an aproach to scenario analysis

Upload: anca-butnariu

Post on 07-Apr-2018

231 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/4/2019 An Aproach to Scenario Analysis

    1/13

    An approach to scenario analysis of the sustainability of an industrialsector applied to clothing and textiles in the UK

    J.M. Allwood*, S.E. Laursen, S.N. Russell, C. Malvido de Rodrguez, N.M.P. Bocken

    Department of Engineering, Institute for Manufacturing, University of Cambridge, Mill Lane, Cambridge CB2 1RX, UK

    Received 31 January 2007; received in revised form 13 May 2007; accepted 5 June 2007

    Available online 2 August 2007

    Abstract

    Companies aiming to be sustainability leaders in their sector and governments wanting to support their ambitions need a means to assess the

    changes required to make a significant difference in the impact of their whole sector. Previous work on scenario analysis/scenario planning

    demonstrates extensive developments and applications, but as yet few attempts to integrate the triple bottom line concerns of sustainability

    into scenario planning exercises. This paper, therefore, presents a methodology for scenario analysis of large change to an entire sector. The

    approach includes calculation of a triple bottom line graphic equaliser to allow exploration and evaluation of the trade-offs between economic,

    environmental and social impacts. The methodology is applied to the UKs clothing and textiles sector, and results from the study of the sector

    are summarised. In reflecting on the specific study, some suggestions are made about future application of a similar methodology, including

    a template of candidate solutions that may lead to significant reduction in impacts.

    2007 Elsevier Ltd. All rights reserved.

    Keywords: Scenario analysis; Sustainability; Industrial sector; Clothing and textiles

    1. Introduction

    Work to date in the broad area of industrial ecology has

    a strong focus on measurement, and if practical change is

    discussed, it is usually expressed through the instruments of

    policy and economics or in the language of sociology. This

    approach implicitly assumes that a catalogue of potential prac-

    tical changes to production, product design or product usage

    exists, so that the requirements for change are related to selec-

    tion and motivation. However, there appears to be a shortageof work in specifying such a catalogue of the potential changes

    that should be considered.

    Such a casual attitude to practical change would be justified

    if sufficient changes had already been identified and were

    ready for implementation, but this is far from the case. For

    example, within the area of carbon emissions reduction, the

    UKs Carbon Trust is committed to the UK governments tar-

    get of a 60% reduction in emissions from 1990 levels by 2050

    and has explored the range of changes that might lead to this.

    Development of alternative energy supplies is anticipated to

    contribute w16%, and the use of hydrogen based fuel cells

    may extend this. However, the bulk of the reduction (more

    than 60%), which is expected to come from energy efficiency

    measures, is not linked to specific technologies. The Carbon

    Trusts publicity material gives two examples: low energy

    light bulbs and timed lighting switches, but these will makea negligible contribution. Their 2005 abatement curve [1]

    which reports on potential carbon savings for the UK industry

    sector, can identify only around 25% of existing carbon emis-

    sions that can be saved through known technologies. While

    motivation and implementation of change are important

    aspects of developing a sustainable future, it appears that the

    identification of practical solutions that will make a significant

    change is also extremely important, but often overlooked.

    In prior work, a comprehensive survey of known practical

    changes that might lead to increased sustainability has been* Corresponding author. Tel.: 44 1223 338181; fax: 44 1223 338076.

    E-mail address: [email protected](J.M. Allwood).

    0959-6526/$ - see front matter 2007 Elsevier Ltd. All rights reserved.

    doi:10.1016/j.jclepro.2007.06.014

    Available online at www.sciencedirect.com

    Journal of Cleaner Production 16 (2008) 1234e1246www.elsevier.com/locate/jclepro

    mailto:[email protected]://www.elsevier.com/locate/jcleprohttp://www.elsevier.com/locate/jclepromailto:[email protected]
  • 8/4/2019 An Aproach to Scenario Analysis

    2/13

    presented by Russell and Allwood [2], with a broader descrip-

    tion of the context of such change in Russell [3]. The survey

    attempted to provide a categorisation of practical changes to

    product design, production processes and supply chain opera-

    tion that have been implemented to date. However, being a sur-

    vey, this work considered only what has been done previously.

    Looking forwards, what are the really important changes thatshould be considered and may not yet have been attempted?

    How will practical changes that lead to substantial reductions

    in undesirable impacts be identified?

    One approach to answering these questions is to use struc-

    tured scenario analyses to consider the consequences of wide-

    scale change to existing systems, and this approach is gaining

    increasing attention in the pursuit of sustainability solutions.

    This paper presents a methodology to apply scenario analysis

    to examine the future sustainability of a whole industrial

    sector and demonstrates its application in a five person-year

    study of the future supply of clothing and textiles products

    to meet UK demand [4].

    Section 2 of the paper reviews previous work on scenarioanalysis, particularly as applied to sustainability. In Section

    3, a methodology for applying scenario analysis to a whole

    sector is developed, and its specific application to the UK

    clothing and textiles sector is described. The results of that

    study are reviewed in Section 4, and in conclusion, Section

    5 reflects on the approach in the hope that it may be beneficial

    for future studies of other sectors. A brief introduction to the

    clothing and textiles sector is given in the Appendix.

    2. Scenario analysis and its application to sustainability

    A range of future techniques are used by organisationsand policy makers to gain insights into what the future may

    look like, thereby laying the foundation for informed

    decision-making. Pesonen et al. [5] provide a glossary of

    definitions of such futures research methods which include

    forecasting and scenario analysis. One of the major flaws in

    analytical techniques such as forecasting is that patterns

    extrapolated from historical events are imposed with the im-

    plicit assumption that the world will remain relatively stable,

    and the future is predicted based on events of the past. Bood

    and Postma [6] suggest that the rise of (multiple) scenario

    analysis has occurred due to the failure of traditional forecast-

    ing techniques to provide credible results. A survey of indus-

    trial corporations in the USA in the period 1977e1981 showed

    that less than a third expressed satisfaction with traditional

    forecasting techniques [7]. In fact, according to Wack [8], dis-

    satisfaction with formal planning and forecasting techniques

    led to widespread development and use of scenario analysis,

    also termed scenario planning [9e11]. Schoemaker [10] gives

    a comparison between scenario planning and other traditional

    planning techniques.

    The development of scenario analysis has been influenced

    by a number of companies and institutes including the

    RAND Corporation, Stanford Research Institute, Shell and

    others [12]. Shell is credited with the introduction of scenario

    analysis in the private sector, where it was developed for

    strategic purposes and has been in use since the 1960s [13].

    The development was driven by Pierre Wack and Edward

    Newland for strategic decision-making purposes.

    2.1. Definitions and classifications of scenarios

    There is no universal definition of scenarios. Kahn andWiener [14], who explored the possible consequences of

    nuclear proliferation in World War II, define a scenario as

    a hypothetical sequence of events constructed for the purpose

    of focusing attention on causal processes and decision points

    [15]. Godet and Roubelat [16] define a scenario as a descrip-

    tion of a future situation and the course of events which allows

    one to move forward from the original situation to the future

    situation. Van der Heijden [17] describes scenarios as tools

    to research ones understanding of the world. In line with the

    critical realism paradigm the objective is to challenge ones

    own mental model of the future. By stretching these vari-

    ables to their limits of credibility, one tries to create a number

    of possible futures which, while plausible, are significantlydifferent from business as usual. Many authors emphasise

    that scenarios do not predict, rather they allow us to examine

    what might happen [18]. Rienstra [19] outlines the three basic

    elements of scenarios as a description of: (1) the present situ-

    ation, (2) a number of future situations and (3) a number of

    events that may connect the present situation with the future

    one.

    Godet [20], Godet and Rouebelat [16] and van Veen-Croot

    et al. [21] distinguish exploratory and normative or anticipa-

    tory scenarios. Exploratory scenarios developed from descrip-

    tion of the past extrapolated by present trends indicate which

    scenarios might happen and then proceed to describe the pos-sible future outcomes. These types of scenarios are used to

    stimulate thinking about the possible futures and aim to exam-

    ine what can happen. For normative or anticipatory scenarios

    the starting point is the desired future situation and these types

    of scenarios explore how a certain target can be reached after

    which targets are set and paths that will lead to the stated

    future are described. Notten et al. [22] distinguish between

    normative and descriptive scenarios. Normative scenarios

    (also referred to as prospective, strategy, policy or intervention

    scenarios) describe probable or preferable futures while

    descriptive scenarios (also referred to as baseline, reference

    and non-intervention scenarios) explore possible futures.

    Fukushima and Hirao [23] and Notten et al. [22] distinguish

    back-casting and forecasting scenarios depending on the

    vantage point from which the scenario is developed. Back-

    casting scenarios reason from a specific future situation and

    then explore the paths needed to be taken to move towards

    that point and arrive at desirable future situations, while fore-

    casting scenarios take the present as their starting point and

    project todays problems, trends and realistic solutions onto

    the future. Leemhuis [24] classifies scenarios based on the

    time horizons used for the planning process: business cycle

    scenarios for shorter periods of up to five years, archetype sce-

    narios for a horizon of 10e15 years and exploratory scenarios

    for very long-term periods. Ringland [13] classifies scenarios

    1235J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    3/13

    as external, internal or system scenarios based on whether or

    not the influences are under the control of the organisation

    undertaking the study. External scenarios are those that exclu-

    sively consist of influences outside the control of the organisa-

    tion, internal scenarios take only factors under the control of

    the organisation into consideration and system scenarios which

    are mixed forms of external and internal scenarios containingexternal environmental influences as well as internal guidance

    dimensions. Scenarios also differ according to their subject

    of study and are classified as issue-based, area-based and

    institution-based [22]. Issue-based scenarios take societal issues

    as the subject of study; area-based scenarios explore a particular

    geographical area such as a country, region or city; institution-

    based scenarios (also subdivided into so-called macro, global,

    archetypal, framework, external or contextual scenarios on

    one hand and focused, decision, internal or transactional on

    the other) address the spheres of interest of an organisation,

    group of organisations or sector.

    Scenarios may also be classified based on the nature of the

    data as qualitative or quantitative scenarios. Van Veen-Crootet al. [21] term the latter objective methods and the former

    subjective. Notten et al. [22] further classify scenarios based

    on the range of possible futures taken into account as alterna-

    tive or business as usual. They explain that alternative scenar-

    ios describe futures that differ significantly from one another

    and often developed in an effort to raise awareness and under-

    standing about new or uncertain issues, while business as usual

    scenarios adhere to the status-quo or to present trends and their

    extrapolation into the future, where the aim is to fine-tune

    strategy rather than develop new strategy, for example.

    Scenario analysis is claimed by various authors to fulfil

    a wide range of functions. The functions given below arefrom Rienstra et al. [25].

    The signalling function e scenarios provide better insight

    into uncertain situations

    The communication and learning function e scenarios

    stimulate thinking about alternative futures

    The exploring and explaining function e scenarios show

    how solutions for specific problems may become reality,

    given certain policy priorities; they also present possible

    solution strategies

    The demonstration function e scenarios show the conse-

    quences of specific decisions

    The decision support function

    Additional information on other quite similar functions of

    scenarios is given by Bood and Postma [6].

    2.2. Scenario analysis in sustainable development and

    LCA studies

    Interest in using scenario analysis as a means of visualising

    plausible future paths for sustainable development has been

    growing recently. The paradigm of sustainable development

    inherently embraces futures research and thinking as the defi-

    nitions and goals refer to both present and future generations

    and according to Kelly et al. [26], sustainable development

    is generally motivated by a real concern for the long-term

    well being of humanity. Various international organisations

    have developed global-level scenarios to indicate the implica-

    tions of future global actions. For example, UNEP [27] in

    global environment outlook (GEO-3) has developed four sce-

    narios: the Market First scenario envisages a world in whichmarket-driven developments converge on the various expecta-

    tions that prevail in industrialised countries; the Policy First

    scenario in which strong actions are undertaken by govern-

    ment in an attempt to reach specific social and environmental

    goals; the Security First scenario assumes a world of great dis-

    parities, where irregularity and conflict prevail, brought about

    by socio-economic and environmental stresses; the Sustain-

    ability First scenario pictures a world in which a new develop-

    mental paradigm emerges in response to the challenge of

    sustainability. The Global Scenario Group [28] has developed

    three global scenarios: Conventional Worlds in which the

    global system evolves without major surprises; Barbarization

    scenarios which envision the grim possibility that social, eco-nomic and moral underpinnings of civilisation deteriorate;

    Great Transitions scenarios which explore visionary solutions

    to the sustainability challenge. Other global scenarios include

    the IPCCs emissions scenarios [29].

    Quist and Vergragt [30] give an extensive review of work

    using back-casting to anticipate more sustainable futures.

    Their emphasis is on a particular development from work

    largely carried out by analysts to participatory back-casting

    with extensive (and apparently in some cases, virtually exclu-

    sive) stakeholder dialogue used to consider simultaneously the

    definition of desirable futures and the means to attain them.

    They use the SusHouse project [31] as a particular exampleof the back-casting technique, aiming to consider scenarios in

    which future consumption is reduced and which includes

    a study of sustainable futures for clothing care. Such a study

    can clearly be undertaken effectively with a process based

    on stakeholder participation e but is primarily based on opin-

    ions rather than numerical analysis or modelling e for in-

    stance, the economic analysis in the project was achieved by

    means of a questionnaire. Such stakeholder led back-casting

    based on opinion gathering might be less effective in consid-

    ering more broad structural changes to a sector, such as the

    location of production, selection of materials, or changes at

    a macro-economic level.

    Scenario analysis is increasingly being incorporated into

    life cycle assessment (LCA) as a means of analysing possible

    future outcomes on the environment. For example, Ubbels

    et al. [32] used four globalisation scenarios to analyse the

    development of the international transport sector, Tan and

    Khoo [33] analysed the environmental performance of a pri-

    mary aluminium supply chain and Sonesson and Berlin [34]

    analysed the environmental impact of future supply chains

    for dairy products.

    Even though scenarios are in some sense an integral part

    of LCA studies, they were not always dealt with explicitly and

    as such the SETAC-Europe LCA Working Group Scenario

    Development in LCA started work in 1998, with the goal of

    1236 J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    4/13

    focusing on the use of scenarios in life cycle assessment. The

    group defines a scenario in LCA as a description of a possible

    future situation relevant for specific LCA applications based

    on specific assumptions about the future, and (when relevant)

    also including the presentation of the development from the

    present to the future [5]. The two general classifications of

    scenarios in life cycle assessment studies used by the groupare what-if scenarios and cornerstone scenarios: what-if sce-

    narios, the more widely used of the two, are generally used

    to compare two or more options in a situation familiar to the

    researcher, whereby a hypothesis can be defined on the basis

    of existing data. They are often studies where some specific

    changes within the present system are tested and their environ-

    mental impacts studied. The results of an LCA study using

    what-if scenarios are typically quantitative comparisons of

    the selected options, offering operational information in

    short- or medium-term decision-making situations. Cornerstone

    scenarios do not necessarily give quantified results comparing

    different alternatives, but offer guidelines in the field of study

    and typically serve as a base to further research. In LCA studiesusing cornerstone scenarios several options (which can be very

    different) are chosen. The alternatives serve as cornerstones of

    the specific field, allowing for an overall view of the field of

    study.

    3. Methodology for scenario analysis of sustainability

    The introduction to this paper offered a challenge e what

    would lead to major change in the impacts of a sector? The

    review of the previous section showed that scenario analysis

    has been widely developed and is increasingly being used

    for exploring options for future sustainability. This paperaims to build on previous work in two ways.

    Where LCA studies have incorporated scenarios, they

    have focused on purely environmental measures. The pa-

    per proposes additionally the use of quantitative economic

    and social measures to develop a graphic equaliser of

    sustainability indicators to allow comparison of different

    scenarios.

    Previous studies have used scenario analysis to consider

    the future of urban or regional systems, products and sup-

    ply chains. This paper attempts to apply scenario analysis

    to a whole sector.

    According to the classifications in Section 2.1, the paper

    specifically aims at exploratory scenarios: the aim is to

    explore the possibility that an entire sector could be rede-

    signed and to evaluate the potential sustainability of such alter-

    native designs, in order to anticipate targets to direct current

    decision-making. Quist and Vergragt [30] discuss creation of

    future visions but limit their discussion to scenarios which

    conceptualise technological and social innovations that are

    imaginable now [31] and emphasise the creation of follow-

    up agendas and implementation plans. In exploring possible

    long-term futures of a complex sector, such as the clothing

    and textiles sector, the creation of such plans would be

    difficult, and probably largely hypothetical. For this study,

    the development of a transition process to connect the present

    to the future scenarios was deemed less important than identi-

    fying the scenarios that would lead to a major change in

    impacts. This approach could be questioned: are scenarios

    created in this way meaningful if a change process has not

    been specified? In a specific business strategy exercise, theanswer to this question would probably be negative. However,

    in considering long-term ambitions of sustainability the tran-

    sition process is clearly complex, and depends greatly on the

    willingness of customers. In turn, customers are influenced

    by their vision of how their choices might lead to a more sus-

    tainable future and what that future could be. Accordingly, the

    aim of this study is to define targets to give a vector for

    future decision-making and information.

    The literature contains various models for constructing sce-

    nario analyses, all with a similar basic structure. Fig. 1 pres-

    ents the process as described in this paper. The blocks at the

    left of the figure are typical of such processes in the literature,

    and are adapted from Bood and Postma [6]. The process is asfar as possible undertaken sequentially, but it is an interactive

    and iterative process where final scenarios are constantly being

    refined to come up with an agreed set of scenarios to examine

    for a particular situation. The right side of the diagram empha-

    sises the significance of dialogue between the project team and

    representatives of stakeholder groups across the sector.

    The remainder of this section describes key components of

    the process of Fig. 1.

    3.1. Understanding the sector as it is

    A first requirement for analysing the future of a sector is to

    characterise its operation at present. For most sectors, associ-

    ations and analysts will have this knowledge, so from literature

    searches it is possible to develop an initial map of the range of

    businesses required to allow completion of final consumer

    products. However, such a map is in fact a snapshot of how

    the sector operates at present e and represents only one stage

    in its evolution. The current arrangement of businesses, their

    location and size, capabilities and culture, have evolved over

    time to balance various objectives, and will continue to

    change. The objective of scenario analysis is to consider

    whether a different form of the sector would allow a different

    balance between these objectives, specifically as far as this pa-

    per goes, balancing the objectives of the triple bottom line.

    Accordingly, an influence diagram can be prepared e to char-

    acterise the major influences that have led to the current

    format of the sector, and its consequences. Strategy courses

    in business schools typically use a PEST (Political, Economic,

    Social, Technological) framework for analysis of the external

    forces acting on a business or sector e or in a more extended

    form a PESTLECH framework (adding Legal, Ecological,

    Cultural and Historic factors). In the context of sustainability,

    the consequences of the operation of the sector can be grouped

    according to the economic, environmental and social measures

    of the triple bottom line.

    1237J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    5/13

  • 8/4/2019 An Aproach to Scenario Analysis

    6/13

    information) is required to allow comparison between the sce-

    narios. A difficulty in all discussion of sustainability is that

    by definition many different measures must be considered, often

    with different units. In order to assist in comparing the impacts

    predicted for each scenario, the different measures will be pre-

    sented in a consistent form using a graphic equaliser.

    Environmental assessment of scenarios was conducted using

    a standard LCA approach based on the Danish methodology for

    SERVICE

    PROVIDERS

    APPAREL

    INDUSTRY

    CHEMICAL

    INDUSTRY

    RETAIL

    NON-CONVENTIONAL

    TEXTILE PROCESSING

    CHEMICAL FIBRE

    INDUSTRY

    AGRICULTURETEXTILE

    INDUSTRY

    Fibre Yarn FabricFinished

    fabric

    Interior andhome

    textiles

    Textile applications

    Clothing

    (fashion)

    Distribution and retail

    Textileservices

    Private use /consumption

    Commercialuse

    - transport- construction- furniture- agriculture

    - hotels- hospitals- public services

    - high street- specialist- independent store

    - supermarket- online shop

    POST-CONSUMERRECOVERY

    ANDDISPOSAL

    Fig. 2. Sector map for the clothing and textiles sector (from Well dressed?, Allwood et al. [4]).

    Climate change - laundry

    Toxic chemicals - cotton agriculture,pre-treatment, dyeing, printing

    Waste to landfill

    Water consumption - cotton

    7% of world exports

    ~26 million direct employees

    Volume growing prices dropping

    Major export earner for somecountries

    Working hours, safety, child labour

    Labour insecurity

    Minimum living wage/ legal wage

    Rights to association

    AGRICUL-

    TURE

    +MINERALS

    Economic

    Environmental

    Social

    Historical

    Cultural

    Ecological

    Legal

    Technological

    Sociological

    Economic

    Political

    INFLUENCES EFFECTS

    RAW

    MATERIALS

    FIBRE

    FABRIC

    PRODUCTS

    RETAIL

    USE

    DISPOSAL

    REUSERECYCLE

    RECOVER

    Fig. 3. Influence diagram for the clothing and textiles sector.

    1239J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    7/13

    environmental design of industrial products. Calculations were

    performed using the GaBi-EDIP software. The data used for the

    analysis was based on that collected from the Danish EDIPTEX

    project [35] and is believed to be the most comprehensive foranalysis of clothing and textiles products. Boundaries around

    the analysis were set to include inputs of energy, water and

    auxiliaries, but exclude capital goods, services and infrastruc-

    ture. Typically, LCA studies report many impact categories.

    However, in aiming to present measures across the triple bot-

    tom line of sustainability in a comprehensible format, some se-

    lection was required to limit the number of environmental

    impacts reported. Three categories were chosen to attempt to

    represent overall impacts: climate change impact (reported in

    thousand tonnes of CO2 equivalent), waste volume (reported

    in thousand tonnes) and an aggregated environmental index

    which gave a single measure of ozone depletion, acidification,

    nutrient enrichment and photochemical ozone formation. The

    impacts within this aggregated index are usually reported sepa-

    rately, so were expressed in Person Equivalent Targeted

    units e normalised to the fair share of one person e and

    weighted according to political reduction targets. Strictly, LCA

    studies are only comparable within identical boundaries e so

    the use of LCA for scenario analysis which might includealternative processes, materials or locations could be mislead-

    ing. To attempt to minimise discrepancies from such changes,

    it was assumed that all electrical energy was generated accord-

    ing to the profile of a single country. No formal third party

    review of the study and the results has been carried out as this

    is not required by the ISO 14044 standard. However, the LCA

    model and the final results have been discussed with experts

    at LCA center, Denmark. Based on the spot sample carried

    out no significant errors were identified and the overall approach

    was judged to be well suited to fulfil the goals of the study.

    Economic assessment has been carried out via a simplified

    set of National accounts calculated for each country involved

    in the scenario. The accounting system is based on the Euro-pean System of Accounts (ESA) 1995 framework as applied

    in the UK [36] and illustrated in Fig. 4. For each product,

    an account of costs is created, showing the build up of the

    retail price to UK consumers in which all profits taken by

    companies in the supply chain are recorded as costs. Transfer

    prices are calculated as the product passes between tiers of the

    supply chain. Relevant components of the accounts are allo-

    cated to the country in which each activity occurs (it is

    assumed that all companies are owned within the country in

    Table 1

    Themes and scenarios used for analysis

    Theme Scenarios

    Production structure Localise production in the UK

    Localise production in the UK and use innovative

    labour saving technologies

    Localise production in the UK, use innovative

    labour saving technologies, and base productionon locally recycled materials

    Consumer influence Extending the life of clothing

    Best practice in clothes cleaning

    New materials

    and processes

    Alternative fibres

    Green manufacturing

    Smart functions

    Government influence Reduced barriers to free trade

    Imposition of eco-tax

    Table 2

    Three case study products

    Knitted cotton T-shirt Woven viscose blouse Tufted polyamide

    carpet

    Cotton farmed

    and spun in the USA

    Viscose made from

    cellulose harvested

    and processed in

    India

    Polyamide face fibres

    and polypropylene

    primary backing

    made in the USA

    Latex secondary

    backing made in

    the UKYarn knitted, dyed, cut

    and sewn in China

    Fibre is spun,

    woven, dyed,

    cut and sewn

    in India

    Carpet tufted

    and dyed in the

    UK largely

    using automated

    machinery

    Wholesale and retail

    in the UK

    (460 million per year)

    Wholesale and retail

    in the UK

    (33 million per year)

    Wholesale and retail

    in the UK

    (8.5 million m2

    per year)

    Twenty-five washes

    at 60 C with

    tumble drying and ironing

    Twenty-five washes

    at 40 C with

    hang drying and no

    ironing

    Vacuum cleaning

    only in use over a

    10-year life-span

    Incinerated after disposal Incinerated after

    disposal

    Landfill after

    disposal

    Cost of farming/mining

    Transfer price (material)

    Cost of material processing

    Transfer price (fibre)

    Cost of spinning etc

    Transfer price (yarn)

    Cost of weaving/knitting

    Transfer price (fabric)

    Cost of making up

    Transfer price (wholesale)

    Cost of retail

    Price to consumer

    Country 1 Country 2 UK

    + /

    +

    +/

    +/

    +

    +

    Gross National Income

    Balance of Trade

    Operating surplus

    People employed

    Fig. 4. Economic model for scenario assessment (example for cotton T-shirt).

    1240 J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    8/13

    which they operate). A contribution of the product to gross

    national income of each country can be calculated. In addition,

    for the UK, the balance of trade and an operating surplus are

    calculated.

    A key decision in developing an economic model for use in

    this context is to determine how consumer prices vary as pro-

    duction/supply-chain costs change. There are several possibili-ties: prices could be held constant, so that any change in cost is

    reflected in a reduction in profits; profit margins could be fixed,

    sothat any change in cost is reflectedin a change in final price to

    the consumer; a market model could be used to predict the

    proportion of changed costs that would be passed to the con-

    sumer, and to incorporate some form of price elasticity to

    show sales volumes changing with price. Naturally, the

    simplest choice (fixed prices) assumes unrealistic consumer

    behaviour, while more complex models are strongly dependent

    on their assumptions. The intention of this work was to find

    a simple way to link predictions across the triple bottom line

    so a simple model was required. Given that the three case study

    products are commodities, the price is largely determined bycompetitors, so the model used assumed fixed consumer prices

    with all intermediate profits (for instance to raw material sup-

    pliers) treated as costs. Thus a change in supply chain costs leads

    to a change in the profitability of the retailer in the UK, and

    hence to the UKs operating surplus, but the product price and

    demand volume is assumed to be constant. The economic quan-

    titative predictions of the analysis should, therefore, be inter-

    preted as indicators of effects that will create change, not as

    accurate predictions of the final change e the retailer would

    not continue to sell commodity products at a loss and would

    either change supplier or increase the price.

    Quantitative social assessment is problematic as most of thesocial consequences of operation of the sector are difficult to

    quantify in a way that could be meaningfully related to the eco-

    nomic model. However, a reasonable prediction can be made of

    the number of people employed due to the case study products

    based on figures for productivity andworking hours. This is only

    partially meaningful e if a country has full employment then

    any new jobs created within a sector must be substitutes for

    other jobs e so will cause loss of activity elsewhere. However,

    in the case of the clothing sector, in agriculture and production

    in particular, this approach is reasonable, as most jobs are

    relatively unskilled and low paid, so will be available to those

    who might otherwise struggle to find employment.

    For each theme, the quantitative measures described abovewere calculated for the basecase (production of the case study

    products as at present) and under the conditions of each sce-

    nario. The full results are presented in Allwood et al. [4]

    with further details in the associated technical annex to that re-

    port. A summary of the quantitative results of the analysis for

    the cotton T-shirt in the basecase and scenarios for theme 1 (as

    described in Table 1) is given in Table 3, demonstrating the ex-

    tensive data generated by the process.

    Presentation of the results of the analysis in a table requires

    very careful reading to evaluate the trade-offs between mea-

    sures of quite different impacts. To assist in this interpretation

    and provide a visual means to aid discussion of the compro-

    mises implied by a triple bottom line, the results have beenpresented with a graphic equaliser as illustrated for the

    data of Table 3 in Fig. 5. The intention of this display is to

    demonstrate firstly, which scenarios lead to significant change

    in key measures, and secondly to allow comparison between

    changes in different measures. Thus the display is intended

    primarily to allow comparison of relative change in magni-

    tudes, not to make decisions about whether one particular

    measure is more important than another. Accordingly, a scale

    factor was defined for each of the measures e so that the

    graphic equaliser display was consistent between all case

    study products and all scenarios. The resulting images allow

    visual comparison between scenarios, across measures andbetween countries. Both Table 3 and Fig. 5 give the same

    information, but from Fig. 5 it is immediately clear that, in

    the case of the cotton T-shirt: energy use is dominated by laun-

    dry (in the UK); the major economic benefit is in the UK (due

    to the high gross margins of retail); shifting production to the

    UK would cause a surge of jobs in the UK and extremely high

    Table 3

    Summary of quantitative results from scenario analysis for the cotton T-shirt in theme 1

    Environment Social Economic

    Climate change

    (1000 tonnes CO2equivalent)

    Waste

    (1000 tonnes)

    Environmental

    impact(PET/1000)

    Employment

    (in thousands)

    GNI (m) Balance of

    trade(m)

    Operating

    surplus(m)

    USA Basecase 969 161 313 10 252

    Scenario 1 954 161 307 10 252

    Scenario 2 876 148 281 9 231

    Scenario 3 448 75 144 2 47

    UK Basecase 1918 208 266 26 2,318 902 1,887

    Scenario 1 2239 220 301 173 2,968 252 111

    Scenario 2 2169 222 293 27 2,989 231 2,541

    Scenario 3 2523 255 330 32 3,174 46 2,645

    China Basecase 374 12 88 108 650

    Global total Basecase 3261 381 667 144

    Scenario 1 3193 381 608 183

    Scenario 2 3044 369 575 36

    Scenario 3 2971 331 474 34

    1241J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    9/13

    costs (shown in the low operating surplus) which could be

    avoided by new labour saving technology; shifting production

    to the UK has little environmental benefit, but manufacturingwith recycled materials would be significant.

    3.4. Stakeholder dialogue

    The scenarios of Table 1 were analysed according to the

    measures of Section 3.3, and the results were presentedas a draft

    report, including box stories of associated information indi-

    cating major impacts or consequences of each scenario that

    arose from the influences diagram of Fig. 3 but were not cap-

    tured in the quantitative model. This draft report was circulated

    widely to stakeholders across the sector with a request for feed-

    back. The feedback proved strongly valuable e in identifying

    results that were partial or misleading, and confirming or chal-lenging the conclusions from the quantitative analysis. In fact, it

    appears that the highest quality of feedback could be achieved

    once a draft report was completed e as it presented, in some

    cases, a challenge to existing views. A time limit was set for

    receiving comments, and the final report prepared from the draft

    after all feedback had been considered.

    4. Results from analysis of the UK clothing and textiles

    sector

    The graphic equaliser displays for each scenario were

    used to draw conclusions about the future of the sector,

    and for each scenario, discussion based on the influences di-

    agram of Fig. 3 allowed consideration of other consequences,

    and of the challenge to implementing such change. An obvi-ous question arising from each scenario is whether it is likely

    that this might develop e and this led to the realisation that

    any of the scenarios considered could become reality if the

    UKs consumers collectively wished it. From this, it was

    possible to develop a description of the ideal behaviour

    of consumers that would drive change in the sector. Having

    done so, it becomes possible to structure a discussion about

    existing barriers that inhibit development of such

    behaviour e and hence to means by which the barriers might

    be overcome.

    4.1. Scenario assessment

    Analysis of the effects of changed production structure

    showed that for the cotton T-shirt, energy consumption is dom-

    inated by the use-phase, so changes in production structure, in-

    cluding recycling have little effect on energy use e as the

    same use requirements remain. As the energy required for

    laundry typically comes from electricity, this means that waste

    volumes (dominated by mining waste from extracting fossil

    fuels) are also largely unaffected. However, for the viscose

    blouse and the carpet, where energy requirements are concen-

    trated in the material phase, recycling is beneficial. The use of

    energy for transport is proportionately low in all three cases so

    localisation on its own has little benefit, and shifting cutting

    CCI WASTE EI

    BOT

    OS EMPGNI

    United Kingdom: impacts

    CCI WASTE EI

    United States of America: impacts

    EMPGNI

    CCI WASTE EI

    Global: impacts

    CCI WASTE EI

    EMPGNIBase case

    Changing the locationof existing operationsChanged location withnew production technologyChange location with new produc-tion technology and local recycling

    Key

    China: impacts

    Fig. 5. Example of the graphic equaliser display of triple bottom line effects (from Well dressed?, Allwood et al. [4]).

    1242 J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    10/13

    and sewing operations from Asia to the UK, causes the loss of

    jobs in Asia and economic loss in the UK due to the high costs

    of employing such labour. However, if labour saving technol-

    ogy can be used, the UK would benefit from such a shift. Such

    technologies are being developed e 3D knitting machines

    capable of manufacturing whole garments without manual

    intervention are now common in producing underwear, swim-wear, sports wear and T-shirts amongst others. The global

    proportion of seamless underwear production rose from 2%

    in 1998 to 18% in 2003. Automated production of clothes

    from recycled materials would potentially be economically

    attractive in the UK and globally environmentally beneficial

    at the cost of lost jobs in Asia.

    Analysis of changed consumer behaviour showed that

    a consumer decision to reduce the energy used in laundry, par-

    ticularly of cotton products, would be highly significant. The

    two key areas of change arise from switching from tumble

    drying to hang drying (preferably outdoors to avoid any

    demand for increased heating), and washing at lower temper-

    atures. For products in which the material or productionphases dominate impacts, consumer change to extend the

    life of products would be immediately beneficial. Some exam-

    ples of this occur with hiring clothes (for weddings, or for

    work uniforms) which could be extended, and the centuries

    old tradition of clothes repair could be renewed through design

    for repair, labour saving technologies and new approaches to

    the supply of spare parts.

    Conventional cotton agriculture requires intense use of

    toxic chemicals in growth and prior to harvesting. The toxic

    impact of cotton would be greatly reduced by a switch to

    organic cotton, and although this would lead to higher prices

    (organic cotton is currently around 50% more expensivethan conventional cotton) the total cost of cotton in a typical

    7 T-shirt is around 0.28, so the price rise is not significant

    if other processes are unaffected. Currently capacity for or-

    ganic cotton growing is constrained, but potentially this can

    be overcome. A switch from man-made to natural fibres in

    carpet manufacture would have mixed economic and environ-

    mental effects, and the analysis was inconclusive. However,

    innovation with smart functions able to change the behaviour

    of fibres in use appears to be valuable: nano-technology coat-

    ings that extend the life of man-made carpet could reduce

    demand for production of new products and hence have

    a net benefit; novel smart functions able to allow more

    wear of a garment between laundry cycles would have benefit

    for all materials. A potential drawback from such innovations

    is that they may inhibit recycling and they are some way from

    gaining consumer confidence.

    If all remaining trade barriers were removed, production

    structures would be unlikely to change significantly, but

    the removal of current cotton subsidies in the USA (equal

    to roughly 25% of the market price) would make USA cot-

    ton less attractive and allow increased cotton trade from

    poorer countries. In the past five years, removal of trade

    barriers has led to reduced prices for UK consumers e

    and more liberalisation would be likely to promote this effect

    further.

    4.2. The ideal consumer

    For many of the environmental impacts of the clothing and

    textiles sector, change depends largely on consumer choice e

    to launder clothes in a different way, and to buy fewer of them.

    Based on the assessment of the sector, it is possible to propose

    a model of ideal consumer behaviour.

    Second hand purchases, leasing items that would other-

    wise not be used to the end of their natural life, and repair-

    ing (or updating) old garments are all environmentally

    preferable to purchasing new products requiring new

    materials. For cotton this behaviour would significantly

    reduce toxicity and for man-made materials, it would

    significantly reduce energy requirements dominated by

    production.

    Purchasing decisions should be based on accurate infor-

    mation about the environmental impacts of their produc-

    tion and the social conditions of those involved in their

    production. Clothes should be washed less frequently and less inten-

    sively, hang drying and ironing should be avoided where

    possible.

    Used clothing and textiles should be disposed through

    recycling businesses that would return them for second

    hand sale where possible, but otherwise recycle the yarn

    or fibres.

    4.3. Barriers to change and means to address them

    The ideal behaviour described above depends on collectiveaction e heroic behaviour by a few purchasers would have lit-

    tle benefit e and it is inhibited by several barriers.

    Consumers in the UK are generally wealthy enough to

    purchase clothing and textiles as much for fashion as func-

    tion, and to replace them before the end of their natural

    life. Recently, prices have dropped and consumers have

    benefited from fast fashion introducing new styles

    more than four times per year, so may be reluctant to

    pay extra for more responsibly made products. UK con-

    sumers do not necessarily see a connection between their

    purchases and negative (but invisible from the UK) social

    and environmental consequences.

    At present, the profits of UK businesses involved in the

    clothing and textiles supply chain are generally linked to

    volume of sales, so reduced volumes will inhibit profitabil-

    ity unless prices rise.

    UK government policy on the environment considers only

    impacts created within the UK, yet in many of the scenar-

    ios, for global environmental indicators to show an im-

    provement, UK indicators must worsen.

    Repair is generally labour intensive and expensive, and the

    rise of fast fashion has led to a flow of cheaper but lower

    quality garments into the UK that are more difficult to

    repair.

    1243J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    11/13

    UK used clothes collection could be increased with im-

    proved infrastructure, and textiles and clothing recycling

    could be improved with better technologies.

    Most clothes washing aims to remove odour but uses a pro-

    cess capable also of removing stains. A reduced intensity

    process that removes odour but not stains would allow a re-

    duction in current washing frequencies.

    Various possible mechanisms for overcoming these barriers

    emerged through discussions with stakeholders.

    Consumer motivation towards the ideal behaviour could

    be promoted by education e with high quality information

    provided by educators, journalists and campaigners as well

    as by business and government.

    The people who help to develop fashion leadership could

    build the idea of durability into future styles.

    The flow of new material driven by the sector could be

    halved without economic loss if consumers paid twice as

    much for products which last twice as long. Retailers e who are the strongest players in the clothing

    and textile supply chain and are largely UK based e can

    seek alternative forms of revenue through new business

    models including repair services, supply of novel coatings

    and fashion upgrades as an alternative to sales related to

    material flow.

    Investment in technology will support development of

    improved recycling technologies, lower temperature wash-

    ing and coatings or other processes to extend product life

    and reduce washing intensity.

    An eco-tax could penalise products using virgin material

    and be used to fund development of material re-use.Legislation may be able to inhibit some toxic impacts

    but is difficult to apply internationally. The UK govern-

    ment could assert environmental and social responsibility

    as part of its negotiation of future international trade

    agreements.

    5. Discussion

    The paper has presented a methodology for triple bottom

    line scenario analysis of large-scale change to a sector, and

    applied it to the supply of clothing and textiles to the UK.

    The results presented in Section 4 have been validated

    through extensive stakeholder dialogue and appear sensible,

    although it is impossible to estimate how complete this set

    of suggestions is, until time has passed. This section attempts

    to reflect in two ways on the approach that has been proposed:

    how effective was the methodology? Can the results of this

    study be used to anticipate useful strategies in other sectors?

    5.1. Assessment of the methodology

    The challenge of providing a methodology to consider

    wide-scale change to a sector is to find a means of analysis

    that is tractable e that can be completed within reasonable

    timee

    but which is sufficiently comprehensive. A sector,

    such as that for clothing and textiles, is sufficiently complex

    to be beyond the comprehension of an individual, so a com-

    prehensive view depends on collaboration across the sector.

    Broadly, the approach offered here e to map the sector, iden-

    tify an influences diagram, create scenarios based on case

    study products, analyse them using the triple bottom line

    graphic equaliser and present draft analysis to stakeholdersfor feedbacke appears to be a sensible solution. The experi-

    ence of the study on clothing and textiles supply to the UK has

    emphasised the importance of the stakeholder feedbacke par-

    ticularly once the draft analysis was complete, as this proved

    to be the trigger for releasing crucial expert insights, often

    when the draft report presented an opposite to a conventional

    (or convenient) view. Future studies should certainly be struc-

    tured to ensure sufficient time is allocated to gather and pro-

    cess such feedback.

    The economic analysis of the sector is obviously simplistice

    but appeared useful in predicting major effects on National

    income of the scenarios for all countries involved. Calculation

    of an operating surplus for the UK proved a valuable way toindicate the likelihood of a particular scenario being adopted e

    for instance, in showing the significance of novel production

    technologies in facilitating a more localised production system.

    While such aggregated national measures are obviously crude,

    they seem more realistic than attempting a more micro-

    economic analysis which would require many more assump-

    tions to provide sufficient data.

    The environmental analysis e as with all properly con-

    ducted LCA e was arduous. While not a universal solution,

    one observation from the results was that in virtually all the

    scenarios, the three environmental indicators used in the

    graphic equaliser were highly correlatede

    suggesting thata simpler measure of energy use would in this case have

    given a similar quality of information with less effort. This

    arises specifically for this sector as most energy is used in

    the form of electricity, and most waste is assumed to be

    incinerated e so mining waste dominates the waste category,

    and most non-climate change indicators are largely related

    to burning fossil fuel. The exception to this simplified

    approach for the clothing case was in the significance of tox-

    icity in cotton growing e which required detailed analysis, and

    was data intensive. No general rule can be applied to simplify-

    ing the analysis, but probably there is by now sufficient anal-

    ysis of most sectors to allow short-cuts to be taken e analysing

    only those effects which are known to have a large impact.

    The quantified social analysis was restricted to an estimate

    of employment e because of the difficulty of providing quan-

    titative predictions of any other measure. With a much more

    complex micro-economic model, it might be possible to

    predict the impact of investment in training on working condi-

    tions e but this would require vastly more detail in the anal-

    ysis, and would again be highly dependent on assumptions.

    Evidence that this broad approach can be translated to a dif-

    ferent situation is provided by Russell [3] who has used the

    same methodology to consider the effect of localising produc-

    tion of two case study products on the island of Jamaica.

    Although that requires a regional rather than sectoral study,

    1244 J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    12/13

    a big picture scenario analysis is similarly required, and

    apparently gives insightful results.

    5.2. Anticipating the outcomes for other sectors

    Development of the methodology presented here, and the

    associated analysis of the UK clothing and textiles sectorrequired five person-year work. Is it possible to anticipate

    the candidate solutions that might apply to other sectors?

    Broadly, two categories of change have been explored in

    this paper: those which aim to reduce the flow of new material

    entering the sector; those which aim to make the processes

    within the sector as it is, more efficient. These solutions are

    summarised in Fig. 6, which may prove a useful template

    for future studies of wide-scale change in other sectors.

    Acknowledgements

    The analysis of the clothing and textiles sector described in

    Sections 3 and 4 of the paper was funded by the UK landfill tax

    credit scheme administered by Biffaward through RSWT with

    a 10% contribution from the UK clothes and food retailer Marks

    and Spencer. The work of Suzana Russell was supported by the

    Commonwealth Scholarship Commission. The Delphi study

    mentioned in Section 3.1 was completed by Marisa de Brito

    who worked on the first half of this project.

    Appendix. A brief introduction to the clothing and

    textiles sector

    The clothing and textiles sector represents about 7% of

    world cross-border trade, leads to sales of over US$ one

    trillion and employs over 26 million people worldwide just

    in production [37]. Production in the sector is dominated by

    Asian countries, with over one million people employed in

    each of China (7.5 million), Pakistan, Bangladesh, India and

    Indonesia, but with significant activity virtually in every

    country e including more than four million in the EU and

    Mediterranean region and two million in North and SouthAmerica [38]. Products in the sector are predominantly

    made from either cellulosic materials such as cotton or poly-

    ester, with one third of world cotton exports from the USA

    [39], aided by government subsidies. The sector has been sub-

    ject to many international trade agreements, most well known

    being the quotas limiting exports (Multi-Fibre Agreement

    1974e1994 and Agreement on Textiles and Clothing 1995e

    2004) which were phased out since 1st January 2005. The

    removal of these quotas has led to a marked drop in prices

    in the UK, while spending has increased, so UK consumers

    have increased the number of garments they have bought by

    one third over four years. The clothing and textiles sector leads

    to a flow of around 3.2 million tonnes of materials through theUK, of which 0.9 million tonnes are exported, 1.8 million

    tonnes are sent to landfill, and the remaining 0.6 million

    tonnes are split between recycling, and emissions to air fol-

    lowing incineration.

    The major impacts of the sector according to the triple bot-

    tom line of sustainability are as follows.

    Environment: energy use associated with laundry (particu-

    larly of cotton products), operating production equipment,

    and production of materials; use of toxic chemicals, partic-

    ularly in cotton production; release of chemicals in waste

    water, particularly from pre-treatment of fibres, dyeing,finishing and laundry; solid waste as illustrated above.

    Social: employees in the clothing sector, who are generally

    relatively unskilled and receive a low income, may have

    precarious contracts, be vulnerable to abuse from em-

    ployers, and often do not have proper representation.

    Leading retailers are working with first tier suppliers to

    develop codes of practice for employment, but it can be

    difficult to impose these on subcontractors.

    Economic: for developed economics, shifting production

    to other countries has not had a significant economic

    impact, as the largest gross margins occur at the wholesale

    and retail end of the supply chain. However, for develop-

    ing countries, the sector may be the major source of export

    earnings e with Cambodia, Haiti, Bangladesh, Pakistan

    and Lesotho all receiving more than 70% of export earn-

    ings from the sector [40].

    References

    [1] Carbon Trust. The UK climate change programme: potential evolution

    for business and the public sector. Carbon Trust; 2005.

    [2] Russell SN, Allwood JM. Sustainable manufacturing: options for physi-

    cally changing the way in which goods are made. Institute for

    Manufacturing, University of Cambridge, Internal Report; 2007.

    Reduction

    in flow

    Customer supportand education

    Durability in placeof fashion

    New businessmodels

    Second-handpurchasing

    Eco-taxation

    Efficiency gains

    Material

    Substitutealternative materials

    Reduce use ofauxiliaries

    Production

    Processefficiencies

    Reduce use ofauxiliaries

    Distribution

    Localiseproduction

    Tariffs, subsidiesand quotas

    Use

    Best practicein use

    Reduced impactauxiliaries

    Disposal

    Incinerate ratherthan landfill

    Recycle materials

    Reducebatch sizes

    Smartfunctions

    Fig. 6. Template of candidate solutions for large-scale change in other sectors.

    1245J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

  • 8/4/2019 An Aproach to Scenario Analysis

    13/13

    [3] Russell SN. The value of localising production as a strategy for sustain-

    able development, PhD thesis, University of Cambridge; 2006.

    [4] Allwood JM, Laursen SE, Malvido de Rodriguez CM, Bocken NMP.

    Well dressed? The present and future sustainability of clothing and tex-

    tiles in the United Kingdom. Institute for Manufacturing, University of

    Cambridge, ISBN 1-902546-52-0; 2006. Available from: ; 2006.

    [5] Pesonen HL, Ekvall T, Fleisher G, Huppes G, Jahn C, Klos ZS, et al.

    Framework for scenario development in LCA. International Journal of

    Life Cycle Assessment 2000;5(1):21e30.

    [6] Bood RP, Postma TJBM. Scenario analysis as a strategic management

    tool. University of Groningen, Research Institute SOM (Systems, Orga-

    nisations and Management); 1998.

    [7] Linneman RE, Klein HE. The use of multiple scenarios by US industrial

    companies. A comparison study, 1977e1981. Long Range Planning

    1983;16(6):94e101.

    [8] Wack P. Scenarios: uncharted waters ahead. Harvard Business Review

    1985;63(5):73e89.

    [9] Phelps R, Chan C, Kapsalis SC. Does scenario planning affect perfor-

    mance? Two exploratory studies. Journal of Business Research 2001;

    51:223e32.

    [10] Schoemaker PJH. When and how to use scenario planning. Journal of

    Forecasting 1991;10(6):549e64.

    [11] Schoemaker PJH. Scenario planning: a tool for strategic thinking. Sloan

    Management Review 1995;36(2):25e40.

    [12] Mietzner D, Reger G. Advantages and disadvantages of scenario ap-

    proaches for strategic foresight. International Journal of Technology

    Intelligence and Planning 2005;1(2):220e39.

    [13] Ringland G. Scenarios in business. Chichester: Wiley; 2002.

    [14] Kahn H, Wiener AJ. The year 2000: a framework for speculation on the

    next thirty-three years. New York: Macmillan; 1967.

    [15] Swart RJ, Raskin P, Robinson J. The problem of the future: sustainability

    science and scenario analysis. Global Environmental Change 2004;14:

    137e46.

    [16] Godet M, Roubelat F. Creating the future: the use and misuse of scenar-

    ios. Long Range Planning 1996;29(2):164e71.

    [17] Van der Heijden K. Scenarios and forecasting: two perspectives. Techno-

    logical Forecasting and Social Change 2000;65(1):31e

    6.[18] RAND. Scenariosfor examiningcivil aviation infrastructureoptionsin the

    Netherlands.Delft: EuropeaneAmericanCentrefor Policy Analysis; 1997.

    [19] Rienstra SA. Options and barriers for sustainable transport policies: a

    scenario approach. Amsterdam: Department of Spatial Economics, Vrije

    Universiteit; 1998. 254 pp.

    [20] Godet M. The art of scenarios and strategic planning: tools and pitfalls.

    Technological Forecasting and Social Change 2000;65:3e22.

    [21] van Veen-Croot DB, Nijkamp P, van den Bergh JCJM. A scenario study

    for investigating the implications of globalisation on international trans-

    port and the global environment: a case study for the Dutch paper

    industry. In: Serie research memoranda, vol. 2000e3. Amsterdam:

    Department of Spatial Economics, Free University; 2000.

    [22] Notten PWF, Rotmans J, Van Asselt MBA, Rothmans DS. An updated

    scenario typology. Futures 2003;35(5):423e43.

    [23] Fukushima Y, Hirao M. A structured framework and language for

    scenario-based life cycle assessment. International Journal of Life Cycle

    Assessment 2002;7(6):317e29.

    [24] Leemhuis JP. Using scenarios to develop strategies. Long Range Plan-

    ning 1985;18(2):30e7.

    [25] Rienstra SA, Nijkamp P, Smokers RTM, Vleugel JM. Scenarios in

    decision-making. Petten, ECN-Policy Studies. Amsterdam: Vrije Univer-

    siteit; 1996.

    [26] Kelly R, Sirr L, Ratcliffe J. Futures thinking to achieve sustainable devel-

    opment at local level in Ireland. Foresight 2004;6(2):80e90.

    [27] UNEP (United Nations Environment Programme). Global environment

    outlook 3 (GEO-3). United Nations Environmental Programme; 2002,

    ISBN: 9280720872.

    [28] Raskin P, Banuri T, Gallopn G, Gutman P, Hammond A, Kates R, et al.

    Great transition: the promise and lure of the times ahead. Global Scenario

    Group (GSG). Stockholm Environment Institute (SEI); 2002.

    [29] Nakicenovic N, Swart R, editors. Special report on emissions scenarios.

    Cambridge, UK: Cambridge University Press; 2000.

    [30] Quist J, Vergragt P. Past and future of backcasting: the shift to stake-

    holder participation and a proposal for a methodological framework.

    Futures 2006;38:1027e45.

    [31] Green K, Vergragt P. Towards sustainable households: a methodology

    for developing sustainable technological and social innovations. Futures

    2002;34:381e400.

    [32] Ubbels B, Rodenburg C, Nijkamp P. A multi-layer scenario analysis for

    sustainable international transport. Transportation Planning and Technol-

    ogy 2003;26(1):69e103.

    [33] Tan RBH, Khoo HH. An LCA study of a primary aluminium supply

    chain. Journal of Cleaner Production 2005;13:607e18.

    [34] Sonesson U, Berlin J. Environmental impact of future milk supply chains in

    Sweden: a scenario study. Journalof Cleaner Production 2003;11(3):253e66.

    [35] Laursen SE, Hansen J, Knudsen HH, Wenzel H, Larsen HF, Kristensen

    FM. EDIPTEX e environmental assessment of textiles. Working report

    no. 3, 2006. Danish Environmental Protection Agency [in Danish]. Iscurrently being translated to English by DEPA; 2006.

    [36] UK National accounts. UK national accounts concepts, sources and

    methods. Office for National Statistics, ; 1998.

    [37] ILO. Global employment trends brief 2006. International Labour Office,

    ; January 2006.

    [38] UNIDO (United Nationals Industrial Development Organization). IND-

    STAT4 Industrial Statistics database, ; 2006.

    [39] ICAC (International Cotton Advisory Council), ; 2001.

    [40] UNCTAD (United Nations Trade Analysis and Information System data-

    base), ; 2003.

    1246 J.M. Allwood et al. / Journal of Cleaner Production 16 (2008) 1234e1246

    http://www.ifm.eng.cam.ac.uk/http://www.ifm.eng.cam.ac.uk/http://www.statistics.gov.uk/http://www.ilo.org/http://www.unido.org/http://www.icac.org/http://www.unctad.org/http://www.unctad.org/http://www.icac.org/http://www.unido.org/http://www.ilo.org/http://www.statistics.gov.uk/http://www.ifm.eng.cam.ac.uk/http://www.ifm.eng.cam.ac.uk/