries as well as from China (Chart 8). The relatively lower share of the value of merchandise imports from China compared to China's share of embodied GHG emissions reflects the higher GHG intensity per dollar for that country's imports compared to those from the United States and the rest of the world.
- Date modified: