Online Appendix 1
An introduction to Environmentally Extended Input-Output analysis
Environmentally extended Input-Output analysis (EEIOA) is an internationally accepted method for evaluating economic and environmental interdependencies on a large scale. It is an accounting procedure that documents all monetary and environmental flows to and from discrete economic sectors, which together cover all traditional economic activity in an economy thereby accounting for all different stages during which a product has environmental impacts (for example: production, transport, usage, disposal as waste etc.) and measuring these impacts quantitatively.EEIOA can be understood to be ‘top down’ modelling of aggregate products i.e. the environmental impacts of a diet or supply chain, rather than individual products (Heller, Keoleian, & Willett, 2013).
The advantage of EEIOA over process based Life Cycle Assessment (LCA) is that it can be used to calculate the hidden and indirect environmental, economic, and social impacts associated with downstream consumption(Baumann, 2013; Berners-Lee, Hoolohan, Cammack, & Hewitt, 2012; Duchin, 2005; Foran, Lenzen, & Dey, 2005; Kitzes, 2013; Roy et al., 2009) Unlike process based LCA there is no double counting of impacts in complex supply chains. Furthermore with EEIOA, complex problems - such as the altering of diets - can be modelled with greater ease than via process based LCA. Finally IO can provide economic and social analysis in addition to the environmental analysis offered by process based LCA.
EEIOA uses Input-Output (IO) tables with additional ‘satellite’ data sets to model environmental or social impacts(M Lenzen, Moran, Kanemoto, & Geschke, 2013)— the IOTs themselves only provide a ‘snap-shot’ of the size and structure of the economy i.e. the economic impacts (refer to Table 1 for schematic of an IOT including ‘satellite’ data). Explained comprehensively by Miller and Blair (2009)and Murray and Wood(2010), an IO table explains macroeconomic relationships of an economy in an accountable format, tracing monetary flows from the sectors which used the goods and services to the sectors which supplied them and vice versa.
In reality an IO table is a matrix of numbers with the rows and columns labelled as sectors of the economy. The intermediate sectors — the sectors of the economy that produce goods and services — are listed on both axes, allowing for an intersection between every intermediate industry sector: . Primary inputs, ,are inputs to a sector that are exogenous to the economy, and thus cannot be traced through the economy. Final demand,, is the consumption of goods or services in their final form by households, the government, export etc, in this simplified example a vector. It is assumed that for an IO table to account for the economy that Gross output or Total Sales,, when transposed equals Total Expenditure, ’, which in turn means that the column sums of are equal to the row sums of. Dividing each cell of by the corresponding sectors of provides the direct requirements matrix,=where the ‘hat’ over a vector denotes a diagonal matrix with the elements of the vector along the main diagonal and zeros elsewhere, for instance, then. The direct requirements matrix gives the direct proportion of each input needed to create one unit of output; in other words a ‘production recipe’ for each sector(2009).
Table 1 Schematic of an IO table
Intermediate demand / Final DemandPrimary Sector / Secondary Sector / Tertiary Sectors
Agriculture
Mining
Fishing
Forestry / Construction
Plastics
Steel
Automotive / Retail
Banking
Hotels
Waste / Household
Intermediate Inputs / Primary Sector / Agriculture / /
Intermediate sectors /
Final Demand / Total Sales/Output ()
Mining
Fishing
Forestry
Secondary Sector / Construction
Plastics
Steel
Automotive
Tertiary Sector / Retail
Banking
Hotels
Waste
Primary Inputs / Household /
Primary Inputs to Production / Primary Inputs to Final Demand
Government
Capital
Imports
Environmental
Satellite / Waste 1 /
Environmental impacts
generated by Intermediate sectors
…
Waste n
Total Expenditure (’)
The economy as shown within the IO table can be represented by . Solving this equation for yields the following equation
(1)
Where is an Identity matrix, and a vector. Equation (1) is also known as the Leontief inverse matrix or the total requirements matrix, and allows the determination of a production recipe when given an exogenously specified final demand.
IO tables are usually discussed in terms of basic price; the price it costs the producer to produce the product. However, it should also be noted that IO tables are not flat 2 dimensional tables, as multiple tables stack on each other to account for different economic effects. In addition to the basic price table, there can be IO tables for the costs of subsidies, taxes, trade, and transport. Added together theses tables become the purchasers price.
EEIOA not only captures the direct economic or environmental effects of producing a product but also the effects upstream, throughout the production process. This has been addressed in detail in a number of publications(Kitzes, 2013; Miller & Blair, 2009; Murray & Wood, 2010), but in brief, each product has many layers of production underneath; in turn these layers have their own layers of production under them(Figure 1). There can be loops or cycles in production layers i.e. production of water pipes requires steel, the production of steel requires water which in-turn requires water pipes, etc. Regardless of the exact composition of the production layers, they all lead to the population for consumption.
Figure 1. An example of the layers of production, showing the upstream inputs to make a product for final consumption by the population.
Some forms of environmental impact assessment only account for the first (and possibly the second and third) layer of production, ignoring many of the upstream impacts. This truncation error can mean that up to 50% of the environmental effects are unaccounted for (Manfred Lenzen, 2000; Majeau-Bettez, Strømman, & Hertwich, 2011). The power of EEIOA is that it can capture the interrelationships between various layers of production and attribute – in a simple and consistent manner – the total environmental impacts of production () to finished products.
In Table 1, is a vector of the amount of environmental impacts per sector . From this we can create a link between economic activity and environmental impact – the amount of pollution emitted per (monetary) unit of output of sector , . Wiedmann(2013)andKitzes(2013)refers to this as the direct intensity multiplier, or the direct ‘pollution’ coefficient, as it shows how much pollution is directly attributed to $1 spending in sector .
In the first layer of , (i.e. )one dollar of output relates only to the pollution directly emitted so that . In the second layer()thefirst level of the supply chain,, becomes involved. Thus , while is the third level of environmental impacts associated with $1 of output from the economy.
Summing all the layers can be written as
.
Due to the infinite geometric series (similar to a Taylors series) occurring inside the brackets this can be rewritten as
. (2)
Equation (2) gives the total pollution generated by one unit of final demand for products from sector . To find the total amount of environmental impact that households are responsible for, , one need only multiply the total environmental impacts of production, , by the amount of consumption in Final demand,.
(3)
References:
Baumann, A. (2013). Greenhouse gas emissions associated with different meat-free diets in Sweden. Uppsala University.
Berners-Lee, M., Hoolohan, C., Cammack, H., & Hewitt, C. N. (2012). The relative greenhouse gas impacts of realistic dietary choices. Energy Policy, 43, 184–190.
Duchin, F. (2005). Sustainable Consumption of Food: A Framework for Analyzing Scenarios about Changes in Diets. Journal of Industrial Ecology, 9(1-2), 99–114. Retrieved from
Foran, B., Lenzen, M., & Dey, C. (2005). Balancing act: A triple bottom line analysis of the Australian economy. Canberra: CSIRO and the University of Sydney.
Heller, M. C., Keoleian, G. A., & Willett, W. C. (2013). Toward a Life Cycle-Based, Diet-level Framework for Food Environmental Impact and Nutritional Quality Assessment: A Critical Review. Environmental Science & Technology, 47(22), 12632–12647. doi:10.1021/es4025113
Kitzes, J. (2013). An Introduction to Environmentally-Extended Input-Output Analysis. Resources, 2(4), 489–503.
Lenzen, M. (2000). Errors in Conventional and Input‐Output—based Life—Cycle Inventories. Journal of Industrial Ecology, 4(4), 127–148.
Lenzen, M., Moran, D., Kanemoto, K., & Geschke, A. (2013). Building Eora: A global multi-region input-output database at high country and sector resolution. Economic Systems Research, 25(1).
Majeau-Bettez, G., Strømman, A. H., & Hertwich, E. G. (2011). Evaluation of process-and input–output-based life cycle inventory data with regard to truncation and aggregation issues. Environmental Science & Technology, 45(23), 10170–10177.
Miller, R. E., & Blair, P. D. (2009). Input-output analysis: foundations and extensions (2nd ed.). Cambridge University Press.
Murray, J., & Wood, R. (2010). The Sustainability Practitioner’s Guide to Input-Output Analysis. Champaign, Illinois: Common Ground Publishing LLC.
Roy, P., Nei, D., Orikasa, T., Xu, Q., Okadome, H., Nakamura, N., & Shiina, T. (2009). A review of life cycle assessment (LCA) on some food products. Journal of Food Engineering, 90(1), 1–10.
Wiedmann, T. (2013). Environmental Input-Output Analysis - Module 1 . Kitakyushu, Japan: The International School of Input-Output Analysis, 21st International Input-Output Conference.