Handbook of Input-Output Economics in Industrial Ecology

To some, industrial ecology is the field that seeks to understand and replicate the dense network of by-product exchanges found in the famous industrial district of Kalundborg, Denmark. To others, it is the attempt to look to natural systems for models for industrial design and practice. To still others, it is nearly any effort to mesh environmental concerns with production and consumption.

A handbook on input-output analysis needs more clarity than this, both to provide context for the individual chapters and to provide an introduction to those less familiar with industrial ecology. This opening chapter will provide such an introduction by first reviewing the goals, history, elements and state of development of the field. It will then examine six dimensions of industrial ecology in terms of their potential relationship to input-output analysis.

Two fields of scientific inquiry can be interconnected effectively only through a clear conceptual overlap. Moreover, the overlapping (that is, the common) concepts must have proven their internal operational effectiveness separately in each one of the adjoining disciplines (Leontief 1959, 1985).

It has been only a few hundred years since human society escaped from a constant cycle of ebb and flow of population changes. Famines and epidemics were mingled with preindustrial European and Asian history, repeatedly setting the human population of the region several decades to hundreds years back (Braudel 1979). It was industrialization, together with the green revolution, that enabled humans to manipulate the untamed nature, setting the humankind free from the famines and epidemics that kept its population at a much lower level throughout its history. The burst of human population in recent centuries, however, is not only a consequence of industrialization but in a sense also a cause of industrialization. Fulfilling the needs by the unprecedented number of people required intensification and efficiency in industrial and agricultural production, which in turn helped generate more economic surplus enabling consuming even more. The human kind seemed to have won an autonomy, of which the prosperity somehow self-catalyzes and works independently from the means that the nature provides.

Ironically, however, humans became more dependent upon the natural environment both as a source of natural resources and as a sink of wastes and pollution. Despite the remarkable technological developments, population growth and improvements in welfare demanded an unprecedented amount of natural resources withdrawal from and wastes and pollutants disposal to the nature. Global crude oil extraction, iron ore mining, and underground water withdrawal, to name a few, are at their highest to satisfy the needs of the ever wealthier and populous human-kind. Around 26 billion barrels of crude oil are extracted every year (EIA 2008), enough to fill over five Olympic-size stadiums every day (Suh 2004a). Per capita copper use until 1900 is estimated to be below 1 kg/year, which has become around 15 kg/year by 2000 (Gordon et al. 2006). The U.S. total materials use is estimated to be less 200 million metric tons at the dawn of the twentieth century, and it reached nearly 3,000 million metric tons by 1995 (Gardner and Sampat 1998).

Increasing empirical evidence suggests that current levels of anthropogenic environmental pressures on the world-wide level do not comply with requirements of environmental sustainability (for example, WWF et al. 2004). Especially industrialized countries, responsible for the largest share of pressures on the global environment, are demanded to significantly reduce the material and energy resources used for production and consumption and to achieve de-linking (or de-coupling) of economic growth from environmental degradation. The concept of de-linking was adopted by a large number of national, European and international environmental policies (for example, European Commission 2003; OECD 2004). While de-linking in relative terms decreases the resource intensity of economic processes, absolute de-linking is required from a sustainability point of view, in order to keep economic and social systems within the limits of the ecosphere (Hinterberger et al. 1997).

1

Monitoring the transition of societies towards de-linking targets requires comprehensive and consistent information on the relations between socio-economic activities and resulting environmental consequences. In the past 15 years, a large number of approaches were developed providing this information in biophysical terms.

2

These methods proved to be appropriate tools to quantify “ societal metabolism ” (Fischer-Kowalski 1998) and to measure the use of “ environmental space ” (Opschoor 1995) by human activities.

Manufactured capital stocks and their relationships to physical flows of materials and energy are of interest in the fields of industrial ecology and input-output analysis. Manufactured capital stocks embody technologies, which may be characterised by input-output (IO) relations. The rate and nature of technological and structural change in an economy are therefore related to the dynamics of these stocks. Certain capital stocks also act as substantial long-lived stores of materials in the anthroposphere. Additions to and scrapping of these stocks directly generate flows of new and used materials and wastes. This chapter is concerned with two relationships between manufactured capital stocks and material flows, and in particular, how they may be modelled in the field of industrial ecology. Examples are drawn from scenarios developed using the Australia Stocks and Flows Framework (ASFF) (Foran and Poldy 2002).

Section two of this chapter deals with methodological and practical issues encountered in accounting for and modelling manufactured capital stocks. Both commonalities and differences between economic and physical perspectives on capital stocks are discussed. An example is given of historical and projected vehicle stocks in Australia. Section three deals with input-output modelling of technologies embodied in capital stocks, focussing particularly on the ‘bottom-up’ or ‘process modelling’ approach employed in ASFF. An example of process-based IO models for steel production in Australia is provided. Section four is concerned with dynamic models of stocks and flows in Industrial Ecology. A dynamic physical IO model (Lennox et al. 2004) within ASFF is described and an example of material flows associated with electricity generation capacity is given. Section 5 concludes the chapter, providing a brief discussion of key issues in modelling capital stocks in terms of material stores and/or embodied technologies within the field of industrial ecology.

Input-Output Tables (IOT) have long attracted interest from researchers in Industrial Ecology (IE) for two reasons. First, they provide an easily accessible data base to analyze networks of economic processes or activities on national scale. This data base is used to allocate environmental impacts and physical flows to economic activities in Life Cycle Assessment (LCA) (Hendrickson et al. 1998; Matthews and Small 2001; Suh and Huppes 2002, 2005; Suh 2004a, b) and Material Flow Accounting (MA Accounting) (see contributions of Giljum and Nathani in this handbook, also Meyer and Bockermann 1998; Bringezu 2002; Daniels and Moore 2002). Second, IOT is integrated into the general framework of national accounting and therefore provide an interface between MF Accounting on national scale and economic indicators such as GDP. This interface is used to set up Physical Input-Output Tables (PIOT) following the system definitions and partly the procedures of data compilation commonly used in IOT (Weisz and Duchin 2006; Gravgård 1999; Stahmer et al. 1998). But, also more general attempts to integrate physical flow and stock accounting into national accounting rely on system definitions commonly used in IOT to ensure compatibility within the overall accounting scheme (Weisz et al. 2005; Voet et al. 2005; Spangenberg et al. 1999). This is of paramount importance for using mixed indicators from physical and economic accounting such as DMC per GDP to analyze processes of dematerialization.

For numerous research questions, however, an analytic approach based on the system definition given by national IOT is not appropriate. IOT on national scale only provide highly aggregated data, use a given industry classification and measure commodity flows in monetary units. Household activities and capital formation are only partly covered. Such a system definition is insufficient to compare environmental performances of different products (e.g. cars), to evaluate waste management strategies for specific materials (e.g. electronic waste) or identify pathways of substances (e.g. chlorine). The system definition is a key step in any LCA or Material Flow Analysis (MFA) and researchers in IE are naturally reluctant to start with a “pre-defined” system which might not cover all important aspects of the problem under study.

The enormous increase in interest — in the last 2 decades — for environmental issues has led to a markedly upsurge in the collection of data. One of the new types of data sources that have become available is the physical input-output table (PIOT). The production sector in an economy distinguishes industries and the intermediate flows between the industries are measured in the same physical unit, such as billion tons (bt). This is in contrast to the usual monetary input-output tables (MIOTs) that measure the intermediate deliveries in money terms, such as billion dollars. Examples of published PIOTs can be found in Kratterl and Kratena (1990), Kratena et al. (1992), Konijn et al. (1997), Stahmer et al. (1997), Pedersen (1999), Nebbia (2000), Mäenpää (2002), and Hoekstra (2003).

1

On the one hand, PIOTs can be regarded as a natural extension of the so-called hybrid input-output tables, as far as their numerical implementation is concerned.

2

Monitoring the transition of modern societies towards a path of sustainable development requires comprehensive and consistent information on the relations between socio-economic activities and resulting environmental consequences. In the past 15 years, several approaches have been developed providing this information in biophysical terms (see, for example, Daniels and Moore 2002 for an overview). These methods of physical accounting are applied to quantify “ societal metabolism ” (Fischer-Kowalski 1998) and to measure the use of “ environmental space ” (Op-schoor 1995) by human activities. Within the system of physical accounts on the national level (for a classification see United Nations 2003), material flow accounting and analysis (MFA) and land use accounting are regarded as appropriate tools to provide a comprehensive picture of environmental pressures induced by and interlinked with the production and consumption of a country.

In the European Union (EU), a large number of policy documents address high levels of resource use and production of huge amounts of waste and emissions as one major obstacle for the realisation of an environmentally sustainable development in industrialised countries. The sustainable management of natural resources, in order to keep anthropogenic environmental pressures within the limits of Earth 's carrying capacity, is highlighted as one central objective for the implementation of a sustainability strategy within Europe (for example, European Commission 2001a). De-coupling (or de-linking) economic growth from the use of natural resources and environmental degradation is defined as the core strategy to achieve sustainable levels of resource use; raising the resource productivity of production and consumption activities should help producing the same or even more products with less resource input and less waste (European Commission 2003).

In an economic system that aims at sustainable development, material indicators become increasingly more important than monetary indicators, as much of the literature now testifies (Ayres and Ayres 1998). Monetary indicators are often not able to reveal all the implications and interactions between the biosphere and technosphere (Nebbia 2000; De Marco et al. 2001).

The knowledge of these indicators is an essential requisite to evaluate the environmental impacts caused by human activities. The scarcity of information on the amount and the quality of waste flows, from the economic system to the biosphere, makes the evaluation of environmental impacts and the choice of an adequate disposal system both very difficult (Ayres and Ayres 1997; Nakamura and Kondo 2002).

One of the fundamental goals of industrial ecology is to change and reduce material and energy flows related to satisfying the needs of human society, so that volume and quality of these flows can be carried by the natural environment without severe disturbances. Among the many redesign strategies proposed in industrial ecology, the strategy of material efficiency improvement mainly targets bulk material flows and aims at producing and using material goods in a more efficient way. Understood in a wider sense, material efficiency improvement also covers material and product substitution which reduces the environmental burden of material goods. As common in the field of industrial ecology, a life cycle view should be chosen to evaluate the efficiency criterion. Thus a higher material efficiency of an alternative design option should be proven over the complete life cycle of a product.

The growing concern of European citizens with environmental quality and the European Commission's determination to develop stronger environmental policies has contributed to the development and optimization of environmental management tools to support decision-makers in industry and government. These tools help to pro-actively identify sustainable options, optimized according to environmental, social, and economic criteria.

In line with recent European Commission initiatives, an Integrated Product Policy (IPP) approach is to be considered in any economic sector. IPP addresses the whole life cycle of a product, and seeks to avoid shifting environmental problems from one phase of the product life cycle to another.

As an emerging science, industrial ecology needs to identify and develop appropriate quantitative methods (Koenig and Cantlon 1998, 2000; Seager and Theis 2002). One of these primary tools has been Life-Cycle Assessment (LCA). LCA is used for assessing the impacts of products, processes, services, or projects on the environment (Graedel and Allenby 2003). The expression life-cycle indicates a “ cradle-to-grave” approach, beginning with a product 's conception and continuing through to its ultimate recycling or disposal. Thus, a product 's or process' lifetime includes (1) a raw materials acquisition phase, (2) a manufacturing, processing and formulation (3) a distribution and transportation phase (4) a use/re-use/maintenance phase (5) a recycling phase (6) and waste management (end-of-life) phase. LCA traditionally consists of four stages, (1) goal and scope (2) inventory analysis, (3) impact assessment, and (4) improvement analysis. In particular, Life Cycle Inventory (LCI) analysis describes those resources required and pollutants produced over the product 's lifetime (Fava et al. 1991). Major benefits of LCA include: a systematic method to evaluate the overall material and energy efficiency of a system; the ability to identify pollution shifts between operations or media as well as other trade-offs in materials, energy, and releases; and a means to benchmark and measure true system improvements and reductions in releases (Owens 1997).

The International Organization for Standardization (ISO) recommends the use of life-cycle assessment (LCA) to better comprehend and reduce environmental impacts related to manufactured products and services offered to our society. The principles of LCA are presented in the international standard ISO 14040; however, the implementation of the standard is not simple, and a couple of studies have addressed the existing limitations (Khan et al. 2002; Ross and Evans 2002).

One fundamental question is how to characterize a given environmental insult and how to select an appropriate metric to evaluate and minimize their impacts. This problem stem from the multiplicity of environmental insults caused by human activities, which are difficult to compare using a single approach. Moreover, most environmental problems have an intrinsic temporal dimension since environmental impacts persist in the environment for years and in some cases for generations. This yields sustainability concerns, which demand frameworks that allow the comparison of outcomes over time.

Life Cycle Inventory analysis (LCI) is defined as a phase of Life Cycle Assessment (LCA) involving the compilation and quantification of inputs and outputs for a given product system throughout its life cycle (ISO 14040 1998a). The concept of LCI has been adopted for cleaner production as early as the 1960s, and has had broad industrial and academic application in the last decades (Vigon et al. 1993). Com pared to the other phases of LCA, LCI has been considered a rather straightforward procedure except for several issues such as allocation (see e.g. Fava et al. 1991). Reflecting this belief, the method used for LCI compilation has rarely been questioned, although a large number of software, LCI databases and case studies have been released so far. However, contrary to the common belief, different methods have been available for LCI, and they often generate significantly different results. Therefore, it is necessary to assess advantages and limitations of different LCI methods and properly select suitable one(s) for each specific application. It is the aim of this paper to review and compare available methods for LCI compilation, and guide LCA users to properly select the most relevant methods for their analyses in relation to the goal and scope of the study as well as the resources and time available. With adaptations, the results are applicable outside the realm of LCA as well.

This paper is organized as follows: first available methods of LCI compilation are presented. Two computational approaches, process flow diagram and matrix inversion, are assessed, and then methods that utilize economic Input-Output Analysis (IOA) are described with special attention to hybrid analyses. Secondly, these methods are summarized and compared in terms of data requirements, uncertainty of source data, upstream system boundary, technological system boundary, geographical system boundary, available analytical tools, time and labor intensity, simplicity of application, required computational tools and available software tools. Finally, conclusions are drawn, and compliance of these methods to ISO standards and future outlooks are discussed.

The environmental impact of a person, whether measured in terms of average energy consumption, specific CO

2

emissions, or a person's occupation of ecological space, has received sustained interest at least since the 1970s. The need for quantifying drivers, key impact segments and leverage points from a consumption perspective lead to the formulation of various indicator concepts and analysis techniques, amongst which are scenario analysis, the ecological footprint, and structural decomposition analysis. Since

sustainable consumption

has become a key interest of environmental policy makers, not at least through the 2002 World Summit on Sustainable Development, there is an increased interest in such investigations.

Rather than providing a broad literature review of the issue, the purpose of this paper is to concurrently demonstrate example applications of the three input-output based methods mentioned above — scenario analysis, the ecological footprint, and structural decomposition analysis — and by so doing, provide a means for comparison and critique. Whilst all three analysis techniques are concerned with the overall theme of the study of sustainable consumption, they each provide a different focus for assessment and interpretation, and are thus suited to diverging purposes and research questions.

Worldwide ecosystems are under pressure of economic activities. The main impetus for this is human demand for food, other goods and services. How household spend their money is an important factor in the magnitude of the damage inflicted on the environment. The distribution of environmental damage among the different household expenditures can provide insight in how this damage can be reduced.

On the issue of consumption related environmental impacts many studies have been performed. However, most of these studies only focus on a specific part of our consumption (e.g. assessments [LCAs] of goods and services), on a specific impact category (e.g. energy or greenhouse gases) or on consumption on an aggregated level (e.g. footprint assessments of nations or cities). A comprehensive assessment covering the whole of consumption, while allowing detailed insight into its composition, and taking many environmental impacts into account, has not yet been performed.

In recent years a lot of new energy models have been developed to analyze climate change mitigation strategies and the effects such strategies have on economic and technological development. Two main types of models can be identified: Top-down models that focus on the economic interactions within different sectors in an economy and bottom-up models that focus more on physical energy flows and technological aspects. One way of exploiting the advantages of each of these approaches is to link them to create a hybrid approach. Due to their main characteristics (e.g. high degree of disaggregating) input-output models are suitable for integrating technological data in an economic model in a special manner. Therefore, a class of models exists, linking input-output models with disaggregated energy system models. In this paper we present such a hybrid approach which consists of the input-output model Macroeconomic Information System (MIS) and the bottom-up model Instruments for Greenhouse Gas Reduction Strategies (IKARUS) — Market Allocation (MARKAL). For the hybrid MIS/IKARUS-MARKAL model a soft-linking approach is used, where the MIS model supplies data about the development of the different industry and service sectors and the IKARUS-MARKAL model calculates the energy demand of these sectors and the cost-optimal energy production structure. Two different examples for the use of the hybrid MIS/IKARUS-MARKAL approach will be presented: our first example focuses on the question of whether a given emission target (like the Kyoto one) can be reached assuming a desired growth rate and taking technological restrictions into account. The focus of the second example is on the economic impacts of a CO

2

mitigation strategy. In this example, we ask which industries will benefit from the decision of policymakers to take measures to reduce CO

2

emissions and which ones will lose (in the sense of economic growth).

Economists and policy makers refer to carbon tax as an efficient instrument to control CO

2

emissions, but concerns about possible negative effects of its implementation, as for instance the loss of competitiveness on the international market, have been expressed.

In the present chapter the IO model is used to estimate the short-term effects of a carbon tax in Italy (the results can be easily extended to the case of a permission trading scheme), which include the percentage increase in prices and the increase in the imports of commodities to substitute domestically produced ones as intermediate input. The present study is not “behavioral”, in the sense that the change in the consumers' behavior and choice, induced by higher prices, is not taken into account.

National environmental policies, in general, are directed to environmental quality, emission reduction and spatial planning of nature areas, all within the borders of a country. Despite the growth in GDP in the Netherlands in the past decades, Dutch environmental policy has led to a substantial decrease in the emissions of several substances (RIVM 2004a). However, nowadays, policy makers realize that there are still several persistent global environmental issues, such as climate change, the loss of biodiversity and the depletion of natural resources that cannot be tackled on a national scale (VROM 2001). Attempts to deal with these issues require an understanding of the relations between economy and the environment, both within and between countries.

For policy makers to develop policies that have an effect on the cross-border environment, they have to be provided with relevant information on economic activities and their ecological effects. Data on environmental pressure across national borders can be assigned in several ways to both production and consumption activities in a country and international trade. It is up to scientists to use the necessary cross-sections from this data to provide policy makers with the right information. Input-output (IO) analysis in combination with industrial ecology offers tools and methodologies for presenting the same data in different ways directed at specific policy questions concerning national or cross-border issues.

As a supplement to the site, substance and media specific environmental policies, Denmark has had, since 1998, a product-oriented environmental policy (at the European level known as “Integrated Product Policy”). The policy has been organized as prioritized activities in selected sectors and/or product areas. This prioritization was informed by the results from the project “Environmental prioritization of industrial products” (Hansen 1995). Other previous studies with similar objectives, i.e. to identify the most important product groups from an environmental perspective, include Dall et al. (2002) for Denmark, Finnveden et al. (2001) for Sweden, Nijdam and Wilting (2003) for the Netherlands, Nemry et al. (2002) for Belgium, and Labouze et al. (2003) for the EU. The Swedish and Dutch study use the same general methodology as our study, namely environmentally extended IO-analysis (Miller and Blair 1985), while the remaining studies use a bottom-up process based analysis.

Due to the environmental indicators used (energy consumption and resource loss) the product groups that are ranked high by Hansen (1995) are those with either large energy consumption or which are destroyed or dissipated during use. This includes the main energy carriers, transport activities (represented by the vehicles including their use phases), fertilizers, animal feeds, meat and dairy products, building materials, detergents, newspaper, beer and furniture.

In this paper we show how IO equations for sector outputs and prices are used as part of a larger policy analysis modeling system for energy and climate-related studies. The IO framework is particularly useful because it can accommodate the analysis of both price and direct program expenditure impacts. We briefly discuss the advantages of including non-price programs in any serious climate policy or sustainable energy strategy. Further, we contend that the impacts on the economy from a set of price and program expenditure polices can be seen by comparing constructed IO tables for a future year, such as 2030, with and without these polices. We present the All Modular Industry Growth Assessment (AMIGA) modeling system which has the capability to forecast future IO table values.

Performing an energy and environmental analysis, researchers have to face many problems regarding the data quality and availability. Data are often out-of-date, not representative and consistent or, frequently, referred to faraway geographic and productive contexts. The Input-Output (IO) model, due to its simplicity, allows to acquire information regarding the energy and environmental performances of productive sectors.

Ecological resources constitute the basic support system for all activity on earth. These resources include products such as air, water, minerals and crude oil and services such as carbon sequestration and pollution dissipation (Tilman et al. 2002; Daily 1997; Costanza et al. 1997; Odum 1996). However, traditional methods in engineering and economics often fail to account for the contribution of ecosystems despite their obvious importance. The focus of these methods tends to be on short-term economic objectives, while long-term sustainability issues get shortchanged. Such ignorance of ecosystems is widely believed to be one of the primary causes behind a significant and alarming deterioration of global ecological resources (WRI 2000; WWF 2000; UNEP 2002).

To overcome the shortcomings of existing methods, and to make them ecologically more conscious, various techniques have been developed in recent years (Holliday et al. 2002). These techniques can be broadly divided into two categories, namely preference-based and biophysical methods. The

preference-based methods

use human valuation to account for ecosystem resources (AIChE 2004; Balmford et al. 2002; Bockstael et al. 2000; Costanza et al. 1997). These methods either use a single monetary unit to readily compare economic and ecological contributions, or use multi-criteria decision making to address trade-offs between indicators in completely different units. However, preference-based methods do not necessitate compliance with basic biophysical laws that all systems must satisfy, and require knowledge about the role of ecological products and services that is often inadequate or unavailable.

Not only activities of humans in households require energy, occurring in the form of natural gas, coal, petrol and electricity (direct energy requirement), but other consumption goods and services also require energy for their production, transport and trade (indirect energy requirement). In many cases a method is needed to determine the energy requirement associated with consumption patterns that is quick and fairly accurate with respect to the individual consumption categories. In other words, the method should be accurate enough to detect possible differences between the consumption categories (not between individual product variants or brands within a consumption category).

This chapter starts off by briefly discussing two existing methods for analysing the energy requirement of consumption categories, i.e. process analysis and input-output analysis. This is followed by a proposal for creating a tiered hybrid (see Suh et al. 2004) from these two methods to analyse the energy requirement for the various consumption categories. The hybrid method will be illustrated using the refrigerator as an example. The chapter ends with a discussion on the suitability of this method for calculating the total energy requirement of household consumption.

One of the main aims of this study is to explore the links between energy, economy and environment from different perspectives, but always with a policy-oriented focus. This will be done by implementing and developing an input-output model with satellite accounts, to analyze the links between the different economic sectors, energy production and use, and the ‘corresponding’ production of CO

2

emissions in Portugal.

For this, the paper is organized as follows. In Section “The Input-Output Framework”, there will be presented a brief outline of the basic input-output model, and then succinctly discussed the core aspects of its extensions for the consideration of environmental and energy issues. In Section “CO

2

Emissions by the Portuguese Economy”, there will be presented the data sets used for the Portuguese case, and then an extended input-output empirical application, from which is assessed the (sectoral and aggregate) production of CO

2

emissions (derived from fossil fuels use) by the Portuguese economy. There will be also offered a succinct reflection on the use of elasticities of CO

2

emissions with some of the model parameters, as measures of sensitivity analysis of the level of emissions. Accordingly, a summary of the key lessons learned and a discussion of their policy relevance will be offered in Section “Final Comments”.

In international climate change negotiations a country is commonly held responsible for all CO

2

emitted from its domestic territory. In the literature this commonly applied CO

2

accounting method is called “territorial” or “producer responsibility”. Driven by concerns about carbon leakage (Wyckoff and Roop 1994; Kondo et al. 1998; Ahmad and Wyckoff 2003) and equity associated with the structure of trade relations between developing and developed countries (Schaeffer and De Sá 1996; Machado et al. 2001) as well as import and export structures of small open economies (Munksgaard and Pedersen 2001), “consumer responsibility” has been proposed as an alternative CO

2

accounting method.

1

From an accounting perspective the difference between the two concepts lies in the treatment of trade related emissions. Besides its domestic emissions a country can either be held responsible for CO

2

embodied in exports or imports (or a combination of both). With world trade growing more than twice as fast as world GDP,

2

the way how to account for CO

2

emissions becomes increasingly relevant for countries in international climate change negotiations and for successful global mitigation efforts as the equity issue becomes more urgent and the threat of carbon leakage becomes more severe.

Any economic activity generates waste of some kind, which needs to be treated by some waste treatment method. Corresponding to any flow of goods among different sectors of the economy, there exists the associated flow of waste involving waste treatment sectors. The conventional IOA was originally developed to represent the intersectoral flow of goods and hence is not designed to take account of the flow of waste associated with it. Consequently, in its conventional form, IOA is not able to take proper account of the effects that result from the interaction between the flows of goods and wastes.

The pioneering study in the field of environmental IOA (EIO) that is relevant to waste management issues is the Leontief pollution abatement model (Leontief 1970, 1972). Leontief extended the conventional IOA to take account of the emission of pollutants, their elimination activity, and the interdependence between conventional goods-producing sectors and pollution abatement sectors. With regard to their relevance to issues of waste management, the Leontief pollution abatement model and its extension by Faye Duchin (1990) can be characterized by the fact that they assume the existence of a strict one-to-one correspondence between a pollutant (waste) and its abatement (waste treatment) method.

The quantity and quality of wastes modern societies experience today is unparalleled in history. Additionally, a new awareness of the pollution and human health hazards caused by the disposal of waste has emerged.

Waste prevention strategies and careful management of waste have significant scope for limiting the waste flows damage and conserving scarce resources. However, successful prosecution of such policies would be possible only if concerted efforts are made to address the traditional partitioned perception of waste issues. Industry and final consumers have to become aware of their contribution to the problems associated with total waste flows generation (directly and indirectly), as well as their major role in delivering the needed solutions. Addressing this ‘lack of perception’ would ultimately change people's attitudes towards waste management as a whole and increase their involvement in ‘sustainable integrated resource and waste management’ practices. The increasing amounts of solid waste being generated show the necessity, and at the same time offer the opportunity, to look for new approaches.

Japanese national energy consumption increased 3.4 times from FY1965 (1,085 peta cal) to FY2001 (3,676 peta cal). During the same period, household energy consumption increased by 4.9 times, from 107 to 522 peta cal (The Energy Data and Modeling Center (EDMC) 2003). One major cause of such a rapid increase in household consumption is electric appliances, that is household electricity consumption has increased 9.3 times from 24 to 227 peta cal; our convenient everyday life is based on increasing energy consumption, inevitably linked with ecological degradation. Therefore, when we talk about ecology, our lifestyle must also be reviewed in that context. Recently, UNEP (United Nations Environmental Programme) has proposed the concept of “Sustainable Consumption” besides “Sustainable Production.” (United Nations Environmental Programme (UNEP) 2002) It says that sustainable consumption “is the final step in progressive widening of the horizons of pollution prevention,” and that “action focused on consumption has highlighted the need to address the creation of new systems of production and consumption.” Supplier's efforts towards ecological products will be meaningless if people do not use them and stick to their consumption habits.

In our life, there are several ways to do the same thing, but one of them can be more ecological than the others. However, no method has yet been developed to enable quantitative evaluation of the ecological effect of household activities. We need a methodology to analyze consumers' behavior.

This year the international handbook on integrated Environmental and Economic Accounting (SEEA-2003) will be published. This handbook provides a detailed overview of environmental accounting approaches that have been developed in parallel with the System of National (economic) Accounts. In addition to natural resource stock accounts, and environmental protection expenditure accounts, SEEA-2003 pays considerable attention to physical flow accounting. Expanding the national economic accounts with physical data sets facilitates the joint analysis of environmental and economic policy issues. This article discusses the main characteristics of national accounts-oriented physical flow accounting approaches and provides an overview of the kind of indicators they may put forward. Also the analytical advantages of national accounts oriented physical flow accounts are illustrated. The article is not an attempt to provide a comprehensive review of macro-oriented physical flow accounting approaches. For such reviews in this Journal we would like to refer to Daniels (2002) and Daniels and Moore (2002).

This chapter reviews the history and discusses the role of the primary environmental data that have become the foundation of environmental input—output analyses in Japan. It also describes two practical approaches to estimating unit environmental burden: an exogenous estimate approach and an endogenous estimate approach. As a case study, the endogenous estimate approach is used to estimate sectoral unit carbon dioxide (CO

2

) emissions based on the Japanese Input—Output Tables for 2000. The technical problems that exist in determining CO

2

emission for each sector by that approach are explained. In addition, this chapter uses an input—output analysis to calculate the embodied CO

2

emission intensities of the approximately 400 sectors in the Japanese Input—Output Tables, and summarizes the quantitative characteristics of intensities for the major sectors. To examine the relationship between economic final demands and CO

2

emissions, the Japanese CO

2

emission structure in 2000 is illustrated using those intensities.

This paper elucidates the data sources and data preparation procedures used in developing the sectoral environmental data of the U.S. The database described in this paper interlinks (1) Input-Output Table (IOT), (2) environmental emission and resources use statistics, and (3) characterization factors from Life Cycle Impact Assessment (LCIA) that quantify environmental impacts. Each of these three modules was designed to describe (1) the economic process generates environmental interventions, (2) the quantity of the environmental intervention generated, and (3) the process that these environmental interventions realize environmental impacts, respectively. The resulting database encompasses 1,344 different types of environmental interventions generated by 480 commodities of the U.S. input-output table, linked to 86 commonly used LCIA models. This paper aims to share the experiences of and to elucidate the procedures and the data sources used for developing the sectoral environmental database in the U.S.

Ecologists have used input-output analysis since the early 1970s to study flows of materials and energy in complex networks. These ecological networks are very similar to material and energy flows in industrial systems, yet the input-output approach developed by ecologists has not been applied to industrial systems. In this paper, an overview of early work to adapt ecological input-output analysis to industrial systems is presented.

On a scientific as well as a political level, there is wide consensus today that the concept of sustainable development requires integrated approaches to illustrate the interactions between economic, social and environmental concerns. Input-output analysis is regarded as an appropriate framework to provide a comprehensive picture of these linkages as it allows combining bio-physical and social data with economic (monetary) input-output models.

The interrelations between the economic and ecological system affect the flow of material inputs and outputs in many forms. Environmental degradation depends considerably on input quantities, which are taken from and transferred again to the environment in form of emissions and wastes. For the description of these relationships, the concepts of “industrial metabolism” and “societal metabolism” are important. These terms refer to the exchange of materials and energy between ecological and socio-economic systems. According to these concepts physical indicators can be differentiated with respect to input and output indicators.

The System of National Accounts (United Nations 1968, 1993) is a useful tool for measuring not only the effects of technological changes, input composition changes, product-mix changes, and consumption shifts as fundamental structural elements (e.g., Afrasiabi and Casler 1991; Rose and Casler 1996; Dietzenbacher and Los 1998, 2000 for the exposition of structural change analyses) but also the efficiency indices consistent with neoclassical economic theory such as total factor productivity growth and labor productivity growth (e.g., ten Raa 1994,1995a; ten Raa and Mohnen 1994).

1

The key to the system's utility rests on the structural elements provided.

In analyzing environmental consequences of productive systems, the emphasis has been generally on microscopic level, where process-based Life Cycle Assessment (LCA) may be a representative tool (Guinée et al. 2002). The analysis measures physical energy and material requirements for the processes of the product system in question

inside

a well-defined system boundary and provides very detailed inventories of environmental emissions and resources use. Hybrid Life Cycle Assessment model, where the standard input-output model and the process-LCA model are combined, additionally enables us to cover the indirect intermediate inputs

outside

the process-based system boundary (e.g., Moriguchi et al. 1993; Suh 2004; Suh et al. 2004). In this case, the concept of a sector outside the system boundary no longer relates to a process but to a standard commodity or an industry.

Input-output analysis of inter-industry exchange has proved to be useful in LCA. Input-output has a long history in economics. Less known, is that input-output in fluenced

linear programming

(LP) in its early development. In fact, Input-output models can be regarded as special cases of

linear programming

problems. Linear programming is the most useful practical tool in helping us to make the best use of scarce resources when faced with complex decision problems. Firms routinely use linear programming and other optimization techniques in planning their activi ties, for example in logistics of supply chains, production scheduling, and resource allocation in general. In this chapter, we show the historical relationship of input-output analysis and linear programming. Next, we show that an input-output model is a special case of an LP formulation. Through a series of examples, we show how an LP formulation more generally can tackle situations with multifunctional units and multiple technologies. Furthermore, we show how a detailed LP model of an industry sector can be linked with an aggregated IO model. The last section of this chapter provides a brief survey of applications where LCA and input-output analysis has been integrated with optimization models.

Industrial Ecology as coined by Frosch and Gallopoulos (1989)

1

has proven to be one operational and holistic concept for successfully implementing more sustain able policies. However, like many other concepts that have become popular in the post-Brundtland era during the late 1980s and early 1990s, such as Cleaner Pro duction (Baas et al. 1990), Ecological Modernisation (Jänicke 1988) and Industrial Metabolism (Ayres 1989), it has been open to criticism, due to the failure of en vironmental policies to achieve many of their ambitious goals set out during the Rio process. The shared pathology has usually been the technocratic approach and supply-side bias, as most clearly laid out in the

sustainable consumption

debate (UNEP 2002; Princen et al. 2002).

2

Researchers have responded to this criticism by adjusting their policy approaches. Much more emphasise has recently been given to the study of household behaviour and demand side issues (e.g. Gatersleben 2000; Jackson 2004); socio-institutional and demographic concerns have been integrated with environmental-economic ones (e.g. Cogoy 1995; Madlener and Stagl 2001); and more and more effort has been devoted to understanding and disclosing the com plex relationship between consumption activities and well-being (Hofstetter and Madjar 2003; Jackson et al. 2004).

Consumption causes environmental impacts in two different ways. Direct environmental impacts result from consumption when consumers directly burn fossil fuels; for instance, from the petrol used for personal transportation or wood used for space heating. Significant environmental impacts also occur indirectly in the production of consumable goods. When production occurs in the same country as consumption, then government policy can be used to regulate environmental impacts. However, increasing competition from imported products has led to a large share of production occurring in a different country to consumption. Regulating the resulting pollution embodied in trade is becoming critical to stem global pollution levels. Due to increased globalization of production networks, there is increasing interest in the effects of trade on the environment (Jayadevappa and Chhatre 2000; Copeland and Taylor 2003).

With the increased interest in trade and the environment research activity is focusing on methods of accurately calculating the pollution embodied in traded products. Early studies in this area assumed that imports were produced with the same technology as the domestic economy (e.g. Wyckoff and Roop 1994; Lenzen 1998; Kondo et al. 1998; Battjes et al. 1998; Machado et al. 2001), however, using this assumption large errors may result when the countries have diverging technology and energy mixes (Lenzen et al. 2004; Peters and Hertwich 2006a, c).

1

This stimulated research in the use of multi-regional input-output (MRIO) models.