THE  PAPER  TO  DIGITAL  MEDIA  TRANSITION:  DEFINING  SUSTAINABILITY  IN  MEDIA  SUPPLY  CHAINS  by  Justin  G.  Bull    A  THESIS  SUBMITTED  IN  PARTIAL  FULFILLMENT  OF  THE  REQUIREMENTS  FOR  THE  DEGREE  OF    DOCTOR  OF  PHILOSOPHY  in  THE  FACULTY  OF  GRADUATE  AND  POSTDOCTORAL  STUDIES  (Forestry)    THE  UNIVERSITY  OF  BRITISH  COLUMBIA  (Vancouver)      November  2014    ©  Justin  G.  Bull,  2014                 iiAbstract  The  phrase,  “please  consider  the  environment  before  printing  this  email”  has  entered  the  common  vernacular.  It  suggests  that  when  I  consider  the  environment,  paper  media  is  of  particular  concern,  and  by  inference,  digital  media  is  not.  This  thesis  tackles  the  legitimacy  of  this  claim  by  examining  how  media  sustainability  operates  from  three  critical  perspectives:  industry,  consumers,  and  academia.  To  measure  the  paper  industry’s  perspective,  a  series  of  interviews  with  business  executives  along  a  supply  chain  were  conducted.  I  found  that  collaboration  between  supply  chain  actors  is  a  prerequisite  for  improving  environmental  performance.  To  gain  insight  on  the  consumer’s  perspective,  I  surveyed  1,400  individuals  in  North  America,  investigating  media  habits  and  environmental  values.  I  found  that  consumers  are  shifting  from  paper  to  digital  media,  but  that  environmental  values  have  no  influence  over  this  shift.  This  suggests  that  consumers  could  be  detached  from  the  environmental  impacts  of  their  media  choices.  Finally,  the  academic  perspective  was  analyzed  through  a  comprehensive  review  of  life  cycle  assessment  (LCA)  research  that  compares  paper  and  digital  media  from  an  environmental  perspective.  The  studies  found  that  digital  media  is  almost  always  preferable  to  paper,  requiring  less  energy  and  materials.  However,  they  did  not  assess  the  assumptions  required  in  order  to  compare  such  different  products.  More  worryingly,  the  context  of  media  consumption  –  the  industrial  systems  that  produce  paper  and  digital  products  –  was  never  taken  into  account.  I  conclude  that  since  a  significant  media  shift  is  underway  new  methods  are  required  to  consider  sustainability.  The  new  methods  should  be  anchored  in  two  concepts  that  could  improve  considerations  of  the  environmental  performance  of  industrial  systems.  First,  industrial  ecology,  the  idea  that  industry  might  mimic  nature,  can  strengthen  initial  assessments  of  environmental  performance.  Second,  capability  maturity  models  can  assist  in  gauging  the  ability  of  industrial  systems  to  manage  and  improve  environmental  performance  over  time.                 iiiPreface  While  I  completed  all  of  the  writing,  research,  and  analysis  present  in  this  dissertation,  credit  must  be  given  my  collaborators.  I  designed  the  research  program  on  my  own,  although  the  advice  of  my  supervisory  research  committee  –  Drs.  Robert  A.  Kozak,  Paul  McFarlane,  and  David  H.  Cohen  –  was  sought.  The  analysis  of  my  research  results  was  conducted  entirely  by  myself,  unless  noted  otherwise.  There  are  four  components  to  my  research  where  I  had  the  privilege  of  collaborating  with  other  researchers  and  benefitting  from  the  scrutiny  of  peer-­‐review.      The  first  is  Chapter  1.1.1,  “The  Sustainability  Shift.”  A  lengthier  version  of  this  chapter  has  been  peer-­‐reviewed  and  is  set  to  appear  as  “The  meaning  and  means  of  environmental  sustainability:  Forestry  in  a  more  responsible  world”  in  an  untitled  book  published  by  the  Value  Chain  Optimization  Network,  a  research  group  supported  by  the  National  Sciences  and  Engineering  Research  Council  (NSERC).  I  co-­‐authored  this  chapter  with  Dr.  Robert  A.  Kozak,  but  I  was  responsible  for  most  of  the  work.  I  designed  the  research  methodology,  identified  all  relevant  source  material,  conducted  the  analysis  of  research  results,  and  was  lead  author  on  all  drafts  of  the  paper.      The  second  collaborative  effort  is  Chapter  2,  “The  Case  of  Supply  Chains:  Carbon’s  Role  in  Paper  Media.”  A  version  of  this  chapter  was  peer-­‐reviewed  and  published  in  the  Journal  of  Forest  Products  Business  Research  (Volume  8,  Number  2)  in  2011.  While  I  was  lead  author,  I  collaborated  with  Graham  Kissack,  Dr.  Christ  Elliot,  Dr.  Robert  A.  Kozak  and  Dr.  Gary  Q.  Bull  on  the  preparation  and  publication  of  the  paper.  The  design  of  the  research  program  was  a  joint  effort,  although  I  was  tasked  with  implementing  the  majority  of  proposed  research.  I  conducted  interviews  over  the  phone  and  in  person  with  corporate  executives  in  North  America.  I  wrote  all  drafts  of  the  paper  and  incorporated  any  edits  that  were  suggested  by  my  co-­‐authors.  Mr.  Kissack  conducted  the  carbon  footprint  analysis,  although  I  led  the  summation  and  write-­‐up  of  his  work.  The  research  was  conducted  after  receiving  a  certificate  of  approval  from  the  Behavioural  Research  Ethics  Board  (BREB  certificate  #H08-­‐02734).      The  third  collaborative  effort  is  found  in  Chapter  3,  “The  Case  of  Consumers:  Environmental  Values  and  Media  Consumption.”  I  was  lead  author  on  this  chapter,  but  worked  closely  with  Dr.  Robert  A.  Kozak  on  the  design,  implementation  and  analysis  of  the  research  program.  I  used  a  web-­‐based  survey  to  engage  consumers  in  North  America  with  the  approval  for  the  Behavioural  Research  Ethics  Board  (BREB  certificate  #H09-­‐03036).  I  wrote  every  draft  of  the  paper,  including  all  feedback  from  my  co-­‐author.  And                 ivwhile  I  took  the  lead  in  designing  the  survey  and  collecting  results,  the  advice  of  my  co-­‐author  was  invaluable.  Most  importantly,  Dr.  Kozak  played  an  instrumental  role  in  guiding  statistical  analysis.  However,  the  write-­‐up  of  all  results,  discussions,  and  conclusions  were  my  own  efforts.      The  fourth  and  final  collaborative  effort  is  found  in  Chapter  4,  “The  Case  of  Academic  Comparison:  Efforts  to  Measure  Paper  and  Digital  Media.”  A  version  of  this  chapter  was  peer-­‐reviewed  and  published  in  February  of  2014  in  the  journal  Environmental  Impact  Assessment  Review  under  the  title  “Comparative  life  cycle  assessments:  The  case  of  paper  and  digital  media”  and  can  be  found  on  pages  10  through  18.  My  co-­‐author  for  this  paper  was  Dr.  Robert  A.  Kozak.  I  was  once  again  lead  author,  designing  and  implementing  the  proposed  research  program.  I  conducted  the  necessary  literature  review,  gathered  all  relevant  data,  and  wrote-­‐up  all  results.  Dr.  Kozak  provided  input  throughout  the  process,  with  particular  focus  on  the  discussion  and  conclusions  of  the  paper.      Full  Citations:  J.  G.  Bull,  G.  Kissack,  C.  Elliott,  R.A.  Kozak,  and  G.Q.  Bull.  (2011)  Carbon’s  Potential  to  Reshape  Supply  Chains  in  Paper  and  Print.  Journal  of  Forest  Products  Business  Research.  8(2).    J.  G.  Bull  and  R.  A.  Kozak.  (2014).  Comparative  life  cycle  assessments:  The  case  of  paper  and  digital  media.  Environmental  Impact  Assessment  Review.  February  (1),  10-­‐18.    J.G.  Bull  and  R.A.  Kozak.  (In  Press).  The  Means  and  Meaning  of  Sustainability:  The  Role  of  Forestry  in  a  Responsible  World.  In:  S.  D’Amours,  M.  Ouhimmou,  J.  Audy  and  Y.  Feng,  ed.,  Forest  Value  Chain  Optimization  and  Sustainability,  1st  ed.  Boca  Raton:  CRC  Press/Taylor  &  Francis.                   vTable  of  Contents  Abstract  ...............................................................................................................................................  ii  Preface  ...............................................................................................................................................  iii  Table  of  Contents  .................................................................................................................................  v  List  of  Tables  .......................................................................................................................................  vii  List  of  Figures  .....................................................................................................................................  viii  List  of  Abbreviations  ............................................................................................................................  ix  Acknowledgements  .............................................................................................................................  x  Dedication  ...........................................................................................................................................  xi  Chapter  1:  Introduction  ........................................................................................................................  1  1.1   Background  and  Literature  Review  .........................................................................................  1  1.1.1   The  Sustainability  Shift  ...........................................................................................................  2  1.1.1.1   Definitions  of  Sustainability  ............................................................................................  2  1.1.1.2   Drivers  of  Sustainability  ..................................................................................................  8  1.1.1.3   Responses  to  Sustainability  ...........................................................................................  15  1.1.2   The  Media  Shift  ....................................................................................................................  20  1.1.2.1   Digital  Media  .................................................................................................................  21  1.1.2.2   Paper  Media  ..................................................................................................................  25  1.2   Research  Questions  ..............................................................................................................  29  Chapter  2:  The  Case  of  Supply  Chains:  Carbon’s  Role  in  Paper  Media  .................................................  33  2.1   Abstract  ...............................................................................................................................  33  2.2   Introduction  .........................................................................................................................  33  2.3   Background  ..........................................................................................................................  34  2.4   Objectives  ............................................................................................................................  35  2.5   Methodology  ........................................................................................................................  36  2.6   Results  .................................................................................................................................  39  2.6.1   Carbon  Footprint  Analysis  ....................................................................................................  39  2.6.2   Origins  and  Evolution  of  Carbon  Management  ....................................................................  42  2.6.3   Future  Directions  of  Carbon  Management  ...........................................................................  43  2.7   Discussion  ............................................................................................................................  45  2.7.1   Efficient  Supply  Chains  .........................................................................................................  45  2.7.2   Responsible  Supply  Chains  ...................................................................................................  45  2.7.3   Resilient  Supply  Chains  .........................................................................................................  46  2.7.4   Implications  for  Businesses  and  Supply  Chains  ....................................................................  47  2.8   Conclusions  ..........................................................................................................................  48  Chapter  3:  The  Case  of  Consumers:  Environmental  Values  and  Media  Consumption  ..........................  49  3.1   Abstract  ...............................................................................................................................  49  3.2   Introduction  .........................................................................................................................  49  3.3   Background  ..........................................................................................................................  50  3.4   Objectives  ............................................................................................................................  52  3.5   Methodology  ........................................................................................................................  52  3.6   Results  .................................................................................................................................  56  3.6.1   Demographic  Trends  ............................................................................................................  56  3.6.2   Environmental  Values  ...........................................................................................................  59  3.6.3   Media  Consumption  Trends  .................................................................................................  62  3.7   Discussion  ............................................................................................................................  73  3.8   Conclusions  ..........................................................................................................................  75                 viChapter  4:  The  Case  of  Life  Cycle  Comparisons:  Efforts  to  Measure  Paper  and  Digital  Media  ..............  76  4.1   Abstract  ...............................................................................................................................  76  4.2   Introduction  .........................................................................................................................  76  4.3   Background  ..........................................................................................................................  77  4.4   Objectives  ............................................................................................................................  78  4.5   Methodology  ........................................................................................................................  79  4.6   Results  .................................................................................................................................  82  4.6.1   Functional  Unit  Definition  ....................................................................................................  82  4.6.2   Boundary  Selection  ...............................................................................................................  84  4.6.3   Allocation  ..............................................................................................................................  86  4.6.4   Spatial  Variation  and  Environmental  Uniqueness  ................................................................  88  4.6.5   Data  Availability  and  Quality  ................................................................................................  89  4.7   Discussion  ............................................................................................................................  91  4.7.1   Uncertainties  in  Comparative  LCAs  ......................................................................................  91  4.7.2   Assumptions  in  Comparative  LCAs  .......................................................................................  93  4.7.3   The  Drivers  of  Uncertainty  and  Assumptions:  The  Context  of  Moore’s  Law  ........................  95  4.8   Conclusions  ..........................................................................................................................  97  Chapter  5:  Discussion  and  Conclusions  ..............................................................................................  100  5.1   Conclusions  ........................................................................................................................  100  5.1.1   Sustainability  Perspective:  Media  Supply  Chains  ...............................................................  102  5.1.2   Sustainability  Perspective:  Consumer  Media  Habits  ..........................................................  104  5.1.3   Sustainability  Perspective:  Comparing  Media  Choices  .......................................................  106  5.2   Potential  Applications  and  Future  Research  ........................................................................  109  5.2.1   Industrial  Ecology  ...............................................................................................................  109  5.2.2   Capability  Maturity  Models  ................................................................................................  112  5.3   Strengths  and  Limitations  ...................................................................................................  115  5.4   Final  Thoughts  ....................................................................................................................  117  Bibliography  .....................................................................................................................................  118  Appendices  .......................................................................................................................................  138  Appendix  A  ..................................................................................................................................  138  Appendix  B  ..................................................................................................................................  139                   viiList  of  Tables  Table  1:  Supply  chain  emissions  ................................................................................................................  41  Table  2:  The  new  ecological  paradigm  ......................................................................................................  54  Table  3:  Digital  use  score  ...........................................................................................................................  55  Table  4:  Media  activities  ............................................................................................................................  67  Table  5:  Key  findings  of  comparative  LCAs  ................................................................................................  81  Table  6:  Summary  of  functional  units  of  the  comparative  LCAs  reviewed  ................................................  83                   viiiList  of  Figures    Figure  1:  Schematic  definitions  of  sustainability  .........................................................................................  4  Figure  2:  Industrial  ecology  types  ................................................................................................................  5  Figure  3:  Summary  of  research  chapters  and  objectives  ...........................................................................  32  Figure  4:  Map  of  supply  chain  emissions  ...................................................................................................  40  Figure  5:  Gender  breakdown  of  respondents  by  segment  ........................................................................  56  Figure  6:  Age  breakdown  of  respondents  by  segment  ..............................................................................  57  Figure  7:  Education  breakdown  of  respondents  by  segment  ....................................................................  58  Figure  8:  Income  breakdown  of  respondents  by  segment  ........................................................................  59  Figure  9:  Frequency  distribution  of  NEP  scores  by  digital  use  segment  ....................................................  60  Figure  10:  Environmental  trade  offs  breakdown  of  respondents  by  segment  ..........................................  61  Figure  11:  Environmental  outlook  breakdown  of  respondents  by  segment  .............................................  61  Figure  12:  Digital  media  sources  –  NEP  segments  .....................................................................................  62  Figure  13:  Digital  media  sources  –  digital  segments  ..................................................................................  63  Figure  14:  Paper  media  sources  –  NEP  segments  ......................................................................................  64  Figure  15:  Paper  media  sources  –  digital  segments  ..................................................................................  64  Figure  16:  Online  services  –  NEP  segments  ...............................................................................................  65  Figure  17:  Online  services  –  digital  segments  ............................................................................................  66  Figure  18:  Willingness  to  pay  –  NEP  segments  ..........................................................................................  68  Figure  19:  Willingness  to  pay  –  digital  segments  .......................................................................................  68  Figure  20:  Media  habits:  newspaper  or  magazine  delivered  to  home  ......................................................  70  Figure  21:  Media  habits:  books  purchased  from  bookstore  ......................................................................  70  Figure  22:  Media  habits:  books  borrowed  from  library  .............................................................................  71  Figure  23:  Media  habits:  reading  news  on  PC  or  laptop  computer  ...........................................................  71  Figure  24:  Media  habits:  read  news  on  smartphone  or  mobile  device  .....................................................  72  Figure  25:  Media  habits:  books  purchased  for  electronic  reader  ..............................................................  72  Figure  26:  Summary  of  research  chapters  and  conclusions    ...................................................................  102  Figure  27:  Industrial  ecology  types  ..........................................................................................................  110                   ixList  of  Abbreviations  British  Columbia  (B.C.)  Burlington  Northern  Santa  Fe  Railways    (BNSF)  Business  to  business  (B2B)  Capability  maturity  models  (CMM)  Carbon  Disclosure  Project  (CDP)  Cathode  ray  tube  (CRT)  Corporate  social  responsibility  (CSR)  Digital  subscriber  line  (DSL)  Economic  Input-­‐Output  (EIO)  End  of  life  (EOL)  Environmental  Defense  Fund  (EDF)  Environmental  management  systems  (EMSs)  Environmental  non-­‐governmental  organizations  (ENGOs)  Environmental,  social,  and  governance  (ESG)  Information  and  communication  technology  (ICT)  Input-­‐Output  (IO)  International  Panel  on  Climate  Change  (IPCC)  Global  warming  potential  (GWP)  Life  cycle  assessment  (LCA)  Life  cycle  inventory  (LCI)  Megajoules  (MJ)  New  ecological  paradigm  (NEP)  Non-­‐state  market-­‐driven  (NSMD)  Original  equipment  manufacturer  (OEM)  Personal  computer  (PC)  Personal  digital  assistant  (PDA)  Reduced  Emissions  from  Deforestation  and  Degradation  (REDD)  Regional  Greenhouse  Gas  Initiative  (RGGI)  United  Nations  Environment  Programme’s  (UNEP)  United  States  Climate  Action  Partnership  (USCAP)  Washington  Marine  Group  (WMG)  Waste  Electrical  and  Electronic  Equipment  (WEEE)  Western  Climate  Initiative  (WCI)                 xAcknowledgements  I  offer  my  gratitude  to  the  faculty  at  UBC,  who  both  permitted  and  inspired  this  far-­‐reaching  thesis.  I  offer  particular  thanks  to  Dr.  Thomas  Sullivan,  for  showing  me  value  in  the  philosophy  of  science,  not  just  science  itself.  I  am  indebted  to  my  committee  members,  Dr.  David  Cohen  and  Dr.  Paul  McFarlane,  for  their  patience  and  insights.  My  supervisor,  Dr.  Robert  Kozak,  invited  me  to  take  a  long  and  unconventional  path  and  tolerated  my  ever-­‐evolving  interests.  His  penetrating  insights,  high  standards  and  incredible  work  ethic  were  a  great  gift  and  a  tremendous  inspiration.      I  also  need  to  thank  NSERC,  the  Value  Chain  Optimization  Network,  and  the  UBC  Faculty  of  Forestry  for  their  generous  support  throughout  my  matriculation.      Special  thanks  are  reserved  for  my  parents.  To  my  mother,  for  never  letting  me  forget  that  I  was  expected  to  succeed.  And  to  my  father,  Dr.  Gary  Bull,  for  providing  the  perspective  and  encouragement  required  to  get  things  done.                     xiDedication  For  Grandma.  For  Sunny.  For  Mom.  For  Dad.  For  Dulce.  (Oh,  and  Crowley  too.)                   1Chapter  1: Introduction  “Please  consider  the  environment  before  printing  this  email.”  We’ve  all  seen  the  phrase  happily  applied  to  the  bottom  of  emails,  reminding  us  to  refrain  from  printing.  It  is  an  empowering  eight  words,  letting  the  reader  know  that,  if  they  do  the  right  thing,  the  environment  will  thank  them.  It  also  suggests  that  printing  is  bad  for  the  environment,  but  emails  aren’t  something  I  should  worry  about.  After  reading  the  phrase  countless  times,  I  have  finally  decided  to  heed  its  advice.      This  dissertation  takes  the  phrase  seriously,  but  rather  than  simply  ask  whether  I  ought  to  print  an  email  or  not,  I  have  taken  an  academic  approach.  I  want  to  explore  the  big  questions  the  phrase,  pondered  closely,  reveals.  What  does  it  mean  to  “consider  the  environment”?  I  need  to  define  and  understand  sustainability.  Why  is  printing  on  paper  implicitly  criticized?  I  need  to  understand  the  sustainability  of  paper.  Why  is  paper  being  compared  to  an  email?  I  need  to  describe  the  shift  from  one  media  type  to  another,  and  its  environmental  consequences.      If  these  questions  sound  too  big  for  the  scope  of  a  PhD,  it  is  because  they  are.  But  they  have  been  winnowed  down  to  research  questions  and  scientific  approaches  that  will  strengthen  an  understanding  of  what  sustainability  means  in  the  context  of  media  consumption.  The  environment  will  be  considered  from  the  perspective  of  industry,  consumers,  and  academic  research.  The  research  objective  of  this  dissertation  is  to  consider  the  environmental  footprints  of  digital  and  paper  media  from  three  perspectives  –  industry,  consumers,  and  academic  comparisons  –  in  order  to  elucidate  the  concept  of  sustainability.      1.1 BACKGROUND  AND  LITERATURE  REVIEW  In  order  to  properly  assess  the  environmental  impacts  of  media  –  either  digital  or  printed  –  I  need  to  begin  with  a  thorough  grounding  in  the  literature  and  establish  working  definitions  of  the  two  major  topics  this  thesis  examines:  sustainability  and  media.  In  this  section,  I  set  out  to  achieve  two  things:  to  explore  definitions  and  manifestations  of  sustainability,  and  to  identify  and  contextualize  the  changes  in  media  consumption  underway  and  potential  environmental  impacts.  I  call  these  two  trends,  “The  Sustainability  Shift”  and  the  “The  Media  Shift”,  respectively,  terms  I  employ  throughout  the  dissertation.                   21.1.1 The  Sustainability  Shift1  For  such  a  common  word,  sustainability  is  a  difficult  concept  to  define.  At  its  core,  it  implies  doing  something  better  (or  at  least  less  badly).  For  this  dissertation,  I  need  to  define  sustainability  in  light  of  my  broad  research  objective  –  to  understand  the  sustainability  of  media  consumption.  Due  to  the  nature  of  media,  I  must  ground  an  understanding  of  sustainability  in  the  world  of  industry.  It  takes  businesses,  supply  chains,  and  consumers  all  connecting  together  in  order  to  produce,  distribute,  and  consume  media.      With  the  need  to  understand  business  and  sustainability  established,  this  section  describes  the  origins,  evolution,  and  future  directions  of  business  and  sustainability.  My  focus  is  on  environmental  sustainability,  not  social  or  economic  sustainability.  While  these  latter  two  issues  are  important,  economic  sustainability  is  well  understood  and  already  considered  thoroughly  by  business  and  its  profit  motive.  Social  sustainability  –  which  considers  human  rights  and  the  concept  of  inter-­‐generational  equity  –  is  important,  but  sufficiently  distinct  from  environmental  sustainability  to  make  its  inclusion  here  inappropriate.    1.1.1.1 Definitions  of  Sustainability  In  1970,  the  economist  Milton  Friedman  wrote  (Friedman,  1970)  a  seminal  article  called  “The  Social  Responsibility  of  Business  is  to  Increase  its  Profits.”  The  role  of  business  has  since  changed.  Business  can  no  longer  relentlessly  pursue  profits  without  considering  the  environmental  and  social  implications.  A  confluence  of  drivers  has  shaped  this  evolution,  including  increasing  population  (and  consumption)  levels  and  the  emergence  of  climate  change  as  a  global  environmental  issue  have  shaken  the  corporate  sector  and  undermined  public  trust  in  private  industry.  In  2010,  the  Nobel  Laureate  Joseph  Stiglitz  revisited  Friedman’s  narrow  doctrine,  and  suggested,  “We  should  think  about  how  we  can  create  a  global  economic  architecture  which  works  better,  for  more  people,  in  a  more  sustainable  way”  (United  Nations,  2009).                                                                                                                                  1  1  A  longer  version  of  this  section  (1.1.1.)  appears  in  the  following:  J.G.  Bull  and  R.A.  Kozak.  (In  Press).  The  Means  and  Meaning  of  Sustainability:  The  Role  of  Forestry  in  a  Responsible  World.  In:  S.  D’Amours,  M.  Ouhimmou,  J.  Audy  and  Y.  Feng,  ed.,  Forest  Value  Chain  Optimization  and  Sustainability,  1st  ed.  Boca  Raton:  CRC  Press/Taylor  &  Francis.                 31.1.1.1.1 Academic  Definitions  of  Sustainability  A  founding  definition  of  sustainability  comes  from  the  Brundtland  Commission,  which  described  sustainability  –  more  specifically,  sustainable  development  –  as  meeting  “the  needs  of  the  present  without  compromising  the  ability  of  future  generations  to  meet  their  own  needs”  (United  Nations,  1987).  This  definition  evolved,  with  the  concept  of  sustainability  as  a  ‘three-­‐legged  stool’  gaining  popular  adoption  (Newport  et  al.,  2003).  Here,  an  ideal  outcome  considers  economic,  social,  and  environmental  concerns  together,  with  sustainability  occurring  at  the  intersection  of  these  three  values.  A  modified  version  of  this  idea  is  the  embedded  circles  approach  (Lozano,  2008),  which  considers  the  economy  a  subset  of  society,  which  in  turn  is  a  subset  of  the  natural  environment.  This  approach  offers  the  natural  environment  as  the  most  important  variable  worth  considering  (see  Figure  1).    A  variation  of  this  definition  comes  from  the  discipline  of  ecological  economics.  Ecological  economics  is  an  effort  to  promote  understanding  between  ecologists  and  economists.  It  differs  from  neoclassical  economic  approaches  in  that  neoclassical  models  are  defined  by  achieving  an  equilibrium  and  final  state  that  is  independent  of  the  path  taken  (Goodland  and  Ledec,  1987).  Growth  in  the  neoclassical  sense  is  a  function  of  accumulated  savings,  capital  investment,  and  technological  progress.  Consequently,  growth  is  theoretically  exponential  and  limitless.  By  contrast,  ecological  economics  models  are  subject  to  an  exogenous  limit:    the  carrying  capacity  of  the  planet.  Economic  models  are  embedded  in  the  environment,  and  these  models  explicitly  address  the  interdependence  of  human  economies  and  natural  ecosystems  over  time  and  space  (Wackernagel  and  Rees,  1997).  Thus,  sustainability  is  conveyed  as  ‘embedded  circles’  where  social  and  economic  concerns  are  bounded  by  the  broader  context  of  the  environment.  In  this  view,  society  and  economic  systems  are  subsets  of  the  environment,  limited  by  a  finite  amount  of  natural  capital.                 4Figure  1:  Schematic  definitions  of  sustainability    While  these  definitions  are  conceptual,  elegant,  and  inclusive,  they  do  not  necessarily  offer  the  specificity  required  by  different  actors,  such  as  governments,  consumers,  corporations,  and  civil  society.  The  traditional  academic  approach  of  creating  distinct  and  isolated  disciplines  and  reducing  questions  to  testable  hypotheses  is  ill-­‐suited  to  ask  questions  and  provide  answers  about  sustainability.  What  type  of  academic  is  best  suited  to  address  issues  of  sustainability  –an  engineer,  an  ecologist,  an  economist,  a  sociologist,  or  political  scientist?  Each  would  approach  sustainability  in  very  different  ways,  none  more  correct  than  the  others.    Seager  (2008)  suggests  that  academics  are  not  unaware  of  this  mismatch  between  traditional  research  structures  and  urgent  research  questions.  To  that  end,  two  new  academic  disciplines  are  emerging  as  a  means  of  grappling  with  the  complex  and  multifaceted  nature  of  sustainability:  industrial  ecology  and  ecosystem  health.  Seager,  in  identifying  this  mismatch  between  definitions  and  research  capacity,  follows  upon  the  work  of  Mebratu  (1998,  p.493)  who  noted  that  definitions  of  sustainability  “focus  on  specific  elements  while  failing  to  capture  the  whole  spectrum.”  Efforts  to  develop  a  theory  of  sustainability  has  led  to  concepts  and  definitions  that  benefit  specific  groups  and  interests,  but  do  not  reflect  holistic  thinking  on  the  subject  (Mebratu,  1998).    Industrial  ecology  is  focused  on  the  interactions  between  ecological  and  industrial  systems  (Graedel  and  Allenby,  1995).  It  is  founded  on  the  idea  that  much  can  be  learned  from  natural  systems  to  improve  the  environmental  footprint  of  industrial  systems  (Erkman,  1997).  Given  that  industrial  systems  have  a               5relatively  brief  history,  at  least  compared  to  ecosystems,  looking  at  the  structures  and  successes  of  nature  can  provide  insights  into  how  to  devise  more  effective  and  sustainable  systems  (Seager,  2008).    Figure  2  shows  three  types  of  systems  commonly  discussed  in  the  discipline  of  industrial  ecology.  The  first,  Type  I,  is  an  open  loop,  with  energy  and  materials  flowing  in,  being  processed,  and  flowing  out.  It  assumes  an  unlimited  availability  of  resources,  and  an  unlimited  ability  for  the  planet  to  absorb  the  impacts  of  industrial  activities.  Early  industrial  systems  were  similar  to  this  model.  Type  II  describes  most  existing  industrial  systems  as  they  occur  today.  There  are  limited  supplies  of  materials  and  energy,  and  industry  cannot  pollute  freely  without  incurring  costs.  Some  industries  might  be  a  ‘weak’  Type  II,  with  minimal  resource  conservation  and  waste  reduction  efforts,  while  others  could  be  considered  a  ‘strong’  Type  II,  where  both  economic  and  environmental  motives  exist  for  managing  resource  use  and  waste  streams.  A  completely  closed  loop  Type  III  system  is  impossible  for  industry,  as  total  resource  conservation  is  unachievable.  However,  it  is  something  that  industry  can  aspire  to.  Ecological  systems  are  Type  III  systems,  as  nature  has  only  one  input  (solar  energy),  waste  does  not  exist  as  every  material  is  eventually  reused  by  the  system  (Gradel  and  Allenby,  1995;  Erkman,  1997).    Figure  2:  Industrial  ecology  types                   6Industrial  ecology  is  more  than  simply  envying  the  extreme  efficiency  of  nature.  Kalundborg,  Denmark  is  an  example  of  industrial  ecology  that  has  garnered  substantial  academic  inquiry  (see  Ehrenfeld  and  Gertler,  1997  for  a  detailed  account).  A  variety  of  industrial  actors  evolved,  without  any  master  plan,  to  integrate  their  waste  and  energy  flows  to  become  more  environmentally  benign.  A  coal  plant  sells  steam  to  a  variety  of  users,  including  local  households.  A  pharmaceutical  factory  installed  a  two-­‐mile  steam  pipe  rather  than  replacing  their  existing  boilers,  an  investment  that  paid  for  itself  in  two  years.  The  coal  plant  was  also  required  by  regulators  to  install  sulfur  dioxide  scrubbers  at  a  cost  of  $US  115  million.  Fortunately,  the  coal  plant  was  able  to  sell  gypsum,  a  byproduct  of  the  scrubbing  process,  to  a  local  plasterboard  plant.  All  of  these  connections  between  industrial  actors  minimize  the  waste  of  energy  and  materials.  This  symbiosis  was  not  mandated  by  regulations,  although  some  regulatory  imperatives  helped  foster  connections  that  would  not  have  otherwise  existed.  Further,  and  perhaps  more  importantly,  these  decisions  made  a  good  deal  of  economic  sense.    While  industrial  ecology  is  now  emerging  as  a  bona  fide  area  of  inquiry,  ecosystem  health  is  much  less  formalized  and  the  most  nascent  of  the  academic  disciplines  focused  on  sustainability  (Rapport  et  al.,  1999).  It  advances  the  goal  of  assessing  the  health  of  an  entire  ecosystem,  not  just  the  health  of  individual  organisms  or  species.  Further,  it  suggests  that  human  health  is  dependent  on  healthy  ecosystems,  and  that  the  interaction  between  these  two  systems  is  vital.  Like  ecological  economics,  it  considers  a  healthy  ecosystem  to  be  a  necessary  factor  of  production  in  a  functioning  society  and  economy.  There  are  three  key  concepts  that  have  shaped  research  in  ecosystem  health,  and  thus,  a  conception  of  sustainability:  vigour  –a  measure  of  total  activity;  organization  –  a  measure  of  the  quality  and  diversity  of  interactions  between  different  parts  of  the  system;  and  resilience  –  the  ability  of  the  system  to  recover  from  injury  (Seager,  2008;  Costanza  and  Mageau,  1999).    1.1.1.1.2 Corporate  Definitions  of  Sustainability  The  corporate  world  offers  its  own  versions  of  sustainability,  with  corporate  social  responsibility  (CSR)  being  the  most  common  tool  for  pursuing  sustainability  objectives.  CSR  requires  a  firm  to  think  beyond  legal  and  profit  imperatives  by  considering  its  impacts  on  society  and  the  environment  (van  Merrewijk,  2003).  By  no  means  is  CSR  a  simple  concept;  the  literature  identifies  nearly  40  competing  and  complimentary  definitions  (see  Dahlsrud,  2008).  CSR  can  be  referred  to  by  many  names,  with  the  terms  triple  bottom  line  accounting,  stakeholder  management,  corporate  citizenship,  or  corporate  responsibility  being  used  interchangeably.  Auld  et  al.  (2008a)  describe  CSR  efforts  as  either  ‘win-­‐win’               7solutions  or  ‘win-­‐lose’  solutions.  Win-­‐win  solutions  involve  internal  changes  that  are  socially  and/or  environmentally  beneficial,  but  also  serve  to  increase  profit.  In  other  words,  this  is  the  ‘low  hanging  fruit’  of  CSR.  Win-­‐lose  solutions  are  more  complex,  as  a  change  that  provides  environmental  or  social  benefits  may  decrease  profits.  As  a  result,  the  structures  that  influence  profits  (such  as  government  policy  and  consumer  preferences)  would  need  to  be  adapted  to  reward  the  new,  but  initially  less  profitable,  solution.  If  no  such  changes  take  place,  the  win-­‐lose  solution  eventually  fails.    Firms  can  be  proactive  in  trying  to  develop  new  markets  that  reward  what  are  initially  win-­‐lose  efforts,  or  they  can  anticipate  changes  in  government,  consumer,  or  civil  society  behaviour  that  will  eventually  result  in  win-­‐lose  solutions  becoming  profitable.  However,  when  win-­‐win  or  win-­‐lose  solutions  fail  to  generate  profits  in  the  long-­‐term,  CSR  efforts  can  result  in  a  competitive  disadvantage,  and  become  marginalized  rather  than  transformative  in  influencing  firm  behaviour  (McWilliams  and  Siegel,  2001).    There  is  a  danger  that  CSR  efforts  might  bias  a  conception  of  sustainability.  Does  CSR  mean  that  a  company  will  transform  to  become  truly  more  sustainable?  Or  does  a  company  engage  in  CSR  in  the  hopes  that  some  of  its  industrial  practices  can  be  viewed  –  without  much  corporate  effort  –  as  being  sustainable?  Another  important  idea  is  that,  ‘you  can’t  make  an  omelet  without  breaking  eggs.’  Businesses  will  always  have  an  impact  on  the  environment,  and  it  is  prudent  to  remember  this  when  considering  some  of  the  more  utopian  ideals  surrounding  definitions  of  sustainability.    1.1.1.1.3 The  Spectrum  of  Sustainability    Sustainability  manifests  differently  within  academic  and  corporate  contexts.  But  in  many  ways,  the  multitudes  of  definitions  complement  one  another,  offering  guidance  at  practical  and  abstract  levels.  A  common  element  is  that  sustainability  is  a  function  of  both  time  and  space.  For  example,  the  time  dimension  of  sustainability  is  revealed  when  describing  the  use  of  a  reusable  mug  instead  of  a  disposable  cup,  or  altering  an  entire  economy  to  rely  on  the  production  of  services  rather  than  goods,  or  reducing  consumption  as  a  means  of  achieving  intergenerational  equality.  The  space  dimension  refers  to  the  fact  that  sustainability  initiatives  can  occur  at  the  scale  of  a  single  consumer  or  at  the  scale  of  a  factory,  supply  chain,  or  country.  When  defining  sustainability,  we  must  embrace  diversity,  rather  than  struggle  to  find  a  unifying  articulation.  Seager  (2008,  p.447)  grapples  with  this  tension  between  inclusivity  and  specificity  and  arriving  at  a  definition  that  allows  different  actors  to  consider  sustainability  relative  to  their  own  interests:               8“Sustainability  might  best  be  defined  as  an  ethical  concept  that  things  should  be  better  in  the  future  than  they  are  at  present.  Like  other  ethical  concepts  such  as  fairness  or  justice,  sustainability  is  best  interpreted  conceptually  rather  than  technically.”    Seager  (2008)  takes  this  notion  that  sustainability  is  an  ethical  concept,  and  produces  spectrum  of  sustainability  that  reflects  the  different  dimensions  of  time  and  space  over  which  sustainable  outcomes  can  occur.  On  one  side  is  a  version  of  sustainability  that  is  static.  It  is  defined  by  the  maintenance  of  the  status  quo,  which  to  some  actors  (such  as  law  enforcement)  would  be  a  sustainable  outcome.  Next  is  steady-­‐state  sustainability,  which  is  defined  by  reliability.  Here,  the  ability  for  a  system  to  sustain  the  same  function  over  the  long-­‐term  is  prioritized.  Further  along  is  dynamic  sustainability,  where  the  ability  of  a  system  to  adapt,  regrow,  or  evolve  in  light  of  changing  circumstances  is  prioritized.  To  the  far  side  is  episodic  sustainability,  which  is,  in  essence,  a  version  of  Schumpeter’s  ‘creative  destruction’  (Schumpeter,  1942).  In  this  view,  sustainability  is  defined  as  an  evolving  system  passing  through  multiple  states.      1.1.1.2 Drivers  of  Sustainability  It  is  clear  the  sustainability  has  a  range  of  meanings  and  interpretations.  It  should  come  as  no  surprise  that  a  wide  variety  of  factors  motivate  the  adoption  of  sustainability  practices.  Each  actor  along  any  given  supply  chain  is  motivated  by  their  own  definition  of  sustainability  and  self-­‐interests.  That  being  said,  many  common  themes  are  emerging  within  this  burgeoning  domain  and,  in  the  first  part  of  this  section,  I  consider  the  main  ‘drivers’  of  corporate  sustainability:  governments;  investors;  eco-­‐efficiency;  the  marketplace;  advocates;  and  climate  change.  These  categories  emerged  from  an  extensive  literature  review.  While  other  categories  certainly  exist,  these  provide  a  framework  that  captures  both  the  breadth  and  the  complexity  of  sustainability  and  its  manifestations.      1.1.1.2.1 Sustainability  Driver:  Governments  Governments  motivate  sustainability  by  a  variety  of  means  and  for  a  variety  of  reasons.  Governments  are  ‘stewards’  of  the  environment;  they  are  responsive  to  the  demands  of  voters  who,  at  times,  prioritize  environmental  issues  (Kraft,  2001).  Governments  are  pressured  by  environmental  advocates  and  corporations,  and  attempt  to  strike  a  policy  balance  between  environmental  conservation  and  the  need  for  economic  growth.  How,  when,  and  why  a  government  pushes  for  corporate  environmental  sustainability  is  dependent  on  the  definition  of  sustainability  employed  and  the  environmental  issue  at               9hand.  Graedel  and  Allenby  (2003)  provide  a  thorough  accounting  of  how  governments  influence  corporate  sustainability.  They  suggest  that  government  action  rests  on  a  spectrum.  On  one  hand,  governments  can  encourage  specific  behaviour,  by  setting  standards,  implementing  bans,  or  mandating  the  use  of  best  available  technology.  At  the  other  end  of  the  spectrum,  governments  can  act  by  correcting  a  lack  of  information,  using  product  labels,  life  cycle  accounting,  and  awards  and  recognition  to  encourage  sustainable  outcomes.  In  the  middle  rests  governmental  actions  which  serve  to  change  incentives,  either  by  introducing  or  removing  subsidies,  taxes,  or  regulatory  reform.      Governments  have  motivated  environmental  sustainability  in  the  forest  sector  in  a  variety  of  ways.  In  British  Columbia  (B.C.),  emitters  are  required  to  pay  a  tax  on  their  carbon  emissions  (Schnoor,  2014).  This  can  be  a  significant  cost  for  major  emitters,  such  as  pulp  and  paper  mills  (Bumpus,  2014).  The  provincial  government  has  committed  to  becoming  carbon  neutral,  providing  opportunities  for  forest  companies  to  sell  credits  to  the  Pacific  Carbon  Trust.2  B.C.  is  also  part  of  the  Western  Climate  Initiative,  which  could  evolve  into  a  major  market  for  trading  carbon  emissions  in  North  America.  The  federal  government  in  Canada  has  subsidized  the  retrofitting  of  pulp  and  paper  facilities  to  produce  bioenergy  and  biochemicals  as  a  means  of  participating  in  the  new  so-­‐called  ‘green  economy’  (Laan  et  al.,  2009).  In  Europe,  governments  have  implemented  sustainable  a  variety  of  procurement  policies,  including  policies  to  ensure  that  the  government  only  buys  forest  products  from  legally  verified  suppliers  (Commission  of  the  European  Communities,  2003).    1.1.1.2.2 Sustainability  Driver:  Investors  Investors  –  those  who  provide  capital  to  companies  or  hold  corporate  stock  –  have  been  a  key  driver  of  corporate  sustainability.  But  to  characterize  all  investors  as  drivers  of  sustainability  would  be  untrue.  Many  are  focused  on  only  profits  and  quarterly  returns.  However,  large  institutional  investors,  like  pension  funds,  have  taken  a  different  approach,  adopting  ESG  (environmental,  social,  and  governance)  metrics  in  their  investment  strategies  (Flatz  et  al.,  2001).    The  United  Nations  Environment  Programme’s  (UNEP)  Finance  Initiative  is  a  group  of  over  seventy  investors  that  have  adopted  Principles  for  Responsible  Investing.  The  investors  describe  their  motives  for  considering  ESG  issues  as  threefold  (UNEP,  2006).  First  and  foremost,  ESG  variables  can  help                                                                                                                            2  The  Pacific  Carbon  Trust  is  currently  being  wound-­‐down,  with  the  provincial  government  folding  its  services  back  into  the  provincial  Environmental  Agency  (Moore,  2014).               10investors  make  more  money.  Their  hypothesis  is  that,  over  time,  integrating  thorough  and  systematic  reviews  of  ESG  issues  will  result  in  better  financial  performance  (Tang  et  al.,  2012).  Second,  some  investors  consider  ethical  values  in  their  investment  decisions,  and  are  happy  to  ignore  marginal  impacts  that  ESG  criteria  might  have  on  financial  performance.  Finally,  channeling  investment  flows  in  accordance  with  ESG  issues  will  support  the  long-­‐term  sustainable  development  of  the  global  economy.    The  authors  of  the  UNEP  report  dismiss  the  notion  that  investors  are  unfamiliar  with  the  financial  impacts  of  environmental  or  social  risk,  offering  as  evidence  the  financial  liabilities  associated  with  asbestos  in  building  materials  in  the  1980s  and  1990s,  or  the  costs  faced  by  Union  Carbide  following  the  Bhopal  explosion.  Speaking  to  agnostic  or  skeptical  investors  who  think  ESG  issues  have  a  tenuous  connection  with  financial  performance,  the  authors  suggest  (UNEP,  2006,  p.4):  “Investors  who  are  not  sure  what  to  do  would  do  well  to  refer  to  Pascal’s  wager  concerning  the  existence  of  God:  make  believe  as  if  the  hypothesis  is  true,  thus  at  little  or  no  cost  avoiding  the  worst-­‐case  scenario.”    Another  example  of  investors  encouraging  environmental  sustainability  comes  from  the  Carbon  Disclosure  Project  (CDP).  The  CDP  is  a  non-­‐profit  supported  by  534  investors  with  assets  totaling  $US  64  trillion.  The  CDP  administers  an  annual  survey  to  over  4700  corporations,  but  focuses  its  efforts  on  the  so-­‐called  ‘Global  500’,  a  group  of  companies  that  collectively  accounts  for  11%  of  annual  greenhouse  gas  emissions  (CDP,  2010).  Of  these  top  500  companies,  82%  responded  to  the  survey  in  2010,  the  highest  rate  since  the  inaugural  survey  in  2003.  In  their  annual  report  for  2010,  the  CDP  identifies  several  important  trends.  They  see  the  demand  for  primary  corporate  climate  change  data  to  be  growing,  with  platforms  like  Bloomberg  Finance  now  offering  carbon  data  alongside  financial  data.  Previously  unresponsive  sectors,  such  as  shipping  and  transportation,  have  now  begun  to  share  carbon  data  (CDP,  2010).  CDP  also  sees  a  shift  in  corporate  perceptions  of  climate  change,  with  more  respondents  in  2010  identifying  climate  change  as  an  opportunity  (average  of  9  out  of  10  respondents)  than  a  risk  (average  of  8  out  of  10  respondents).  While  the  CDP  is  not,  by  itself,  a  driver  of  sustainability,  it  is  the  result  of  pressure  from  investors  to  identify  environmental  risks  in  their  investment  portfolios,  which  is  a  strong  driver  of  disclosure  and  transparency  on  environmental  footprints.                     111.1.1.2.3 Sustainability  Driver:  Eco-­‐Efficiency  Eco-­‐efficiency  is  the  most  obvious  reason  for  a  corporation  to  consider  sustainability.  At  its  core,  eco-­‐efficiency  involves  “producing  and  delivering  goods  while  simultaneously  reducing  the  ecological  impact  and  use  of  resources.”  (Molina-­‐Azorín  et  al.,  2009,  p.1082).  From  this  perspective,  the  generation  of  pollution  is  seen  as  inefficient.  However,  the  financial  benefits  of  eco-­‐efficiency  are  not  always  immediate  or  evident,  and  its  pursuit  could  render  a  firm  temporarily  or  permanently  uncompetitive.    Ambec  and  Lanoie  (2008)  identify  ways  in  which  eco-­‐efficiency  can  help  businesses  to  improve  their  bottom  lines,  for  instance,  by  increasing  their  levels  of  risk  management  and  external  stakeholder  engagement  or  by  decreasing  their  costs  of  materials,  energy,  and  services.  They  also  identify  three  opportunities  afforded  by  eco-­‐efficiency  to  directly  increase  revenue  streams:  through  better  access  to  certain  markets;  through  product  differentiation;  and  through  selling  pollution-­‐control  technology.  This  variety  of  opportunities  is  demonstrative  of  how  eco-­‐efficiency  can  be  implemented  in  a  manner  that  makes  good  fiscal  sense  to  businesses.  Eco-­‐efficiency  can  also  involve  attempts  to  reduce  the  transportation  infrastructure  required  to  deliver  a  product  to  market.  This  is  most  prevalent  in  the  food  industry  where  ‘food  miles’  are  used  to  measure  the  environmental  impact  of  delivering  food  to  market  (Engelhaupt,  2008).      Based  on  a  review  of  32  case  studies,  Molina-­‐Azorín  et  al.  (2009)  present  a  three-­‐stage  process  that  relates  environmental  management  to  financial  performance.  Functional  specialization  is  the  first  stage,  and  involves  a  company  reacting  to  the  pressures  of  environmental  regulation.  Internal  integration,  the  second  stage,  involves  management  developing  and  monitoring  corporate  objectives  on  environmental  management.  The  third  stage  is  external  integration.  Here,  a  firm  incorporates  environmental  performance  into  its  overall  business  strategy.  Many  of  the  opportunities  identified  above  by  Ambec  and  Lanoie  (2008)  result  from  this  three-­‐stage  process.  In  pursuing  environmental  goals,  new  markets  of  consumers  who  prioritize  ecological  issues  open  up  and  firms  can  become  leaders  in  their  sector.  As  a  result,  they  may  develop  technologies  or  processes  that  put  them  in  an  advantageous  position,  or  they  may  be  able  to  patent  and  resell  their  innovations,  thereby  gaining  first  mover  advantage.  Molina-­‐Azorín  et  al.  (2009)  find  that  firms  who  achieve  a  positive  relationship  between  environmental  management  and  financial  performance  are  able  to  invest  more  in  their  environmental  strategies,  resulting  in  a  positive  feedback  loop.                   121.1.1.2.4 Sustainability  Driver:  Marketplace  There  are  primarily  two  market  forces  that  can  drive  corporate  sustainability:  business-­‐to-­‐business  (B2B)  markets  and  consumer  markets.  B2B  markets  influence  corporate  sustainability  when  one  company  seeks  suppliers  who  have  integrated  environmental  thinking  into  their  business  model  (Mariadoss  et  al.,  2011).  For  example,  the  Home  Depot  implemented  a  policy  where  the  purchase  of  certified  wood  is  prioritized  (Home  Depot,  2011).  Wal-­‐Mart  has  undertaken  similar  responsible  purchasing  policies.  Wal-­‐Mart  has  developed  a  widely  praised  supplier  assessment  mandatory  for  anyone  who  intends  to  sell  goods  at  a  Wal-­‐Mart  store  (Wal-­‐Mart,  2011).  Providing  information  on  carbon  emissions,  water  usage,  and  plans  to  increase  eco-­‐efficiency,  amongst  other  things,  is  required.  Wal-­‐Mart  is  in  the  process  of  requiring  its  largest  suppliers  to  also  conduct  the  same  assessment  of  their  own  suppliers.  As  a  result,  Wal-­‐Mart’s  environmental  priorities  are  trickling  up  the  supply  chain,  influencing  companies  who  might  not  even  sell  directly  to  Wal-­‐Mart.  Given  the  purchasing  power  of  Wal-­‐Mart  these  efforts  hold  enormous  potential.    Consumers  can  also  drive  corporate  sustainability.  Price  being  equal,  consumers  generally  prefer  to  purchase  a  ‘green’  product  (Wilk,  2001)),  and  some  are  willing  to  pay  a  premium  for  more  sustainable  products  (Ottman,  2011).  The  rapid  rise  of  organic  food  reflects  an  increasing  consumer  awareness  of  environmental  issues.  Major  consumer  products  manufacturers,  such  as  Coca-­‐Cola,  are  beginning  to  promote  their  green  initiatives,  in  this  case,  promising  to  make  bottles  that  contain  at  least  30%  plant-­‐based  materials  (Houpt,  2011).  Information  and  communications  technology  companies  are  offering  mobile  phones  with  plastic  casing  made  from  recycled  water  bottles  (German,  2009).  Hygienic  paper  products  with  only  recycled  content  have  been  developed  and  are  capturing  market  share  (Tissue  World,  2010).  There  are  similar  examples  from  almost  every  sector  and  product  type.    But  there  is  a  risk  that  corporate  efforts  to  increase  demand  for  green  products  offer  little  environmental  benefit  or,  even  worse,  are  misleading  (Athanasiou,  1996).  This  ‘greenwashing’  is  not  surprising  given  the  relative  immaturity  of  markets  for  environmentally  responsible  goods.  In  the  future,  however,  it  will  be  harder  for  companies  to  mislead  consumers  (Parguel  et  al.,  2011).  The  Internet  and  social  media  are  creating  savvy  consumers  who  facilitate  the  rapid  dissemination  of  embarrassing  information  on  corporate  practices.  Green  marketing  experts  suggest  that,  in  this  transparent  information  age,  CSR  should  be  authentic  or  not  attempted  at  all  (Ottman,  2011).                 131.1.1.2.5 Sustainability  Driver:  Advocates  Advocates  are  individuals  or  organizations  who  attempt  to  influence  corporate  behaviour,  with  environmental  non-­‐governmental  organizations  (ENGOs)  serving  as  the  most  prominent  example.  Other  advocates  might  include  local  communities  that  are  impacted  by  a  firm’s  activities  or  academics  whose  research  leads  them  to  believe  that  a  firm’s  behaviour  should  be  improved.  Advocates  will  exert  influence  by  different  means  at  various  levels  on  a  variety  environmental  issues  (Stevis  and  Assetto,  2001).    If  there  is  a  corporation  creating  an  environmental  impact,  there  is  likely  an  ENGO  advocating  for  its  amelioration  (Jad,  2007).  They  do  so  by  a  variety  of  means  (Obar  et  al.,  2012):  connecting  directly  with  a  firm  to  express  their  concerns  and  perhaps  offer  solutions;  lobbying  relevant  government  agencies  to  suggest  regulatory  changes  that  address  the  situation;  or  communicating  with  the  public  to  raise  awareness.  Some  ENGOs  have  a  more  adversarial  and  combative  relationship  with  the  corporate  world,  publicly  shaming  a  company  by  releasing  details  of  an  environmental  harm.  Other  ENGOs,  in  contrast,  work  directly  with  corporations  to  help  improve  their  environmental  performance,  lending  authenticity  and  credibility  to  CSR  efforts.  ENGOs  themselves  admit  that  working  with  the  private  sector  is  a  potent  opportunity  to  promote  sustainability.  Speaking  about  his  relationship  with  the  CocaCola  Company,  the  CEO  of  WWF  Canada  stated  the  following  (Houpt,  2011,  para.  8):  "Coke  is  the  No.  1  purchaser  of  aluminum  on  the  face  of  the  earth  -­‐  which  is  one  of  the  most  carbon-­‐intensive  commodities.  The  No.  1  purchaser  of  sugar  cane.  The  No.  3  purchaser  of  citrus.  The  second-­‐largest  purchaser  of  glass,  and  the  fifth-­‐largest  purchaser  of  coffee.  We  could  spend  50  years  lobbying  75  national  governments  to  change  the  regulatory  framework  for  the  way  these  commodities  are  grown  and  produced.  Or  these  folks  at  Coke  could  make  a  decision  that  they're  not  going  to  purchase  anything  that  isn't  grown  or  produced  in  a  certain  way  –  and  the  whole  global  supply  chain  changes  overnight.  And  that  in  a  nutshell  is  why  we're  in  a  partnership."    Local  communities  affected  by  a  company’s  environmental  footprint  can  similarly  advocate  for  change.  How  and  where  they  advocate  will  depend  on  the  particular  issue,  their  capacity,  and  the  responsiveness  of  government  to  community  concerns.  Like  ENGOs,  what  is  important  from  the  perspective  of  CSR  is  that  advocacy  is  likely  to  occur  for  any  number  of  issues,  at  many  levels,  and  can  be  combative  or  cooperative  (Lund-­‐Thomsen  and  Wad,  2014).  How  a  corporation  responds  to  advocacy  is               14perhaps  indicative  of  how  seriously  it  has  integrated  CSR  practices  into  its  business  strategy.  Those  that  have  taken  the  steps  to  measure  and  manage  their  environmental  footprint,  and  have  done  so  in  a  methodologically  rigorous  and  transparent  way,  will  be  better  positioned  to  engage  with  advocates  (Doh  and  Guay,  2004).    1.1.1.2.6 Sustainability  Driver:  Climate  Change  Climate  change  and,  by  extension,  carbon,  deserve  particular  attention  as  a  driver  of  sustainability.  Climate  change  is  a  unique  environmental  issue  because  it  is  global  in  scale  and  has  dire  implications  (IPCC,  2007).  It  is  unique  politically  because  so  many  government  bodies  at  the  international,  national,  provincial,  and  municipal  levels  have  put  forth  regulations  and  set  targets  with  the  goal  of  reducing  carbon  emissions.  Corporations,  perhaps  seeing  the  writing  on  the  wall,  have  embraced  carbon  as  a  major  environmental  issue  (SwissRe,  2007).  Given  that  carbon  emissions  are  somewhat  analogous  with  energy  use,  reducing  carbon  often  equates  with  reducing  energy  costs.  Reducing  emissions  can,  therefore,  be  a  classic  ‘win-­‐win’  CSR  initiative  (Stanny  and  Ely,  2008).    Carbon  simply  being  ubiquitous  does  not  make  it  a  driver  of  environmental  sustainability.  The  real  driver  is  the  likelihood  of  carbon  as  a  priced  commodity.  Governments,  advised  by  scientists  and  advocates  that  carbon  emissions  should  be  reduced,  have  devised  schemes  to  put  a  price  on  carbon  (Garnaut,  2008).  In  the  European  Union,  the  European  Trading  System  governs  the  buying  and  selling  of  emission  allowances  (Europa,  2005).  International  agreements  have  tried,  thus  far  failing,  to  emulate  this  system.  Regional  approaches,  such  as  the  Western  Climate  Initiative  or  the  Regional  Greenhouse  Gas  Initiative,  have  taken  steps  to  price  carbon  (Western  Climate  Initiative,  2009;  RGGI,  2009).  Some  jurisdictions,  like  British  Columbia,  have  imposed  a  carbon  tax.  All  of  these  efforts  result  in  one  obvious  conclusion:  corporations  anticipate  having  to  pay  for  carbon  (Kolk  et  al.,  2008).  The  prospect  of  a  new  expense  embedded  in  all  business  transactions  has  been  remarkably  effective  in  driving  businesses  to  consider  their  environmental  footprints.    The  monetization  of  carbon  is  what  economists  consider  internalizing  a  cost.  Economic  systems  often  fail  at  internalizing  environmental  costs,  and  carbon  is  the  first  attempt  at  a  global  scale  (Lohmann,  2009).  It  should  not  be  a  surprise  that  reaching  an  international  consensus  on  how  to  price  carbon  is  an  elusive  goal.  The  lack  of  action  at  the  international  level  has  not  stopped  corporations  from  learning  how  to  measure,  manage,  and  reduce  their  emissions,  and  these  efforts  are  reaping  benefits  (Lee,  2011).               15Measuring  the  carbon  footprint  of  a  large  multinational  corporation  is  a  technical  feat,  requiring  expertise,  scientific  knowledge,  and  an  evolved  corporate  culture  (Halldórsson  et  al.,  2009).  However,  now  more  than  ever,  companies  have  an  increased  capacity  and  willingness  to  begin  measuring  this  aspect  of  their  environmental  footprints  (CDP,  2010).  Furthermore,  lessons  learned  here  could  be  used  in  measuring  the  environmental  impacts  of  using  other  public  goods,  like  water.    1.1.1.3 Responses  to  Sustainability    Here  I  consider  responses  to  sustainability  in  the  form  of  CSR  practices,  basing  the  analysis  on  a  framework  developed  by  Auld  et  al.  (2008a)  that  categorizes  CSR  innovations.  These  seven  categories  capture  how  a  corporation  might  respond  to  the  sustainability  imperative:  individual  firm  efforts;  individual  firm  and  NGO  agreements;  public-­‐private  partnerships;  information-­‐based  approaches;  environmental  management  systems  (EMSs);  industry  association  codes  of  conduct;  and  non-­‐state  market-­‐driven  (NSMD)  governance  in  the  form  of  private-­‐sector  hard  laws.    1.1.1.3.1 Sustainability  Response:  Individual  Firm  Efforts    Individual  firm  efforts  occur  when  a  firm  independently  makes  a  decision  to  become  more  environmentally  responsible.  Such  efforts  are  not  responses  to  regulations,  but  may  be  attempts  to  preempt  government.  Firms  may  uncover  win-­‐win  CSR  opportunities  or  adopt  win-­‐lose  strategies  that  hold  long-­‐term  financial  potential.  In  general,  internal  firm  efforts  are  not  subject  to  externally  imposed  prescriptive  requirements,  and  firms  control  the  processes  and  policies  developed.  This  flexibility  and  the  fact  that  win-­‐win  solutions  tend  to  be  the  focus  of  corporate  attention,  explain  why  individual  firm  efforts  represent  the  most  prevalent  and  widespread  manifestation  of  CSR  (Hill  et  al.,  2003).    1.1.1.3.2 Sustainability  Response:  Individual  Firm  and  Individual  NGO  Agreements  Agreements  refer  to  CSR  efforts  in  which  a  firm  engages  with  an  ENGO  or  other  stakeholders  to  address  the  environmental  impact  associated  of  a  firm’s  operations  (Hartman  and  Stafford,  1997;  Dauvergne  and  Lister,  2012).  In  general,  an  ENGO  will  come  to  view  an  environmental  issue  from  a  different  perspective  than  a  firm.  Firms  benefit  from  these  partnerships  by  adding  legitimacy  to  their  CSR  efforts  by  collaborating  with  traditional  adversaries  (Yaziji  and  Doh,  2009).    An  example  of  firm-­‐ENGO  collaboration  comes  from  the  Environmental  Defense  Fund  (EDF)  and  Wal-­‐Mart.  Wal-­‐Mart,  the  epitome  of  a  big  box  American  retailer,  has  integrated  environmental  variables  into               16its  practices  for  a  variety  of  reasons.  EDF  decided  to  leverage  this  opportunity,  seeing  collaboration  with  Wal-­‐Mart  as  a  chance  to  highlight  what  they  perceive  to  be  a  sincere  CSR  effort,  but  also  to  promote  best  practices  among  retailers  and  along  Wal-­‐Mart’s  massive  supply  chain  (EDF,  n.d.).      1.1.1.3.3 Sustainability  Response:  Public-­‐private  Partnerships  Public-­‐private  partnerships  are  similar  to  firm-­‐ENGO  agreements,  but  involve  other  interests,  such  as  governments,  or  are  made  up  of  several  firms  and  ENGOs  acting  in  concert.  These  partnerships  can  emerge  as  efforts  to  address  standards  development,  to  implement  self-­‐regulation,  or  to  develop  collaborative  co-­‐regulation  schemes  (Andonova,  2010).  These  partnerships  are  grounded  in  the  idea  that  private-­‐public  collaboration  can  provide  an  efficient  means  of  enforcing  costly  legislation.    An  early  example  of  public-­‐private  partnerships  comes  from  the  United  States  and  the  Environmental  Protection  Agency  (EPA),  which  developed  the  33/50  program.  Implemented  in  1991,  the  goal  of  the  program  was  to  reduce  emissions  of  17  toxic  chemicals  by  33%  in  1993,  and  50%  by  1995  (EPA,  1994).  Rather  than  set  strict  prescriptive  requirements,  the  EPA  gave  industry  flexibility  in  meeting  these  goals.  The  program  managed  to  achieve  its  targets  one  year  ahead  of  schedule  (Khanna  and  Damon,  1999).  Forestry  has  similar  examples.  The  Great  Bear  Rainforest  Agreement  and  the  Canadian  Boreal  Forest  Agreement  both  demonstrate  the  potency  of  partnerships  between  ENGOs,  industry,  and  government.  The  former  protected  a  large  area  of  rainforest  on  the  coast  of  British  Columbia,  allowing  ENGOs  to  claim  victory  while  providing  industry  with  a  degree  of  security  from  ENGO  criticism  ((Howlett  et  al.,  2009).  In  the  latter,  ENGOs  and  industry  (including  traditional  adversaries  like  Greenpeace  and  Weyerhaeuser)  agreed  on  environmental  conservational  goals  and  the  need  to  pursue  sustainable  forest  management  and  provide  jobs  for  forest-­‐dependent  communities  (FPAC,  2010).  This  provided  a  détente  of  sorts,  allowing  industry  to  focus  on  economic  issues  without  the  constant  risk  of  negative  attention  from  ENGOs.    1.1.1.3.4 Sustainability  Response:  Information-­‐based  Approaches  Many  CSR  efforts  revolve  around  the  provision  of  information  related  to  a  firm’s  behaviour  (Kharrazi  et  al.  2014).  Some  efforts  are  voluntary,  while  others  are  mandatory.  Government  sponsored  systems,  such  as  the  EPA’s  Toxic  Release  Inventory  (Khanna  et  al.,  1998),  require  companies  to  disclose  information  on  an  extensive  list  of  chemicals  that  their  activities  produce.  This  was  brought  about  by  advocates  who  argued  that  communities  have  the  ‘right  to  know’  about  what  corporations  are  doing.  In               17Canada  and  the  United  States,  like  most  other  developed  economics,  major  emitters  of  carbon  are  required  to  report  on  their  emissions  to  the  federal  government  (Jones  and  Ratnatunga,  2012).  Similarly,  food  producers  must  report  on  the  nutritional  requirements  of  their  products.  Wal-­‐Mart,  while  describing  the  future  of  its  Supplier  Sustainability  Index  has  suggested  that  a  carbon  label  might  be  required  for  all  of  the  goods  that  it  sells  (Wal-­‐Mart,  2011),  which  could  potentially  impact  sustainability  reporting  and  information  based  approaches  across  the  globe  (Gereffi  and  Christian,  2009).    These  examples  speak  to  an  increasing  emphasis  on  transparency.  By  requiring  the  disclosure  of  information  or  by  advocating  for  voluntary  disclosure,  transparency  becomes  the  norm  rather  than  the  exception.  Firms  that  do  not  participate  risk  losing  market  access.  Other  forms  of  information  sharing  also  revolve  around  this  notion  of  transparency.  In  the  United  States,  the  Lacey  Act  mandates  that  importers  of  forest  products  trace  the  origin  of  their  imports.  This  has  forced  the  development  of  sophisticated  supply  chain  practices,  such  as  radio-­‐tags  that  get  attached  to  felled  trees  at  harvest  sites  (USDA,  2011).  Non-­‐compliance  with  the  Lacey  Act  can  lead  to  embarrassment.  Gibson  Guitars,  a  high-­‐end  luthier  in  the  United  States,  experienced  public  scrutiny  when  its  operations  were  shut  down  by  government  officials  for  violating  the  Lacey  Act.  An  investigation  by  the  World  Resources  Institute  purchased  32  books  randomly  from  a  retailer  and  found  that  three  of  them  used  illegal  fibre  in  the  paper  (Nogueron  and  Hanson,  2010).  A  continued  need  for  transparency  and  measurement  along  the  supply  chain  remains.  Similar  legislation  for  conflict  minerals  (minerals  from  the  Congo  Basin  whose  sale  funds  armed  conflict)  are  being  implemented  in  the  United  States  which  would  require  manufactures  to  trace  the  source  of  certain  metals  in  their  products  (OpenCongress  ,2010).    The  Global  Reporting  Initiative  (GRI)  is  an  example  of  information  sharing  that  deserves  particular  attention.  The  GRI  has  become  the  de  facto  standard  (it  is  technically  a  guideline  rather  than  standard,  as  it  is  free  to  use  and  does  not  mandate  verification  or  auditing)  for  corporate  sustainability  reporting,  and  is  perceived  by  executives  as  second  only  to  the  ISO  14001  standard  in  influencing  their  CSR  practices  (Brown  et  al.  2009).  The  GRI  is  a  multi-­‐stakeholder  process,  involving  industry,  governments,  scientists,  and  civil  society.  Various  working  groups  develop  specific  guidelines  for  different  sectors.  Additionally,  the  guidelines  evolve  relatively  quickly;  the  third  version  of  GRI  is  in  place  after  only  9  years.  The  GRI  has  gained  widespread  support  for  a  variety  of  reasons.  It  is  a  ‘win-­‐win’  solution  as  corporations  adopt  a  reporting  framework  that  meets  the  requirements  of  many  social  actors  (notably,               18advocates  and  investors).  The  GRI  also  carries  with  it  a  strong  sense  of  legitimacy  and  inclusivity  and  is  a  partner  institution  of  the  UNEP.    At  the  opposite  end  of  the  scale  from  enterprise-­‐level  GRI  reporting  is  another  information-­‐based  approach:  Environmental  Product  Declarations  (EPDs.)  These  statements  are  equivalent  to  food  nutrition  labels,  disclosing  the  energy,  materials,  water  impacts  and  waste  emissions  associated  with  a  particular  product.  EPDs  are  based  off  ISO  standards  (ISO  14025,  2006)  and  can  only  be  prepared  once  Product  Category  Rules  (PCRs)  have  been  developed  using  a  multi-­‐stakeholder  process.  The  forestry  sector  has  embraced  EPDs,  as  they  are  increasingly  being  required  of  products  used  in  the  construction  industry  (Zackrisson  et  al.  2008).  EPD’s  have  been  developed  for  highly  specific  products,  like  Western  Red  Cedar  Decking,  Western  Red  Cedar  Bevel  Siding,  Glue  Laminated  Timbers  and  Softwood  Timber,  to  name  a  few  (FII,  2013).      1.1.1.3.5 Sustainability  Response:  Environmental  Management  Systems  Environmental  management  systems  (EMS)  are  externally  defined  and  imposed  criteria  about  how  a  firm  should  approach  its  environmental  footprint  (Melnyk,  2003).  The  ISO  14001  standard,  and  more  recently  the  26001  standard,  are  examples  of  widely  adopted  EMSs  that  organizations  adopt  in  order  to  standardize  their  internal  processes  for  handling  environmental  issues.  Firms  use  these  systems  to  measure  and  monitor  their  footprint  with  the  goal  of  reducing  inefficiencies  and  identifying  risk.  They  carry  with  them  credibility,  and  the  opportunity  to  attach  a  recognized  logo  to  a  product,  but  there  are  costs  associated  with  implementation  and  third-­‐party  verification  of  EMS  standards.    Firms  can  also  employ  other  systems  to  measure  their  environmental  footprint.  A  common  tool,  in  both  academia  and  the  corporate  world,  is  life-­‐cycle  assessment  (LCA)  (Gauthier,  2005).  While  ISO  standards  are  used  to  define  protocols  on  how  to  conduct  an  LCA,  LCAs  are  rarely  verified  by  a  third  party.  Increasingly,  LCAs  are  utilizing  data  from  life  cycle  inventory  (LCI)  databases  that  are  privately  owned  and  not  subject  to  peer-­‐review.  While  these  databases  make  the  conduct  of  an  LCA  cost  effective,  they  may  also  hide  uncertainties  or  gaps  in  data  (Auld  et  al.,  2008a).    1.1.1.3.6 Sustainability  Response:  Industry  Association  Codes  of  Conduct  Industry  codes  of  conduct  are  typically  found  in  sectors  where  companies  tend  to  share  a  collective  reputation.  This  can  happen  when  the  behaviour  of  one  firm  can  impact  an  entire  industry,  regardless  of               19individual  firm  performance  (Bondy  et  al.,  2004).  Most  often  these  are  sectors  that  produce  primary  or  intermediary  goods,  like  forestry,  mining,  or  petrochemicals  (Jenkins  and  Yakovleva,  2006).  Recently,  other  sectors  have  adopted  codes  of  conduct:  the  coffee  industry,  which  sells  directly  to  consumers,  has  several  schemes  in  place  (Giovannucci  and  Ponte,  2005);  and  the  textile  industry  in  the  United  States  has  created  a  Sustainable  Apparel  Coalition  (Sustainable  Apparel  Coalition,  2014)  with  the  backing  of  ENGOs  (EDF),  government  bodies  (EPA),  and  major  clothing  manufacturers  (Levi  Strauss,  Nike,  and  Mountain  Equipment  Co-­‐op,  to  name  a  few).  These  codes  of  conduct  are  often  principles-­‐driven  and  administered  by  industry  associations;  they  do  not  provide  specific,  prescriptive  guidance  on  how  firms  should  behave.  Rather,  concepts  are  defined  and  adherents  promise  to  meet  them,  although  verification  and  enforcement  may  not  be  as  stringent  as  in  other  types  of  CSR  innovations.    1.1.1.3.7 Sustainability  Response:  Nonstate  Market-­‐driven  Governance  Nonstate  market-­‐driven  (NSMD)  governance  is  a  type  of  CSR  innovation  defined  by  Auld  et  al.  (2008a,  2008b).  It  differs  from  all  of  the  above  CSR  tools  in  that  it  requires  mandatory  and  enduring  behaviour  (so-­‐called  ‘hard  law’),  but  is  not  enforced  by  the  state.  Further,  it  differs  from  traditional  hard  law  by  continually  adapting  over  time  with  the  input  of  various  stakeholders  (Bernstein  and  Cashore,  2007).  NSMD  governance  also  differs  from  other  CSR  innovations  in  that  it  is  rooted  in  the  supply  chain.  NSMD  governance  requires  a  firm  to  understand  and  manage  the  upstream  implications  of  its  activities,  but  also  requires  a  firm  to  look  downstream  to  understand  that  it  is  ultimately  the  market  that  drives  NSMD  schemes  to  be  adopted.  Finally,  NSMD  governance,  unlike  most  CSR  efforts,  requires  third-­‐party  verification  (Cashore  et  al.,  2007).  This  imposes  a  significant  cost  on  firms,  but  also  ensures  that  NSMD  schemes  remain  something  akin  to  hard  law.    The  most  obvious  example  of  NSMD  is  certification.  Certification  schemes  exist  in  several  industries  (seafood,  for  example,  has  seen  recent  growth  in  a  variety  of  schemes).  But  forestry  has  led  the  way  in  the  development  of  certification  systems.  Perhaps  the  most  widely  known  scheme  is  the  Forest  Stewardship  Council  (FSC).  Borne  out  of  dialogue  around  sustainability  at  the  Rio  Environmental  Summit  in  1992,  FSC  is  an  independent  body  jointly  controlled  by  an  environmental  committee  and  an  industry  committee  (see  Rametsteiner  and  Simula,  2003  for  a  more  thorough  discussion  of  forest  certification  schemes).  Other  prominent  schemes  are  the  Sustainable  Forestry  Initiative  (SFI,  an  industry-­‐backed  scheme  in  North  America)  and  the  Programme  for  the  Endorsement  of  Forest  Certification  (PEFC,               20dominant  in  Europe  with  a  significant  Canadian  presence  and  composed  of  a  central  body  that  allows  for  country-­‐level  definitions  of  certification).    Whether  a  firm  choses  to  participate  in  a  forest  certification  scheme  relates  to  the  character  of  the  firm  (publicly  or  privately  owned),  the  product  they  are  manufacturing  (magazine  paper,  dimensional  lumber,  pulp,  etc.),  trade  dependence  (major  exporter  or  domestically  oriented),  ENGO  pressure,  supply  chain  procurement  policies,  and  government  support.  Small  operations  may  find  compliance  too  costly  and  not  worth  the  effort  (Crals  and  Vereeck,  2005).  Large  forestry  firms,  in  contrast,  are  more  vulnerable  to  pressure  from  advocates  and  can  better  bear  the  costs  of  compliance.  In  fact,  in  many  instances,  certification  among  larger  supply  chain  actors  is  fast  becoming  the  norm  in  the  marketplace  (Cashore  et  al.,  2005).  Other  firms,  such  as  those  that  sell  intermediary  products,  must  comply  with  more  than  one  scheme  in  order  to  meet  various  customers’  demands  (Mikkilä  and  Toppinen,  2008).  For  example,  a  paper  company  might  have  to  be  SFI-­‐certified  to  sell  to  an  American  client,  FSC-­‐certified  to  sell  to  a  Canadian  client,  and  PEFC-­‐certified  to  sell  to  a  European  client.    1.1.2 The  Media  Shift  We’ve  described  and  defined  sustainability  and  its  facets,  possibilities,  and  contexts.  With  these  drivers,  definitions,  and  responses  in  mind,  in  this  section  I  set  out  to  characterize  the  media  shift  that  is  underway.  Describing  the  paper  to  pixel  transition  is  not  simple.  There  is  such  variety  in  the  many  transitions  that  no  single  metric  that  can  capture  the  diversity  in  the  scope,  volume,  or  intensity  of  the  transition.  My  goal  is  not  to  describe  specific  transition,  but  instead  to  consider  the  types  of  transitions.  Given  the  diversity  of  media  that  can  be  consumed  either  on  paper  or  digitally,  this  approach  was  necessary.    The  complexity  of  the  transition  is  simple  to  illustrate:  media  that  previously  could  be  consumed  only  on  paper  is  now  available  on  a  vast  array  of  digital  devices.  Paper  media  can  be  a  newspaper,  magazine,  book,  catalogue,  flyer  insert,  directory,  printing  or  writing  paper,  invoices,  and  bills,  etc.  All  of  these  serve  as  a  basis  for  communicating  information.  Digital  media  is  similarly  equipped  to  communicate  over  a  wide  variety  of  platforms  including  desktop  computers,  laptops,  internet-­‐enabled  cell  phones,  tablet  computers,  or  E-­‐Readers.  With  this  many  products  and  types  of  information,  the  scope  for  different  transitions  is  vast  (Hetemäki  and  Nilsson,  2005).                 21There  are  other  characteristics  of  media  that  need  to  be  considered  when  describing  the  paper  to  pixel  transition.  Media  consumption  is  not  necessarily  a  zero-­‐sum  game.  If  the  marketplace  consumes  more  digital  media,  it  does  not  necessarily  to  come  at  the  expense  of  paper  media.  This  is  not  to  suggest  that  increased  digital  consumption  does  not  impact  paper  consumption,  it’s  just  not  a  direct  trade-­‐off.      Nor  is  the  transition  towards  digital  media  the  result  of  consumers  becoming  tired  of  the  offerings  provided  on  paper  media  (Hujala,  2011).  Digital  devices  are  capable  of  providing  a  broad  range  of  services,  far  beyond  the  ability  to  communicate  information  previously  on  paper.  For  example,  digital  devices  can  email,  bank,  shop,  play  videos,  take  pictures,  and  act  as  GPS  devices  –  this  multi-­‐functionality  is  a  large  part  of  their  attraction  to  consumers.  And,  most  importantly,  it  also  implies  that  the  environmental  impacts  of  digital  devices  are  spread  over  multiple  uses.      In  this  section,  I  review  digital  and  paper  media,  their  lifecycles,  and  the  range  of  environmental  issues  associated  with  media  production,  distribution,  consumption,  and  end-­‐of-­‐life.  The  objective  is  to  bring  into  focus  the  characteristics  of  each  sector,  and  to  identify  key  trends  that  indicate  a  shift  is  underway.  By  combining  this  review  with  definitions  of  sustainability  (see  section  1.1.1  The  Sustainability  Shift)  I  can  define  research  questions  that  will  add  depth  and  context  to  our  understanding  of  the  sustainability  of  media.      1.1.2.1 Digital  Media    I  discuss  trends  in  the  digital  media  according  to  various  lifecycle  stages.  The  goal  is  to  better  understand  the  diffuse,  complex  and  rapidly  changing  digital  media  landscape.  I  organize  the  review  in  four  parts    –  media  production,  distribution,  consumption,  and  end-­‐of-­‐life.  In  each  category,  I  examine  recent  trends  and  the  environmental  context  of  each.  Digital  media  is  employed  as  a  holistic  term;  it  incorporates  both  the  virtual,  final  product  (the  digital  content  viewed  on  a  screen),  as  well  as  all  of  the  necessary  Information  and  Communication  Technology  (ICT)  that  supports  the  product,  delivery,  and  consumption  of  digital  media.  This  includes  hardware  and  networking  devices,  consumer  devices  like  computers  and  smartphones,  as  well  as  the  factories  required  to  manufacture  all  of  this  equipment.      1.1.2.1.1 Digital  Media  Production  To  actually  produce  digital  media  requires  a  global  and  complex  supply  chain  of  ICT.  To  deliver  digital  media  requires  a  complex  supply  chain,  from  raw  materials  to  material  processing  to  manufacturing               22facilities,  to  computer  hardware  and  networking  equipment.  Over  its  lifespan  ICT  equipment  hosts  a  massive  variety  of  products,  dispersing  its  footprint  across  a  range  of  digital  media.  But  the  key  point  is  that  digital  media  is  not  virtual  media:  it  depends  on  a  physical,  tangible,  and  complex  network  of  material  goods.      Digital  media  begins  making  hardware,  involving  a  variety  of  raw  materials,  including  chemicals,  metals  and  minerals.  Paul  and  Campbell  (2011)  investigate  the  environmental  footprints  of  rare  earth  minerals3,  substances  that  are  present  in  in  almost  all  ICTs.  In  all  stages  of  mining  and  refining  rare  earth  minerals,  there  are  adverse  impacts  on  the  environment.  After  sourcing  and  refining  raw  materials,  complex  industrial  processes  are  required  to  produce  ICT  equipment.  Lau  et  al.  (2002)  investigate  the  footprints  associated  with  electronics  manufacturing,  finding  that  a  host  of  environmental  problems  arise  in  the  production  and  then  eventual  disposal  of  computer  material  components.  Adamon  et  al.  (2005)  find  the  range  of  environmental  impacts  from  ICT  is  difficult  to  manage:  impacts  are  dispersed  along  the  supply  chain,  and  ICT  users  are  limited  in  their  ability  to  mitigate  footprints.      1.1.2.1.2 Digital  Media  Distribution  Once  ICT  hardware  and  networks  are  in  place,  digital  media  needs  to  be  distributed.  This  means  data  flowing  over  wired  and  wireless  networks,  being  processed  by  data  centers,  and  finally  arriving  at  the  various  devices  that  can  be  used  to  access  digital  media.  Over  the  past  two  decades,  there  has  been  an  almost  exponential  increase  in  the  volume  of  data  being  delivered  worldwide,  and  the  trend  is  accelerating.  According  to  Cisco  (2014),  devices  like  smartphones,  tablets,  and  laptops  are  expected  to  increase  in  number  from  7  billion  in  2013  to  10.3  billion  in  2018.  Over  the  same  period,  the  amount  of  data  consumed  over  mobile  networks  is  expected  to  increase  by  15.9  exabytes  per  month4.  Mobile  networks  are  also  evolving  away  from  3G  (third  generation)  to  4G  (fourth  generation)  networks.  This  requires  a  massive  build-­‐out  of  switches,  relays,  and  transmission  towers  in  order  to  support  the  new,  faster  mobile  frequencies  (Suk  Yu  Hui  and  Kai  Hau  Yeung,  2003).    As  the  volume  of  data  and  digital  networks  grow,  more  and  more  data  centers  are  required.  Terms  like  “the  cloud”  and  “big  data”  have  become  ubiquitous  as  firms  like  Google,  Apple,  Facebook,  Microsoft,  and  Amazon  measure  and  collect  more  data,  while  finding  ways  to  profit  from  the  process.  Koomey                                                                                                                            3  Rare  earth  minerals  are  used  in  a  variety  of  applications,  and  have  nuclear,  metallurgical,  chemical,  catalytic,  electrical,  magnetic,  and  optical  properties  that  are  sought-­‐after  by  industrial  users.  (Koltun  and  Tharumarajah,  2014)  4  To  put  that  figure  in  perspective,  an  exabyte  is  one  million  terabytes,  and  one  terabyte  (a  thousand  gigabytes)  is  about  the  size  of  large  hard  drive  in  a  desktop  computer.               23(2008)  reviews  the  explosive  growth  in  data  centers  that  started  in  the  early  2000s.  Early  in  the  growth  curve,  data  centers  were  smaller  and  their  operators  paid  little  attention  to  energy  consumption.  But  as  the  number  and  size  of  data  centers  grew,  they  began  to  consume  almost  2%  of  total  electricity  produced  in  the  United  States.  Operators  have  now  begun  to  manage  for  energy  consumption,  attempting  to  reduce  both  environmental  and  financial  costs  (Koomey,  2008).  But  not  all  companies  share  a  concern  for  reducing  the  environmental  footprint  of  their  data  centers.  A  recent  investigation  (BBC,  2014)  found  that  not  all  data  centers  are  created  equally.  While  some  providers  publicly  disclose  their  environmental  footprints  and  set  targets  for  absolute  reductions  in  emissions,  others  are  more  sanguine  about  their  facilities  impacts  and  energy  use  (Greenpeace,  2010).      1.1.2.1.3 Digital  Media  Consumption  The  production  and  distribution  of  digital  media  has  experienced  explosive  growth.  Unsurprisingly,  the  consumption  of  digital  media  –  both  digital  hardware  and  virtual  media  –  has  undergone  a  similar  experience.  Between  2009  and  2012,  global  shipments  of  smart  phones  and  tablets  per  quarter  rose  from  15  million  to  almost  190  million  today  (Business  Insider,  2014).  Over  the  same  period  shipments  of  traditional  personal  computers  (PCs),  like  desktops  and  laptops,  experienced  little  growth  (AVC,  2014).      In  mature  markets  like  North  America  and  Europe,  people  have  long  consumed  digital  media,  first  on  desktops  with  dial-­‐up  modems,  shifting  to  laptops  with  broadband  connections,  and  now  dominated  by  mobile  devices  with  wireless  connections.  According  to  the  MIT  Technology  Review  (2010),  developing  countries  like  India  are  experiencing  growth  that  skips  the  wired,  stationary  ICT  devices  and  are  moving  straight  towards  cheap,  mobile  devices.  While  mature  markets  still  dominate  the  total  consumption  of  mobile  data,  the  fastest  growth  is  occurring  in  regional  markets  like  Latin  America,  Africa,  and  Eastern  Europe.      Smartphones,  laptops,  and  tablets  have  been  steadily  declining  in  price  for  years,  facilitating  the  massive  growth  in  data  consumption  (Zuberbühler  Associates,  2013).  ICT  companies  know  that  not  all  consumers  can  afford  $700  smart  phones  and  are  creating  cheaper  devices  for  developing  markets.  But  this  trend  means  that  devices  tend  to  last  for  only  a  few  years  and  companies  may  in  fact  been  engaged  in  so-­‐called  ‘planned  obsolescence’  where  devices  are  deliberately  designed  to  have  a  limited  lifespan  (CBC  News,  2013b).  Why  make  a  device  intended  to  last  a  long  time,  when  after  just  12  or  24  months  a  new  device  with  better  performance  is  available?  Unfortunately,  this  approach  results  in  manufacturing               24complex  devices  with  embedded  environmental  impacts  that  have  relatively  short  lifespans,  clearly  suboptimal  from  the  perspective  of  environmental  sustainability.      1.1.2.1.4 Digital  Media  End-­‐of-­‐Life  Given  these  trends  in  production,  distribution,  and  consumption,  it  is  no  surprise  that  the  end-­‐of-­‐life  (EOL)  of  digital  media  is  problematic.  Ongondo  et  al.  (2011)  provide  a  thorough  review  of  the  management  of  electrical  and  electronic  waste  flows  worldwide.  Digital  media  relies  on  ICT  that  is  composed  of  abiotic,  non-­‐renewable  resources  that  are  difficult  and  expensive  to  process.  Processing  ICT  at  EOL  is  further  complicated  by  the  fact  that  dozens,  if  not  hundreds,  of  different  metals,  chemicals,  and  plastics  are  present  in  small  quantities.  This  makes  it  difficult  to  retrieve  any  particular  substance  at  a  large  scale,  removing  any  economic  imperative  to  manage  EOL.  As  a  result,  large  volumes  –  between  20  and  25  million  tonnes  per  year,  by  some  estimates  (Robinson,  2009)  –  of  ICT  waste  ends  up  in  landfills  or  exported  to  countries,  like  China,  India,  or  Ghana,  where  the  environmental  and  social  impacts  of  processing  ICT  waste  are  often  severe  (Ni  and  Zeng,  2009).  Some  jurisdictions  have  tried  to  improve  handling  of  ICT  waste.  Europe  passed  the  WEEE  (Waste  Electrical  and  Electronic  Equipment)  Directive,  requiring  the  handling  of  ICT  waste  domestically.  In  North  America,  the  province  of  British  Columbia  charges  consumers  an  up-­‐front  waste  handling  fee,  and  processes  all  ICT  waste  at  a  domestic  facility  operated  with  funds  from  the  up-­‐front  fees  (Driedger,  2001).  But  these  initiatives  are  the  exception,  not  the  norm  (Kahhat  et  al.,  2008).  In  the  face  of  explosive  growth  in  ICT  hardware  production,  ICT  device  consumption  and  digital  media  production,  the  EOL  impacts  of  ICT  and  digital  media  remain  a  pressing  environmental  concern  (Yu  et  al.,  2010).    1.1.2.1.5 Summary:  Digital  Media  Shift  Digital  media  is  intrinsically  connected  to  ICT  hardware,  devices  and  networks,  all  of  which  are  undergoing  significant  growth  in  production,  distribution,  and  delivery.  The  digital  media  shift  can  be  characterized  as  accelerating  quickly  and  increasingly  global.  It  is  complex  and  involves  hundreds  of  industrial  sectors  and  actors.  This  interconnected  web  of  networks,  data  centers,  and  devices  is  the  platform  that  hosts  all  digital  media.  And  this  platform  is  growing  globally,  evolving  as  new  technologies  emerge.  In  light  of  all  the  drivers  and  responses  to  sustainability  identified  earlier,  the  digital  media  shift  is  a  challenge  to  environmental  sustainability.  The  diffuse  and  rapid  change  makes  it  difficult  to  measure  and  manage  environmental  impacts,  both  negative  and  positive.                   251.1.2.2 Paper  Media    In  this  section,  I  explore  four  stages  of  the  paper  media  lifecycle  –  production,  distribution,  consumption,  and  end-­‐of-­‐life.  For  each  stage,  I  identify  the  most  important  trends  and  environmental  issues.  Clearly,  paper  is  a  very  different  media  form  than  digital.  Humans  have  been  producing  paper  for  thousands  of  years,  and  the  printing  press,  pioneered  in  the  1430s,  portended  the  almost  universal  adoption  of  paper  media  across  the  world.  Paper  media,  much  like  digital,  depends  on  expansive  industrial  system  –  in  this  case  forestry  –  in  order  to  be  produced.  Again,  the  goal  here  is  not  to  define  every  environmental  impact  or  sustainability  issue  in  forestry.  Instead,  I  review  the  production,  distribution,  and  consumption  of  paper  and  the  environmental  impacts  that  arise.    An  important  factor  to  consider  is  that  not  all  paper  is  created  equally.  Trees  used  to  make  paper  could  come  from  a  sustainably  managed  forest  or  from  an  illegal  harvest  site  in  an  area  of  high  conservation  value.  A  paper  mill  could  be  run  at  a  world-­‐class  level,  with  all  environmental  impacts  measured  and  managed,  or  it  could  originate  from  an  old  mill  in  a  jurisdiction  with  weak  environmental  controls.  Paper  products  also  vary  widely,  from  an  enduring  product  like  a  classic  novel  or  a  passport,  to  something  transient  like  junk  mail  or  shopping  catalogue.  The  Environmental  Paper  Network  (EPN)  produces  a  paper  utility  matrix  that  identifies  four  classes  of  paper  (EPN,  2013):    • High  Utility  /  Low  Volume:  passports,  birth  Certificates,  letters,  photographs,  important  documents  • High  Utility  /  High  Volume:  books,  newspapers,  hygienic  papers  • Low  Utility  /  Low  Volume:  local  advertising,  advertising  posters  • Low  Utility  /  High  Volume:  junk  mail,  catalogues,  over-­‐prints  of  books  and  magazines,  over-­‐packaging    These  four  categories  demonstrate  the  wide  variety  of  paper  media  that  exists.  And  in  each  category,  digital  media  is  playing  a  disruptive  role,  either  augmenting  or  replacing  a  paper  product.      1.1.2.2.1 Paper  Production  The  impacts  of  paper  production  begin  in  the  forest,  where  logging  takes  place.  Berg  and  Lindholm  (2005)  review  the  energy  impacts  of  logging,  finding  that  removal  and  haulage  contribute  significantly  to  the  footprint  of  harvesting.  Boltz  et  al.  (2003)  compare  various  logging  methods,  finding  that  reduced-­‐impact  logging  can  significantly  decrease  the  footprint  of  harvesting.  After  a  log  is  removed  from  the  forest,  it  must  be  processed  (typically  at  a  sawmill)  and  chips  moved  from  the  sawmill  to  a  paper  mill.               26Here  another  host  of  environmental  challenges  emerge.  Carlberg  and  Stuthridge  (1996)  review  the  scope  of  environmental  impacts  of  paper  making,  from  the  origin  of  fibre  to  the  long-­‐term  impact  of  paper  mills  on  the  surrounding  environment.  Bajpai  (2011)  offers  a  more  recent  perspective,  identifying  how  paper  mills  have  improved  their  chemical,  effluent,  and  waste  handling  procedures  to  minimize  environmental  impacts.  Galloway  et  al.  (2004)  consider  how  paper  mills  and  their  effluents  can  adversely  impact  wildlife,  including  fish,  with  a  commensurate  impact  on  ecosystem  health  and  biodiversity.  This  reinforces  a  core  truth  about  forestry  and  papermaking:  it  involves  harvesting  and  processing  trees  (with  the  exception  of  recycling  various  forest  products),  thereby  disrupting  natural  environments  at  a  large  scale.  The  impacts  of  this  disruption  can  be  managed  sustainably,  through  adequate  planning  and  site-­‐level  remediation,  or  they  can  be  ignored,  resulting  in  deforestation  and/or  the  degradation  of  the  natural  environment.    In  the  production  of  paper,  there  is  significant  scope  for  variation.  In  some  instances,  forest  certification  schemes  may  be  used  to  encourage  sustainable  forest  management.  These  schemes,  either  industry  sponsored  or  a  collaborative  approach  involving  non-­‐state  actors  and  industry,  have  varying  levels  of  effectiveness.  Beyond  a  scheme  and  the  standards  it  sets,  the  role  of  governance  and  transparency  is  important  (Overdevest,  2010).  Recent  studies  (Visseren-­‐Hamakers  and  Pattberg,  2013)  have  suggested  that  it  is  inconclusive  whether  certification  meaningfully  and  consistently  improves  environmental  performance.        1.1.2.2.2 Paper  Distribution  Distributing  paper  media  has  its  own  set  of  environmental  impacts.  There  is  an  inevitable  reliance  on  transportation  networks,  including  sea  shipping,  railways,  trucks,  as  well  as  the  vehicles  of  consumers  who  drive  to  retailers  to  purchase  paper  media.  Borggren  et  al.  (2011)  suggest  that  the  distribution  impacts  of  paper  media  can  vary  widely.  Some  paper  media,  like  books,  are  easily  shared  and  can  last  for  decades.  How  a  consumer  actually  receives  a  book  matters  too:  do  they  drive  a  significant  distance  for  the  express  purpose  of  purchasing  a  book?  Or  do  they  take  public  transit,  and  walk  to  a  bookstore  as  part  of  their  daily  routine?  Or  perhaps  the  book  is  delivered,  part  of  a  network  of  trucks  and  distribution  centres  that  deliver  a  wide  variety  of  goods.                   27To  understand  the  footprint  of  distributing  paper  media,  there  is  no  better  case  study  than  that  of  mail.  The  delivery  of  mail  has  been  disrupted  by  the  advent  of  digital  media.  SLS  Consultants  (2008)  conducted  a  study  of  the  evolving  mail  industry  in  the  United  States.  Personal  correspondence,  as  well  as  invoices  and  statements,  used  to  dominate  the  mail  system.  Now  people  rely  on  emails  and  electronic  invoices  to  correspond.  Mail  volumes  have  declined,  but  have  also  evolved.  Unsolicited  mail,  or  junk  mail,  now  makes  up  the  majority  of  mail  delivered.  The  type  of  vehicle  used  for  delivery  can  influence  environmental  impacts:  older,  less  efficient  vehicles  can  potentially  double  the  footprint  of  delivering  mail.  The  density  of  a  mail  system  also  matters.  Delivering  paper  media  in  remote  areas  often  requires  driving  significant  distances.  Postal  workers  operating  in  urban  areas,  in  contrast,  can  easily  do  so  on  foot.      1.1.2.2.3 Paper  Consumption  The  impacts  of  paper  consumption  are  best  understood  through  the  concept  of  ‘paper  utility’  introduced  at  the  beginning  of  this  section.  Different  paper  types  are  of  high  or  low  utility  and  are  consumed  in  high  or  low  volumes.  A  paper  product  like  a  book  can  be  shared  and  stored  for  a  significant  period  of  time.  Some  have  suggested  (Skog  and  Nicholson,  2000)  that  paper  products  can  even  act  as  a  carbon  sink,  thereby  assisting  in  efforts  to  mitigate  and  abate  climate  change.  In  contrast,  high  volume  low  utility  products  –  junk  mail  being  the  obvious  example  –  can  have  a  significant  and  adverse  impact  on  the  environment.  Consumers  often  immediately  discard  some  forms  of  paper,  gaining  no  utility  from  its  production  and  distribution.    1.1.2.2.4 Paper  End-­‐of-­‐Life  At  the  end-­‐of-­‐life  (EOL),  paper  media  faces  two  options:  it  can  be  recycled  or  it  can  be  discarded.  Recycling  paper  itself  has  an  impact:  it  takes  energy  and  water  to  break  paper  down  to  its  constituent  fibres  and  recycle  these  into  a  new  product.  Virtaten  and  Nilsson  (1993)  offer  a  broad  view  of  all  the  environmental  impacts  of  paper  recycling.  They  find  that  recycling  has  the  potential  to  actually  use  more  energy  and  water  than  virgin  fibre,  although  specific  energy  sources  and  processes  employed  matter  as  well.  If  a  paper  product  is  not  recycled  at  its  EOL,  chances  are  it  ends  up  in  a  landfill.  Research  has  suggested  (Agriculture  Today,  2010)  that  paper  in  a  landfill  can  decompose  in  an  oxygen  starved  environment,  thereby  resulting  in  methane  emissions,  a  potent  source  of  greenhouse  gases.  Some  estimates  (Micales  and  Skog,  1997;  Wang  et  al.,  2013)  suggest  that  decomposition  in  a  landfill  and  the               28associated  methane  can  actually  produce  the  most  greenhouse  gas  emissions  of  any  stage  in  the  paper  lifecycle.    1.1.2.2.5 Paper  Media  Shift  I  have  briefly  reviewed  the  various  lifecycle  stages  –  from  forest  floor  to  delivery  truck  to  landfill  –  where  paper  has  the  potential  to  impact  the  environment.  But  for  the  purposes  of  this  thesis,  I  need  to  also  review  how  paper  media  is  changing.  As  digital  media  has  emerged,  it  has  clearly  impacted  the  consumption  of  paper  media.  I  want  to  contextualize  and  deepen  the  understanding  of  this  impact.      A  big  change  in  paper  media  has  been  the  decline  of  traditional  paper  media  outlets.  The  best  example  of  the  paper-­‐to-­‐digital  media  shift  is  that  of  newspapers.  For  well  over  a  century,  newspapers  dominated  the  media  landscape,  generating  revenue  by  selling  physical  copies,  advertisements  as  well  as  classifieds.  But  with  the  advent  of  digital  media,  this  revenue  model  faced  an  existential  threat.  Websites  like  Craigslist  offered  a  platform  for  posting  employment  advertisements,  real  estate  listings,  and  various  other  products.  The  monopoly  of  newspapers  on  distributing  targeted  and  local  advertising  was  suddenly  broken.  Between  2003  and  2012,  newspapers  in  the  United  States  lost  $11  billion  in  classified  revenue  (Edmonds,  2013).  This  amplified  a  downward  spiral  of  declining  readership  and  advertising  revenue  brought  about  by  changing  consumer  preferences.  Industry  believes  that  the  transition  has  stabilized  (Newspaper  Association  of  America,  2013),  but  big  brands  like  the  Wall  Street  Journal  and  the  New  York  Times  are  now  relying  on  digital  subscriptions  to  augment  their  revenue  streams.      Newspapers  are  not  the  only  paper  medium  to  be  impacted:  magazines  have  suffered  too.  Prominent  brands  like  Life  Magazine  and  Newsweek  have  ceased  paper  publications,  unable  to  sell  enough  magazines  or  advertising  to  remain  profitable  (Media  Life  Magazine,  2014).  While  niche  magazines  endure,  it  is  only  prominent  brands  (e.g.  The  Economist)  or  publications  with  generous  corporate  backing  (e.g.  Time  Magazine)  that  have  survived  the  media  shift  (Pew  Research  Journalism  Project,  2013).  Experts  predict  that  the  future  of  media  is  both  paper  and  digital  (Ferguson,  2009).  Consumers  vary  in  their  preferences,  and  the  advent  of  cheap  and  accessible  digital  technology  has,  by  no  means,  eradicated  all  print  media.  Established  paper  media  brands  (like,  as  mentioned  previously,  the  New  York  Times)  are  able  to  leverage  their  editorial  staff  and  reputation  to  generate  significant  digital  revenues.                 29I  have  already  discussed  the  decline  in  mail  volumes  that  the  digital  to  paper  media  shift  has  induced,  but  is  worth  revisiting.  Between  1980  and  2013,  mail  volume  went  from  around  60  billion  pieces  mailed  per  year  to  a  peak  of  103  billion  in  2002  and  back  to  65  billion  (United  States  Postal  Service,  2014).  With  current  mail  volumes  similar  to  levels  in  the  1980s,  it  is  no  surprise  that  postal  services  worldwide  have  had  to  aggressively  curtail  services  and  cut  costs  in  order  to  operate  without  losing  large  sums  of  money.      One  final  shift  worth  considering  is  that  of  the  office  place.  In  the  workplace,  paper  and  digital  media  form  a  complex,  and  at  times  symbiotic,  relationship.  Digital  tools  are  used  to  produce  paper  products  (like  a  printed  email).  And  while  the  use  of  digital  media  has  exploded,  the  decline  of  paper  media  has  not  experienced  an  equal  and  inverse  implosion.  York  (2006)  sees  the  concept  of  the  ‘paperless  office’  as  a  quintessential  example  of  Jevon’s  paradox,  which  posits  that  any  increase  in  efficiency  is  often  offset  by  an  increase  in  use.  In  the  office  environment,  this  paradox  manifests  in  increased  digital  media  consumption  along  with  only  modest  declines  in  paper  consumption.  While  digital  media  may  be  more  efficient  at  delivering  memos,  emails,  or  reports,  paper  media  is  still  being  consumed.  As  a  result,  the  aggregate  environmental  footprint  of  office  workers  may  have  increased.  Researchers  who  study  the  effectiveness  of  media  types  suggest  that  a  mix  of  paper  and  digital  media  is  likely  to  persist,  as  each  format  has  its  own  attributes  and  strengths  (Ashby,  2011).    1.2 RESEARCH  QUESTIONS  I  have  demonstrated  that  there  has  been  a  shift  in  media  consumption,  and  that  concurrent  to  that  shift  has  been  the  emergent  priority  of  sustainability.  To  probe  these  trends  further,  and  to  take  the  time  to  “consider  the  environment”,  I  have  arrived  at  three  research  questions  that  seek  to  unpack  the  relationship  between  sustainability  and  media.  These  questions  look  at  the  same  issue  –  media  sustainability  –  from  three  distinct  perspectives:  the  consumer,  industry,  and  academic  comparisons  of  media  choices.    Given  what  I  have  learned  about  the  sustainability  shift  and  the  media  shift,  it  is  clear  that  multiple  perspectives  are  required.  I  have  identified  three  research  questions  that  will  deepen  an  understanding  of  why  I  should  consider  the  environment  when  printing  an  email.  As  already  mentioned,  clearly  the  paper  sector  deserves  particular  attention,  having  been  classified  as  the  product  requiring  environmental  consideration.  I  have  also  seen  that  media  consumption  is  shifting,  and  want  to  find  out  whether  a  survey  of  consumers  provides  further  evidence  of  this  transition.  Finally,  I  have  identified  the  need  to  compare  paper  and  digital.  To  that  end,  I  need  to  conduct  a  thorough  analysis  of  the  research  that  has  compared  both  paper  and  pixels.                 30  Influencing  and  motivating  this  line  of  inquiry  are  the  two  shifts  I  discussed  in  detail  above.  As  the  world  responds  to  environmental  challenges  by  pursuing  sustainability,  and  as  the  media  landscape  evolves  in  light  of  new  technologies,  I  need  to  better  understand  what  this  means  for  the  environment.  Given  the  breadth  and  depth  of  the  trends  that  motivate  this  dissertation,  I  cannot  focus  simply  on  a  discrete  trade-­‐off  between  digital  and  paper  media.  A  discrete  trade-­‐off,  say  a  consumer  reading  a  book  digitally  instead  of  in  print,  is  an  important  question.  But  it  does  not  capture  the  systemic  shifts  in  sustainability  and  media.  It  might  say  something  about  the  impacts  of  a  single  consumptive  choice,  yet  reveals  very  little  about  the  impacts  of  entire  industries  across  the  planet  and  across  decades  of  time.  This  dissertation  seeks  to  elucidate  the  broad  and  deep  questions  that  a  comparative  analysis  of  two  industrial  sectors  requires.  The  uncertainties  and  assumptions  that  shape  an  understanding  of  paper  and  digital  media  need  to  be  identified  so  that  I  might  better  understand  the  ecology  of  the  industries  involved,  and  contribute  to  a  theory  of  sustainability  that  is  both  nuanced  and  evolved.      But  scientific  inquiry  and  academic  research  needs  to  be  grounded  in  testable  hypotheses,  sound  methods,  and  clear  reasoning.  While  a  long  and  philosophical  exploration  of  sustainability  is  tempting,  grounding  this  dissertation  in  sound  research  will  ensure  that  results  and  findings  are  theoretically  robust  and  relevant.  Reflecting  these  trends  and  objectives,  the  three  research  questions  are  as  follows:  1. Given  the  focus  of  environmental  concern  on  paper  media,  how  does  sustainability  operate  along  a  paper  media  supply  chain?  2. Given  the  shift  of  consumers  from  paper  to  digital  media  sources,  has  sustainability  impacted  consumers’  media  consumption  habits?  3. Given  the  proliferation  of  media  choices,  have  life  cycle  comparisons  been  made?  And  if  so,  what  are  their  findings?    I  start  with  an  inquiry  into  paper  media  supply  chains.  Because  paper  is  seen  as  an  environmental  pariah,  I  want  to  investigate  how  that  sector  is  managing  sustainability  issues.  Rather  than  focus  on  one  company  or  facility,  I  want  to  investigate  a  media  supply  chain  to  learn  more  about  how  sustainability  impacts  the  way  business  is  done.  I  identify  a  specific  environmental  variable  in  order  to  fully  understand  how  supply  chains  are  coping  with  sustainability.  Sustainability  itself  is  a  broad  concept,  and  focus  is  needed.  I  use  carbon  as  an  environmental  variable  to  explore  for  a  variety  of  reasons:  it  is  intrinsically  connected  to  climate  change,  the  pressing  environmental  issue  of  our  age;  it  is  a  widely               31understood  and  managed  concept;  and  it  is  a  proxy  for  cost,  as  carbon  represents  energy  which  is  an  economic  cost  faced  by  all  businesses.  In  short,  I  am  using  carbon  as  a  proxy  for  environmental  sustainability.  In  selecting  carbon  as  a  vehicle  for  studying  sustainable  supply  chain  management,  I  utilize  a  variable  that  has  traction  and  a  shared  understanding  in  the  business  community.  I  hypothesize  that  supply  chains  play  an  important  role  in  managing  environmental  variables,  as  they  capture  a  large  part  of  the  lifecycle  of  a  media  product.  To  conduct  this  portion  of  research,  I  use  expert  interviews,  grounded  theory  methods,  and  a  quantitative  analysis  of  the  carbon  footprint  of  a  magazine.      I  then  follow  with  an  examination  of  the  environmental  values  of  media  consumers.  I  want  to  evaluate  whether  consumers  are  shifting  in  their  media  consumption  habits,  and  whether  environmental  values  have  any  role  to  play  in  that  shift.  I  hypothesize  that  consumers  who  care  more  about  the  environment  consume  less  media,  thereby  reducing  their  environmental  footprint.  I  also  hypothesize  that  consumers  are  shifting  in  their  media  consumption  habits,  with  digital  media  playing  an  increasingly  large  role.  While  I  did  present  evidence  of  this  shift  earlier  in  Chapter  1,  I  want  to  validate  and  add  nuance  to  my  understanding  of  changing  consumer  habits.  I  survey  consumers  in  North  America  using  an  online  survey  tool,  asking  a  series  of  questions  on  environmental  values  and  media  consumption  habits.  Where  appropriate,  I  apply  statistical  methods  to  strengthen  and  validate  my  results.  I  employ  the  New  Ecological  Paradigm  (NEP)  Dunlap  et  al.  (2000)  in  order  to  segment  my  survey  sample,  while  also  connecting  my  research  with  a  broader  academic  discourse.      Finally,  I  evaluate  the  findings  of  academic  research  that  compares  the  environmental  footprint  of  paper  and  media  products.  Beyond  a  simple  literature  review,  however,  I  construct  an  analytical  framework  so  that  I  can  better  understand  the  limitations  of  academic  research,  and  what  gaps  in  the  environmental  lifecycle  of  media  products  traditional  methodologies  might  miss.  I  hypothesize  that  the  standard  academic  method  of  comparing  environmental  footprints,  the  life  cycle  assessment  (LCA)  has  limitations  when  assessing  the  footprint  of  digital  media.  I  use  an  analytical  framework  developed  by  Reap  et  al.  (2008)  to  structure  my  research.  The  framework  identifies  the  key  areas  of  concern,  as  well  as  opportunities  for  improvement,  in  the  conduct  of  LCAs.  This  approach  allows  us  to  compare  LCAs  using  a  common  framework.    Our  results  are  presented  in  three  chapters,  organized  according  to  Figure  3  on  the  subsequent  page.  The  first,  Chapter  2,  presents  the  results  of  my  investigation  into  a  paper  media  supply  chain.  The  second,  Chapter  3,  examines  consumer’s  perceptions  of  media  consumption  and  environmental  values.               32The  third,  Chapter  4,  provides  a  framework  for  analyzing  comparisons  between  paper  and  digital  media  products.  Together,  this  research  constructs  a  more  nuanced  definition  of  the  sustainability  of  media.  I  then  synthesize  the  findings  into  a  broader  discussion  of  media  and  sustainability.  I  identify  patterns  that  connect  my  research  chapters,  looking  for  opportunities  to  strengthen  and  expand  an  understanding  of  sustainability.  My  goal  is  to  move  beyond  results  towards  new  ideas  that  might  help  us  better  understand  what  it  is  to  “consider  the  environment”,  contributing  to  the  existing  literature  while  offering  guidance  for  further  research.     Figure  3:  Summary  of  research  chapters  and  objectives               33Chapter  2:  The  Case  of  Supply  Chains:  Carbon’s  Role  in  Paper  Media5    2.1 ABSTRACT  The  purpose  of  this  case  study  is  to  identify  the  origins  and  evolution  of  carbon  management  along  a  supply  chain  in  the  paper  and  print  sector.  I  selected  carbon  as  the  environmental  metric  to  track  since  it  is  a  common  and  well-­‐understood  pollutant,  and  companies  have  developed  systems  and  processes  for  managing  their  own  performance.  But  carbon  can  also  be  used  as  a  proxy  for  sustainability  and  environmental  performance  in  general.  This  chapter  offers  an  investigation  of  how  supply  chains,  and  in  particular  paper  supply  chains,  consider  the  environment.  The  results  from  in-­‐depth  interviews  are  used  to  suggest  a  framework  for  assessing  carbon’s  impact  on  the  supply  chain.  I  find  that  the  use  of  biomass  for  energy  and  low-­‐carbon  transportation,  such  as  rail  and  sea-­‐based  barges,  can  reduce  the  carbon  footprint  of  a  paper  product.  The  interviews  reveal  that  upstream  and  downstream  supply  chain  actors  are  shaped  by  different  pressures.  Energy-­‐intensive,  upstream  actors  manage  their  carbon  footprints  in  order  to  save  energy  and  in  anticipation  of  regulated  carbon  emissions.  Downstream  actors,  in  contrast,  manage  carbon  in  order  to  strengthen  their  corporate  brand  and  maintain  market  share.  Businesses  trying  to  balance  short-­‐term  costs,  long-­‐term  profitability,  and  the  maintenance  of  a  corporate  brand,  have  identified  carbon  as  a  means  for  progress  on  all  three  fronts.      2.2 INTRODUCTION  The  challenge  of  climate  change  is  significant.  Shifts  in  global  temperatures,  weather  patterns,  and  sea  levels  may  adversely  impact  ecosystems  and  societies  alike  (Garnaut  Review,  2008).  To  address  this  challenge,  governments,  corporations,  civil  society,  and  scientists  are  collaborating  in  efforts  to  mitigate  and  adapt  to  the  impacts  of  climate  change  (U.S.  Climate  Action  Partnership,  2008).  Mitigation  strategies  are  focused  on  efforts  to  limit  and  reduce  carbon  emissions,  often  through  the  monetization  of  carbon.  Carbon,  in  short,  is  poised  to  have  a  price.  Any  business  that  emits  carbon  will  pay  for  its  emissions.  This  development  has  implications  for  supply  chains.  Monetized  carbon  may  force  supply  chains  to  consider  a  new  cost.                                                                                                                                5A  version  of  this  chapter  (2.1.)  was  peer  reviewed  and  published  as  follows:  “J.  G.  Bull,  G.  Kissack,  C.  Elliott,  R.A.  Kozak,  and  G.Q.  Bull.  (2011)  Carbon’s  Potential  to  Reshape  Supply  Chains  in  Paper  and  Print.  Journal  of  Forest  Products  Business  Research,  8(2)               34It  is  the  purpose  of  this  chapter  to  fill  a  gap,  exploring  the  potential  carbon  management  to  reshape  supply  chains  in  the  paper  and  print  industries.  This  will  be  done  in  two  parts.  I  begin  with  a  case  study  from  the  paper  and  print  sector  that  explores  the  role  of  the  supply  chain  in  managing  carbon.  I  follow  with  a  framework  grounded  in  my  case  study  that  explores  how  carbon  will  influence  supply  chains.  The  case  study  synthesizes  results  of  interviews  with  six  corporations  along  a  supply  chain  in  the  paper  and  publishing  industries.  I  explore  the  origins,  evolution,  and  future  directions  of  carbon  management  for  supply  chain  actors.  I  also  present  a  carbon  footprint  for  the  supply  chain.  I  suggest  that,  in  the  future,  a  thorough  understanding  of  carbon  may  become  a  central  issue  in  supply  chain  design  and  business  operations.      2.3 BACKGROUND  There  is  a  convergence  around  carbon.  Many  sectors  of  society,  from  corporations  to  NGOs  to  governments  and  international  bodies,  have  embraced  carbon  management  as  central  to  addressing  climate  change.  Governments  have  begun  to  regulate  carbon  emissions.  Internationally,  the  Kyoto  Protocol  and  subsequent  processes  are  working  towards  binding  international  commitments  for  reducing  carbon  emissions  (UNFCC,  2009;  Hedegaard,  2008).  At  the  regional  level,  the  European  Union  Emissions  Trading  Scheme  is  capping  and  monetizing  carbon  emissions.  Nationally,  Australia,  Canada,  and  the  United  States  are  moving  towards  binding  emissions  targets.  There  are  also  regional  and  local  initiatives  to  cap  and  trade  carbon  emissions  —  the  Western  Climate  Initiative  (WCI)  and  the  Regional  Greenhouse  Gas  Initiative  (RGGI)  are  prominent  examples.    The  corporate  world  is  also  responding,  treating  carbon  as  both  a  risk  and  an  opportunity.  There  is  a  growing  awareness  of  carbon  amongst  consumers  (Semenza  et  al.,  2008),  and  government  regulations  and  policies  are  increasing  well  (Neil  Adger  et  al.,  2005),  shaping  the  corporate  response.  Those  who  differentiate  their  products  based  on  its  carbon  footprint  may  be  rewarded  (Benjaafar  et  al.,  2013).  Insurance  companies  are  including  climate  change  in  their  long-­‐term  cost  projections  and  designing  incentives  to  reward  climate  change  mitigation  strategies  (SwissRe,  2008).  Investors  are  adjusting  their  decisions  to  include  climate  change  criteria  (UN  Principles  for  Responsible  Investing  2009).  Carbon  also  represents  a  possible  source  of  savings,  or  perhaps  revenue,  for  business.  As  carbon  gains  a  price,  business  may  be  forced  to  pay;  those  that  are  able  to  emit  less  will  spend  less  relative  to  their  competition  and  may  gain  market  advantages  (Luo  and  Bhattacharya,  2006).                   35Civil  society  is  active  in  shaping  the  response  to  climate  change  in  a  variety  of  ways.  For  example,  advocates  are  arguing  to  include  REDD  (Reduced  Emissions  from  Deforestation  and  Degradation)  schemes  in  a  post-­‐Kyoto  global  climate  accord  (IUCN,  2008).  Scientists  are  also  central  in  this  response,  from  the  Intergovernmental  Panel  on  Climate  Change  (IPCC),  a  body  composed  almost  entirely  of  scientists  and  recipient  of  the  Nobel  Peace  Prize  for  its  efforts  (IPCC,  2007)  to  the  Stern  Review  (Stern,  2006),  science  is  enjoying  a  prominent  voice  in  discussions  on  climate  change.  Demonstrative  of  this  consensus  behind  carbon  are  positions  taken  by  the  U.S.  Climate  Action  Partnership  (USCAP),  a  coalition  of  several  major  corporations  (including  Shell,  General  Electric,  and  ConocoPhillips),  and  four  prominent  NGOs  (Environmental  Defense,  Natural  Resources  Defense  Council,  Pew  Center  on  Global  Climate  Change,  and  World  Resources  Institute).    Beyond  the  convergence  of  governments,  corporations,  and  civil  society,  there  is  a  strong  theoretical  grounding  for  the  management  and  monetization  of  carbon  (Helm  et  al.,  2012).  Typically,  the  producer  of  a  good  pays  only  the  private  costs  associated  with  production.  Public  costs,  like  carbon,  are  not  captured  in  the  price.  The  evolution  of  carbon  into  a  regulated  pollutant  is  an  example  of  internalizing  a  negative  externality.  Economic  theory  has  established  the  “polluter  pays  principle,”  which  has  appeared  in  academic  literature  and  policy  for  over  30  years  (Gaines,  1991).  What  is  different  now  is  the  scale  and  the  application  to  carbon  (Woerdman  et  al.,  2008).  Carbon  is  the  largest  attempt  at  internalization  to  date,  and  with  a  pollutant  that  is  so  ubiquitous  and  intangible,  challenges  arise.  The  specifics,  means,  and  mechanisms,  while  well-­‐grounded  in  theory,  are  less  understood  in  practice.      2.4 OBJECTIVES  In  this  chapter,  there  are  three  objectives:  1. To  measure  the  relative  contribution  of  various  paper  supply  chain  stages  to  the  carbon  footprint  of  a  magazine.  2. To  elucidate  why  carbon,  in  particular,  has  the  potential  to  encourage  collaboration  along  supply  chains  in  the  paper  and  print  sector.    3. To  develop  a  framework  for  gauging  how  supply  chains  can  collaborate  on  environmental  issues,  thereby  changing  their  composition  and  behaviour  over  time.                   362.5 METHODOLOGY  Our  research  measured  the  carbon  footprint  of  a  supply  chain  with  a  focus  on  describing  the  influence  that  carbon  has  on  business-­‐to-­‐business  relationships  between  supply  chain  partners.    This  chain  oriented  around  the  production  of  a  major  American  magazine,  which  chose  to  remain  anonymous.  Using  a  case  study  approach  (Yin,  2009),  I  wanted  to  explore  carbon  not  only  as  a  quantifiable  emission,  but  as  a  phenomenon  that  influences  supply  chain  dynamics.  I  interviewed  six  corporations  for  this  case  study,  after  attaining  ethics  approval  (UBC  BREB  certificate  #H08-­‐02734).  I  investigated  the  origins,  evolution,  and  future  directions  of  carbon  management  in  each  company.  In  order  to  strengthen  the  internal  validity  of  the  case  study,  I  shared  the  findings  with  all  those  interviewed  and  incorporated  their  feedback.      The  steps  in  the  supply  chain  were  as  follows.  Catalyst  Paper  sourced  fibre  from  Western  Forest  Products  (WFP),  which  harvested  trees  on  Vancouver  Island,  Canada.  Residual  fibre  from  WFP’s  operations  was  shipped  to  Catalyst’s  mill  in  Port  Alberni  by  truck.  After  the  paper  was  manufactured,  it  was  shipped  by  truck  and  then  on  barges  operated  by  the  Washington  Marine  Group  (WMG)  to  Catalyst’s  distribution  center  in  Richmond,  Canada.  From  here,  it  was  shipped  by  Burlington  Northern  Santa  Fe  Railways  (BNSF)  to  Quebecor  World’s  printing  plant  in  Merced,  California,  where  it  was  printed  and  then  distributed  across  North  America.      The  standard  approach  in  quantifying  carbon  emissions  is  to  conduct  a  life  cycle  assessment  (LCA).  LCAs  of  magazine  products  exist  (Boguski,  2010;  Gower  et  al.  2006)  and  there  is  agreement  that  the  paper  manufacturing  process  is  significant  in  the  total  footprint  of  a  product.  For  example,  Boguksi  (2010)  found  that  79%  of  lifecycle  energy  is  accounted  for  by  the  cradle-­‐to-­‐gate  (meaning  from  harvest  in  the  forest  to  final  product  at  the  paper  mill)  for  coated  magazine  paper.  Gower  et  al.  (2006)  found  that  the  paper  manufacturing  process  accounts  for  61%  to  77%  of  total  lifecycle  carbon  emissions.  The  approach  of  focusing  on  the  supply  chain  partners  in  direct  contact  with  Catalyst  Paper  meant  that  I  measured  emissions  that  made  up  the  bulk  of  the  magazine’s  footprint.  I  therefore  felt  confident  that  the  methods  were  in  line  with  those  employed  in  other  LCAs,  and  chose  not  to  prioritize  the  replication  of  LCAs  that  already  exist.    Further,  there  are  problems  with  LCA  that  I  did  not  want  to  introduce  into  my  analysis  given  the  tandem  focus  on  quantifying  emissions  and  qualifying  their  influence  on  business  relationships.  Gadreault  et  al.               37(2007)  reviewed  forty  LCAs  in  the  pulp  and  paper  industry  and  found  that  sound  methodologies  for  assessing  land  use  and  demonstrating  the  carbon  storage  advantages  of  paper  were  absent.  They  also  noted  that  generalized  coarse-­‐level  LCAs  are  not  as  robust  as  LCAs  that  rely  on  primary  data  and  describe  specific  processes  or  products.  Reap  et  al.  (2008)  took  a  broader  view,  and  discussed  unresolved  problems  in  LCA  methodologies.  They  found  that,  at  each  stage  in  the  conduct  of  LCA,  there  are  several  challenges.  Most  prominent  in  the  context  of  a  paper  magazine  were  the  problems  of:  local  environmental  uniqueness;  spatial  variation;  time  horizons;  and  data  availability/quality.  Summarizing  these  issues,  Reap  et  al.  (2008,  p.384)  quotes  Bare  et  al.  (1999)  in  stating  that  it  is  hard  to  know  “where  to  draw  the  line  between  sound  science  and  modeling  assumptions.”      Pulp  and  paper  LCAs,  in  sum,  are  not  without  problems.  I  opted  to  avoid  the  full  LCA  methodology  not  only  because  of  these  problems,  but  because  I  also  had  the  advantage  of  a  unique  level  of  cooperation  and  accessibility  to  corporate  executives,  as  well  as  primary  data  for  several  stages  of  the  supply  chain.  I  felt  that  a  hybrid  approach,  quantifying  what  I  could,  while  describing  the  qualitative  influence  of  carbon,  would  lead  to  a  more  nuanced  understanding  of  how  carbon  can  influence  a  supply  chain.  When  I  did  quantify,  I  used  the  most  granular  data  available.  I  tracked  logs  from  specific  logging  operations  to  a  specific  mill,  and  along  specific  transport  routes  to  a  specific  printer.  I  avoided  generalized  emissions  factors  in  favor  of  specific  data  whenever  possible.    It  should  be  noted  that  I  did  not  include  the  distribution  footprint  for  the  product.  I  attempted  to  estimate  this  figure  using  several  approaches,  but  each  proved  highly  sensitive  to  assumptions  made.  Variables  under  consideration  were:  the  average  distance  traveled  by  each  copy  of  the  magazine;  the  volume  distributed  by  retail  outlets  compared  to  home-­‐delivery;  and  the  precise  geography  of  distribution.  Because  actual  data  was  not  available,  and  the  assumptions  produced  unacceptable  variation  in  results,  I  omitted  the  distribution  process  from  the  supply  chain  footprint.  This  is  an  area  where  further  research  is  warranted.      I  also  refrained  from  modeling  the  carbon  emissions  of  the  magazine  after  disposal  by  the  consumer.  Again,  there  was  too  much  potential  variation.  Whether  the  magazine  was  recycled,  incinerated,  archived,  or  buried  in  a  landfill  strongly  influenced  the  results.  Since  my  study  is  distinct  from  a  traditional  LCA,  I  used  data  for  specific  facilities  and  processes  rather  than  aggregate  data.  The  goal  was               38to  describe  the  carbon  emissions  of  the  supply  chain  stages  examined,  and  to  describe  the  relative  emissions  of  those  stages,  not  the  entire  lifecycle  of  the  magazine.      The  interviews  were  developed  using  qualitative  methods  for  exploring  complex  and  intricate  phenomena  that  are  difficult  to  express  quantitatively  (Cresswell,  1998;  Strauss  and  Corbin,  1998;  Yin,  2009).  Given  the  emergent  nature  of  the  topic  at  hand,  this  approach  was  deemed  the  most  appropriate  for  providing  a  better  understanding  of  carbon’s  potential  to  reshape  supply  chains.  The  following  companies  (five  of  which  consented  to  be  identified)  and  individuals  participated  in  the  interviews.  I  have  separated  the  companies  from  the  individuals  interviewed  to  further  protect  anonymity.     • Companies  o Anonymous  Magazine  Publisher6  An  internationally  distributed,  monthly  magazine.  o Burlington  Northern  Santa  Fe  Railways    (BNSF)  A  railway  operator  with  an  extensive  network  in  western  North  America.  o Catalyst  Paper  Corporation  A  paper  manufacturer  with  operations  on  the  west  coast  of  Canada.  o Quebecor  World  Inc.  One  of  the  largest  printing  companies  in  North  America.  o Washington  Marine  Group  A  shipping  company  with  operations  on  the  west  coast  of  North  America.  o Western  Forest  Products  (WFP)  A  forestry  company  with  operations  primarily  on  the  west  coast  of  Canada.    • Individuals  o Chief  Executive  Officer  o Chief  Operating  Officer  o Director,  Environmental  Affairs  o Director,  Paper  Procurement,  Environmental  Affairs  o General  Director,  Environmental  o Vice-­‐President,  Corporate  Relations  and  Social  Responsibility  o Vice  President,  Health,  Safety  and  Environment  o Vice-­‐President,  Manufacturing    Data  were  collected  through  in-­‐depth,  semi-­‐structured  interviews.  Participants  received  the  interview  questions  in  advance  (a  copy  of  the  interview  protocol  employed  is  found  in  Appendix  A).  Two  researchers  conducted  each  interview,  either  in  person  or  via  telephone.  Each  interview  was  recorded  and  transcribed.  The  interviews  focused  on  three  themes  of  interest  in  carbon  management:  origins;                                                                                                                            6  The  magazine  chose  to  remain  anonymous  out  of  fear  of  being  criticized  by  ENGOS  for  using  virgin  fibre  in  their  product  rather  than  recycled  fibre.               39evolution;  and  future  directions.  I  asked  companies  to  identify  how  carbon  manifested  as  a  management  priority,  to  describe  the  role  that  the  supply  chain  played  in  shaping  their  perceptions  of  carbon,  and  to  identify  specific  examples  of  interactions  with  external  actors  that  shaped  their  carbon  strategy.  Given  the  elite  status  of  those  being  interviewed,  I  adopted  methods  (Dexter,  1970)  that  acknowledged  the  expertise  of  the  interview  subjects.  The  interview  protocol  guided  each  interview,  but  when  the  interview  subject  demonstrated  additional  knowledge  or  interest  in  the  subject,  I  used  a  semi-­‐structured  approach  to  ensure  I  could  remain  responsive  to  a  subject’s  expertise.    Upon  completing  the  interviews,  I  supported  the  analysis  with  a  review  of  existing  literature  around  carbon  and  the  supply  chain.  This  review,  influenced  by  qualitative  methods  developed  by  Glaser  and  Strauss  (1967)  and  Strauss  and  Corbin  (1998),  led  to  the  construction  of  a  framework  on  efficient,  responsible,  and  resilient  supply  chains.  It  is  within  the  context  of  this  framework  that  I  discuss  the  interview  results.      Throughout  this  chapter,  I  use  the  term  ‘carbon  management’  in  an  intentionally  ambiguous  way.  It  can  mean  the  measurement  of  carbon  emissions,  or  the  acknowledgement  that  carbon  is  an  important  issue,  or  steps  taken  to  control  carbon  emissions.  In  other  words,  its  specific  meaning  varies  depending  on  the  context.      2.6 RESULTS  I  present  the  results  of  my  research  below,  beginning  with  a  review  of  the  carbon  footprint  of  Anonymous  Magazine.  I  follow  with  an  analysis  of  the  interviews,  highlighting  common  themes  and  particular  concerns  of  individual  companies.      2.6.1 Carbon  Footprint  Analysis  The  results  were  in  line  with  those  of  other  studies  (Boguski,  2010;  Gower  et  al.  2006)  that  measured  the  environmental  footprint  of  magazine  products.  I  found  that  paper  production  made  up  the  bulk  of  greenhouse  gas  emissions  associated  with  the  magazine,  but  that  transportation  of  paper  to  the  printing  facility  was  a  significant  source  of  emissions  in  and  of  itself.  Figure  4  shows  the  different  stages  in  the  supply  chain,  while  Table  1  provides  the  specific  contributions  from  each  supply  chain  stage.  I  follow  with  a  detailed  breakdown  of  the  results  and  assumptions  made  for  each  supply  chain  stage.    The  data  is  expressed  as  greenhouse  gas  equivalent  (referred  to  as  carbon  dioxide  or  C02  throughout)  per  air-­‐dried  tonne  (ADt).               40  The  question  of  how  much  carbon  was  emitted  from  the  felling  of  the  tree  compared  to  how  much  was  stored  in  the  final  product  was  not  addressed  in  our  analysis.  Life  cycle  research  has  often  treated  emissions  from  forest  activities  as  carbon  neutral,  but  this  assumption  has  been  challenged  (Helin  et  al.  2012).  Other  research  suggests  considerable  variability  in  the  stocks  and  flows  of  a  forest  product’s  life  cycle  (Winjum  et  al.,  1998).  Some  have  found  that  that  forest  products,  when  produced  efficiently  and  from  well-­‐managed  forests,  can  even  act  as  a  carbon  sink  (Marland  and  Schlamadinger,  1997;  Liu  and  Han,  2009).  The  lack  of  consensus  on  how  to  treat  embedded  emissions,  along  with  the  peripheral  nature  of  this  question  relative  to  the  objectives  of  this  section,  led  me  to  exclude  further  consideration  of  the  issue.       Figure  4:  Map  of  supply  chain  emissions      Harvesting,  road-­‐building  felling,  transport  to  sawmills    The  data  used  here  was  based  on  a  study  by  the  Forest  Engineering  Research  Institute  of  Canada  (FERIC,  1997)  that  found  that  6.9L  of  diesel  is  used  per  m3  harvested.  This  is  equivalent  to  18.5  kg  C02,  0.000816  kg  CH4  (methane),  and  0.000466  kg  N2O  (nitrous  oxide),  which  expressed  in  C02  equivalency,  is  18.66  kg  C02  per  m3  harvested  wood  converted  to  dimensional  lumber.  The  Western  Forest  Products  Alberni  Pacific  Division  generated  269,000m3  dimensional  wood,  175,000m3  chips,  307,000m3  hog  in  2008.  The  effective  carbon  footprint  on  all  products  is  18.66*269000  /  (269000+175000+307000)  =  6.7  kg  C02  /  m3  chips.  The  final  carbon  footprint  from  harvest  is  6.7*8.2m3/ADt  =  55  kg  C02  /  ADt.                 41Sawmilling  fibre  into  dimensional  and  residual  products  Based  on  WFP  2008  carbon  footprint,  Alberni  Pacific  Division  scope  I  &  II  emissions  =  5.5  kg  C02  /  m3  chips.  At  8.2  m3  chips  per  ADt,  sawmill  carbon  footprint  on  paper  basis  =  45  kg  C02  /  ADt.    Transport  of  chips  to  mills  Estimate  average  return  trip  of  chip  trucks  between  WFP  operations  and  Catalyst’s  mill  is  100  km.  Using  International  Panel  on  Climate  Change  (IPCC)  emission  factors  of  1.02  kg  C02/km  from  IPCC  EF  ID  19043,  emissions  associated  with  delivery  of  chips  =  102  kg  per  truckload  (at  3500  ft3  equal  ~  100  m3)  =  1.0  kg  C02/m3.  At  8.2  m3  chips  per  ADt,  carbon  delivery  footprint  =  8  kg  C02  /  ADt.    Conversion  of  chips  to  paper  at  Catalyst  Based  on  data  provided  by  Catalyst  for  its  carbon  footprint  in  2008  of  Scope  I  &  II7  emissions  =  185  kg  C02/  ADt.    Transport  of  paper  to  Quebecor  World  in  Merced  Based  on  identified  supply  chain  distances,  transportation  type,  and  recognized  IPCC  emissions  factors,  transport  footprint  =  127  kg  C02  /  ADt.    Printing  of  paper  at  Quebecor  World  in  Merced  Based  on  the  Heinz  (2006)  study,  the  surveyed  printing  facilities  carbon  footprint  is  36  kg  C02  /  ADt.       Table  1:  Supply  chain  emissions  Activity  Carbon  Emissions  (C02/ADt)   Percentage  of  Total  Harvesting,  road-­‐building,  felling,  transport  to  sawmills   55  kg   12%  Sawmilling  into  dimensional  and  residual  products   45  kg   10%  Transport  of  chips  to  mill   8  kg   2%  Paper  manufacturing  process   185  kg   41%  Transportation  to  print  facility   127  kg   28%  Printing  process   36  kg   8%  Total   456  kg   100%                                                                                                                              7  Scope  I  means  all  direct  Greenhouse  Gas  Emissions.  Scope  II  is  embedded  carbon,  meaning  all  indirect  emissions  from  consumption  of  purchased  electricity,  heat,  or  steam.               422.6.2 Origins  and  Evolution  of  Carbon  Management  The  interview  process  revealed  three  motives  for  managing  carbon:  as  a  performance  metric  in  pursuing  operational  excellence;  as  a  basis  for  product  differentiation;  or  as  a  strategic  priority  to  satisfy  corporate  commitments  to  environmental  responsibility.  The  motivations  of  a  company  depended  on  three  variables:  the  proximity  of  the  corporation  to  the  end  consumer;  the  degree  of  compliance  required  by  regulations;  and  the  need  to  enhance  the  corporate  brand.  Despite  these  diverse  motives,  carbon  provided  an  opportunity  for  cooperation  and  shared  understanding  between  supply  chain  partners.        I  classified  motives  for  carbon  management  into  two  categories:  internal  and  external.  The  only  internal  origin  identified  was  the  need  to  manage  energy  costs,  and  by  extension,  carbon.    BNSF  and  Catalyst  were  both  motivated  by  internal  origins.  BNSF,  which  spent  approximately  $4.6  billion  on  diesel  fuel  in  2008,  saw  carbon  as  an  entry  point  for  managing  fuel  costs.  To  them,  “not  only  does  carbon  make  economic  sense,  we  see  it  as  an  opportunity  to  differentiate  ourselves  from  an  environmental  perspective.”    Catalyst,  in  a  similar  vein,  wanted  to  reduce  its  energy  use.  To  do  so,  “Catalyst  invested  heavily  in  the  right  equipment  to  turn  waste  into  energy  for  their  operations,  and  as  they  did,  their  reliance  on  fossil  fuels  decreased  to  almost  zero.”  Catalyst  then  moved  beyond  operational  benefits  towards  a  more  sophisticated  marketing  strategy.  By  controlling  costs,  it  also  produced  a  unique  product;  paper  produced  while  emitting  as  little  carbon  as  possible.    External  origins  took  several  forms,  and  were  threefold  in  their  origins:  compliance  with  regulations;  response  to  pressures  from  civil  society;  or  relationships  between  supply  chain  partners.  The  latter  was  the  most  important  in  the  case  study.  WMG  cited  a  meeting  with  senior  executives  of  Catalyst  Paper  as  the  origin  of  its  carbon  management.  Catalyst,  in  trying  to  reduce  its  carbon  footprint,  engaged  with  WMG  to  maximize  the  use  of  fuel-­‐efficient  barges  in  moving  its  product.  WMG  cited  this  engagement  as  vital  in  its  own  consideration  of  carbon.  WMG  was  not  alone  in  crediting  its  relationship  with  Catalyst  as  an  origin  for  its  understanding  of  carbon.  With  the  exception  of  BNSF,  every  interviewee  had  been  actively  engaged  with  Catalyst  on  carbon  issues.  Catalyst  also  helped  connect  its  supply  chain  partners  with  World  Wildlife  Fund  Canada,  an  ENGO  that  assisted  Catalyst  in  the  measurement  of  its  carbon  emissions.  Two  respondents,  Catalyst  and  BNSF,  also  credited  regulatory  requirements  as  an  important  motive  for  managing  carbon.                 43A  company’s  position  along  the  supply  chain  also  affected  carbon  management.  Upstream  and  downstream  actors  face  different  pressures.  In  the  case  study,  upstream  suppliers  consume  more  energy  than  their  downstream  counterparts,  but  are  less  visible  to  consumers.  As  a  result,  they  are  more  likely  to  undertake  carbon  management  in  order  to  derive  cost  savings  or  comply  with  regulations.  Downstream  suppliers,  in  contrast,  use  less  energy  and  have  fewer  financial  incentives  to  undertake  carbon  management.  However,  non-­‐financial  incentives  do  influence  downstream  actors.  Quebecor  World,  for  example,  suggested  that  despite  an  economic  downturn,  there  was  a  strong  interest  in  sourcing  environmentally  preferable  paper.  To  meet  this  demand,  the  company  developed  a  database  of  carbon  emissions  for  all  of  the  paper  products  that  it  offers.  Suppliers  are  requested  to  fill  out  comprehensive  surveys  that  contain  information  on  the  carbon  emissions  of  their  products.  Quebecor  World  receives  an  almost  perfect  response  rate  to  this  survey  request.    I  found  that  most  supply  chain  actors  in  this  case  study  were  in  the  early  stages  of  developing  a  carbon  management  policy.  The  evolution,  therefore,  was  not  fully  understood  because  most,  with  the  exception  of  Catalyst  and  BNSF,  had  only  begun  to  develop  their  carbon  management  plans.  All  had  begun  to  take  the  first  steps  to  do  so,  but  were  still  in  the  formative  stages.  Specific  reduction  targets  were  the  exception  rather  than  the  norm.  This  speaks  to  the  current  regulatory  uncertainty  that  exists  in  North  America.  National-­‐level  schemes  are  evolving,  while  regional  initiatives,  such  as  the  WCI  and  RGGI,  may  have  impacted  some  of  the  corporations  interviewed,  but  were  not  significant  regulatory  priorities.  British  Columbia,  Canada,  proved  an  exception,  as  companies  operating  there  (such  as  Catalyst,  WFP,  and  WMG)  are  subject  to  a  carbon  tax  on  their  fossil  fuel  use.        2.6.3 Future  Directions  of  Carbon  Management  When  asked  where  they  thought  the  future  of  carbon  management  lay,  each  respondent  gave  an  answer  specific  to  its  own  corporation  and  industry.  The  economic  volatility  at  the  time  of  the  interviews  (January  through  March  2009)  influenced  answers.  It  should  be  noted  that  four  of  the  corporations  interviewed  were  in  the  print  industry  (WFP,  Catalyst,  Quebecor  World,  and  Anonymous  Magazine);  the  particular  hardships  of  this  industry  shaped  interviewees’  responses  on  future  directions.      Respondents  universally  agreed  that,  in  the  future,  carbon  and  sustainability  will  be  considered  more  closely.  They  recognized  carbon  as  a  potential  cost,  risk,  and  opportunity.  They  also  felt  that  the  marketplace  would  increasingly  demand  information  about  carbon  emissions.  They  suggested  that  the               44market  was  unwilling  to  pay  significant  premiums  on  carbon-­‐light  products  (that  is,  products  that  are  designed  and  manufactured  with  the  goal  of  reducing  carbon  emissions).  However,  carbon-­‐light  products  may  be  given  preference  if  cost  competitive.      Interviewees  identified  the  ability  for  carbon  to  create  differentiated  products.  They  described  carbon’s  role  in  the  marketplace  as  a  three-­‐step  transition.  The  first  step  was  simple  differentiation.  The  second  involved  the  marketplace  rewarding  carbon-­‐light  products  with  increased  market  share;  further,  they  anticipated  some  scenarios  where  the  market  would  pay  a  premium  for  carbon-­‐light  products.  The  final  step  depended  on  how  the  monetization  of  carbon  plays  out.  If  polluters  are  eventually  forced  to  pay  for  emissions,  carbon-­‐light  producers  who  currently  only  enjoy  product  differentiation  may  actually  gain  cost  advantages.  Respondents  saw  this  as  a  medium-­‐  to  long-­‐term  development,  and  felt  that  differentiation  and  market  preference  are  priorities  in  the  short-­‐term.    It  was  suggested  that  carbon  has  the  potential  to  change  the  value  of  existing  industrial  assets.  This  was  particularly  true  for  three  companies  —  BNSF,  Catalyst,  and  WFP.  Their  assets,  and  their  economic  value,  would  change  in  a  low-­‐carbon  economy.  Catalyst  identified  the  possibility  of  using  underutilized  mills  to  produce  electricity  with  biomass.  WFP  saw  potential  in  recognizing  solid  wood  products  as  sinks  of  carbon.  Given  that  dimensional  lumber  can  exist  as  a  carbon  sink  in  a  home  for  decades  and  then  be  recycled,  this  has  the  potential  to  change  the  market  and  pricing  of  wood  products.  BNSF  saw  significant  opportunities  in  the  future  for  increased  use  of  rail  capacity,  as  the  carbon  benefits  of  shipping  by  rail  may  be  enhanced  by  monetized  carbon.        These  same  three  firms  also  expressed  concern  about  the  specifics  of  carbon  regulations.  Catalyst  moved  early  to  reduce  its  carbon  emissions.  If  allowances  under  a  cap  and  trade  system  are  calculated  using  an  average  of  the  previous  10  years  of  emissions,  Catalyst  could  be  in  a  position  where  further  reductions  in  order  to  comply  with  shrinking  allowances  are  almost  impossible.  In  short,  they  could  be  punished  for  good  behaviour.  BNSF  also  identified  similar  risks  with  cap  and  trade.  WFP  identified  uncertainty  around  the  measurement  of  carbon  in  forest  products  as  risky.  Harvested  timber  is  converted  to  solid  wood  products  that  store  carbon;  pulp  is  converted  to  paper  and  can  be  recycled;  and  wood  waste  can  be  used  to  offset  fossil  fuel  use.  The  methods  and  assumptions  behind  the  measurement  of  these  (and  other)  variables  impacts  the  emissions  associated  with  forest  operations.                   452.7 DISCUSSION  Based  on  the  interview  results,  I  suggest  a  framework  that  explains  how  carbon  will  transform  supply  chains.  I  consider  a  three-­‐step  process,  where:  at  first,  efficient  supply  chains  emerge  due  to  carbon’s  equivalency  with  energy;  next,  environmentally  responsible  supply  chains  emerge;  and  finally,  resilient  supply  chains  develop  as  the  risks  of  monetized  carbon  are  mitigated.      2.7.1 Efficient  Supply  Chains  There  are  two  ways  in  which  carbon  efficiency  can  transform  supply  chains.  Corporations  that  are  carbon-­‐efficient  may  become  preferred  suppliers  (a  status  achieved  by  Catalyst  with  Anonymous  Magazine)  and  gain  market  share.  Supply  chains  themselves  may  also  reorient  to  minimize  carbon  emissions,  as  seen  with  Catalyst’s  use  of  barges  and  rail  to  reduce  transportation  carbon.  These  carbon-­‐efficient  supply  chains  will  better  adapt  to  the  regulations  and  cost  structures  of  a  low-­‐carbon  economy.  The  carbon  emissions  associated  with  supply  chains  may  influence  where  business  is  conducted.    Supply  chains  may  evolve  to  locate  particularly  energy-­‐intensive  processes  (such  as  aluminum  smelting  or  paper  making)  near  low-­‐carbon  energy  sources,  such  as  hydroelectric  power  (used  by  Catalyst  to  reduce  the  carbon  footprint  of  its  paper).  Conversely,  processes  using  little  energy  may  relocate  closer  to  efficient  transportation  networks  and  major  markets,  minimizing  emissions  from  transportation  and  distribution.        2.7.2 Responsible  Supply  Chains    As  carbon  emerges  as  a  major  component  of  sustainability,  it  may  play  a  stronger  role  in  corporate  social  responsibility  (CSR)  policies.  The  prevalence  of  the  Carbon  Disclosure  Project  (CDP,  2009)  indicates  that  corporations  already  understand  this.  How  this  will  impact  supply  chains  is  less  certain.  If  carbon  continues  to  gain  importance,  products  with  large  carbon  footprints  relative  to  their  competition  may  fall  out  of  favor.  Companies  that  demonstrate  an  understanding  of  their  supply  chain  footprint,  and  steps  taken  (or  at  least  plans)  to  reduce  it,  will  benefit.      To  achieve  reliable  and  transparent  management  of  carbon,  third  party  auditing  and  verification  will  need  to  be  more  widespread.  While  costly,  the  outcomes  of  this  monitoring  may  lead  to  stronger  engagement  between  supply  chain  collaborators,  a  phenomenon  shown  in  the  case  study.  The  potential  to  audit  supply  chains  for  carbon  has  several  implications.  These  audits  will  provide  a  baseline  measurement,  and  allow  for  improvements  over  time.  Relationships  between  supply  chain  partners  that  developed  around  carbon  may  evolve  to  include  other  issues  in  sustainability.  There  are  potential               46trickle-­‐down  effects  if  the  demands  of  one  customer  change  the  behaviour  of  a  supplier.  The  supplier  in  this  case  study,  responding  to  one  customer’s  demand,  is  able  to  subsequently  provide  carbon-­‐light  products  to  all  of  its  customers.  Carbon  management  can  therefore  diffuse  along  the  supply  chain  due  to  the  requests  of  only  one  supply  chain  partner.          2.7.3 Resilient  Supply  Chains    Efficient  and  responsible  supply  chains  build  more  resilient  connections  between  supply  chain  collaborators.  These  connections,  observed  in  the  case  study,  suggest  that,  in  the  future,  carbon  will  be  considered  closely  in  risk  management.  I  identified  three  types  of  risk:  regulatory  risk;  financial  risk;  and  market  access  risk.      Regulatory  risk  involves  government  control  of  carbon.  Businesses  that  anticipate  this  control  are  in  a  less  risky  position.  Those  that  emit  large  amounts  of  carbon,  but  have  not  begun  to  adjust,  are  exposed.  An  illustrative  example  can  be  found  in  the  American  bond  market,  where  analysts  are  projecting  a  premium  on  corporate  bonds  for  new  coal-­‐fired  power  plants  (Stevenson,  2008),  reflective  of  an  anticipated  cost  of  carbon.  Regulatory  risk  also  involves  how  regulations  are  deployed.  Companies  that  have  already  made  progress  in  reducing  their  footprints  may  be  put  in  a  difficult  position,  a  danger  identified  by  both  Catalyst  and  BNSF.  Good  behaviour  already  underway  faces  the  risk  of  being  punished  by  the  definition  and  allocation  of  allowances.      Financial  risk  involves  the  ability  of  companies  to  secure  capital  in  the  long  term.  Investors  have  indicated  that  they  will  consider  carbon  in  their  investment  decisions  (CDP,  2008).  Their  reasoning  is  simple:  if  a  corporation  emits  a  lot  of  carbon,  they  will  be  obligated  to  pay  for  these  emissions.  Some  industries  cannot  avoid  emissions,  and  investors  may  require  the  disclosure  of  emissions  and  the  beginnings  of  a  carbon  management  plan.  In  other  circumstances,  investors  may  prefer  articulated  targets  and  reduction  strategies.  In  either  scenario,  carbon  may  emerge  as  an  impediment  to  securing  capital  if  emissions  are  not  managed.  Although  no  interviewees  cited  the  specific  connection  between  carbon  and  capital,  sources  such  as  the  Carbon  Disclosure  Project  (2008)  and  the  UN  Principles  for  Responsible  Investing  (2009)  support  the  idea.      Market  access  risk  has  two  components.  Understanding  carbon  emissions  may  become  mandatory  for  participating  in  supply  chains  as  businesses  seek  to  collaborate  with  partners  who  manage  their  carbon.               47Wal-­‐Mart,  for  example,  has  initiated  a  process  requiring  all  suppliers  to  measure  and  disclose  their  carbon  footprints.  Consumers  may  demand  carbon  labeling  on  products  that  they  purchase.  Although  the  appetite  to  pay  a  premium  for  sustainably  produced  goods  is  small  (Manget  et  al.  2009),  consumers  have  a  preference  for  products  that  are  cost-­‐competitive,  but  also  demonstrate  an  environmental  commitment.  Carbon,  given  its  current  prevalence,  has  the  potential  to  emerge  as  an  important  criterion  in  consumer  choice  (MacKerron  et  al.,  2009),  but  awareness  of  environmental  issues  and  demographic  variables  like  income  play  an  important  role  (Moon  et  al.,  2002;  Hansla  et  al.,  2008).      Moving  beyond  efficiency  and  responsibility,  I  found  evidence  that  supply  chains  were  becoming  more  resilient  to  risk  by  managing  carbon.  Regulatory  risk  around  carbon  encouraged  its  management.  Financial  risk  associated  with  securing  adequate  investment  capital  encouraged  carbon  management  as  well.  Finally,  market  access  risk,  with  both  consumers  and  businesses  preferring  environmentally  friendly  products,  motivated  companies  to  manage  carbon.  Together,  these  risk  management  activities  suggest  that  carbon  mechanism  that  can  help  supply  chains  become  more  resilient.    2.7.4 Implications  for  Businesses  and  Supply  Chains  Through  the  interviews,  it  was  clear  that  carbon  is  emerging  as  a  common  cause.  Less  clear  is  how  this  will  induce  change  in  purchasing  decisions,  design  of  supply  chains,  and  perceptions  of  sustainability.  Businesses  trying  to  balance  short-­‐term  costs,  long-­‐term  profitability,  and  the  maintenance  of  a  corporate  brand,  have  identified  carbon  as  a  means  for  progress  on  all  three  fronts.  Supply  chains  composed  of  different  actors  facing  different  pressures  have  been  able  to  align  corporate  strategies  around  a  common  variable.    Location  matters  if  reducing  carbon  emissions  is  a  priority.  As  the  case  study  showed,  printing  contributes  a  small  amount  to  the  total  footprint  of  a  product  and  is  best  done  close  to  markets  and  transportation  hubs  to  reduce  emissions.  Papermaking  contributes  a  large  amount  to  the  total  footprint  of  a  product  and  is  best  done  where  there  are  abundant  supplies  of  renewable  energy  and  efficient  transportation  networks.  Focusing  on  the  emissions  of  just  one  stage  potentially  ignores  the  biggest  emitters  and  the  best  opportunities  for  emission  reductions.      Supply  chains  will  evolve  to  better  reflect  the  carbon  costs  of  transportation.  Physical  locations  of  supply  chain  stages  may  change,  with  low-­‐energy  operations  relocating  to  reflect  the  carbon  costs  of               48transportation,  and  high-­‐energy  operations  moving  to  reflect  the  carbon  costs  of  energy  bottlenecks  in  a  supply  chain.  Regions  with  energy  grids  that  include  significant  access  to  renewable  energy  sources  may  become  increasingly  competitive,  while  regions  reliant  on  carbon-­‐heavy  energy  could  find  themselves  at  a  disadvantage.      At  present,  when  carbon  is  generally  without  a  price,  companies  are  finding  that  reducing  their  carbon  footprint  reduces  their  fuel  costs.  As  carbon  gains  a  price,  these  companies  will  find  other  benefits.  Not  only  will  they  save  on  fuel,  emissions  will  cost  less.  Some  businesses  already  market  a  product  on  its  carbon  footprint,  and  if  carbon  awareness  increases,  these  businesses  stand  to  benefit.  While  current  trends  in  carbon  management  are  predominantly  internal  in  orientation  (steps  to  reduce  employee  travel,  more  efficient  office  lighting,  etc.),  there  is  a  limited  scope  and  diminishing  returns  from  such  efforts.  More  sophisticated  policies  to  manage  and  reduce  emissions  will  look  at  suppliers,  logistics,  and  operations  —  in  other  words,  the  supply  chain      2.8 CONCLUSIONS  Underlying  this  discussion  on  carbon  are  some  significant  assumptions.  It  is  assumed  that  monetized  carbon  will  emerge  in  some  form.  It  is  assumed  that  businesses  will  begin  to  consider  carbon  more  carefully.  These  are  assumptions,  not  guarantees.  However,  in  the  case  study,  I  found  supply  chain  actors  embracing  carbon.  At  the  corporate  level,  energy  efficiency,  long  term  profitability,  and  responsible  branding  converged  around  the  concept  of  managing  carbon.  This  convergence  was  supported  at  the  political  level,  where  government  policies  regulating  carbon  are  present  and  evolving.  Along  the  supply  chain,  the  nature  of  collaboration  between  partners  and  the  physical  designs  of  the  supply  chain  seem  poised  to  evolve.  I  am  not  suggesting  that  carbon  is  the  only  variable  that  influences  the  composition  of  a  supply  chain,  but  I  do  think  it  is  time  to  consider  the  role  that  carbon  may  play.  This  case  demonstrates  the  influence  of  carbon  along  a  supply  chain  and  offers  a  framework  for  further  analysis  of  sustainable  supply  chains.               49Chapter  3: The  Case  of  Consumers:  Environmental  Values  and  Media  Consumption    3.1 ABSTRACT  The  consumption  of  the  written  word  is  changing.  The  printing  press,  pioneered  by  Gutenberg  in  the  1430s,  enjoyed  five  centuries  of  dominance.  But  with  the  emergence  of  the  personal  computer  and  the  Internet,  consumers  are  faced  with  digital  alternatives  to  paper  media.  With  bewildering  speed,  the  last  30  years  have  seen  computers  and  electronic  devices  proliferate.  As  telecommunications  networks  matured,  the  Internet  enabled  media  outlets  to  deliver  content  electronically.  Desktops,  laptops,  tablets,  and  smartphones  can  now  serve  the  same  function  of  printed  newspapers,  books,  and  magazines.  Every  reader  of  this  dissertation  will  have  experienced  this  change,  either  at  work  or  home.  Concurrent  to  this  media  revolution  has  been  the  emergence  of  environmental  sustainability  as  a  pressing  –  perhaps  the  pressing  –  challenge  of  our  time.  This  chapter  is  a  study  of  the  environmental  values  of  media  consumers.  I  used  a  common  method  for  studying  environmental  values  –  the  New  Ecological  Paradigm  (NEP)  –  to  investigate  the  ecological  views  of  two  groups  of  media  consumers,  those  who  consume  a  lot  of  digital  media  and  those  who  do  not.  I  arrived  at  two  key  findings:  media  consumers  are  quickly  shifting  towards  digital  sources,  and  media  consumption  habits  have  no  influence  over  environmental  values.  These  findings  affirm  the  idea  that  a  media  transition  is  under  way.  Further,  they  suggest  that  media  footprints  are  not  of  a  source  of  environmental  concern  to  consumers.    3.2 INTRODUCTION  This  chapter  focuses  on  the  views  of  consumers  in  the  midst  of  two  shifts:  the  media  shift  and  the  sustainability  shift.  I  ask  whether  consumers  are  thinking  about  the  environmental  footprint  of  their  media.  Do  they  see  their  media  consumption  habits  changing?  Do  consumers  with  different  environmental  values  have  different  consumption  habits?  To  address  questions  like  these,  I  designed  a  three-­‐part  survey  that  asked  North  American  consumers  to  report  on  their  demographic  profile,  their  environmental  values,  and  their  media  consumption  habits.  This  chapter  begins  with  a  brief  background  on  why  an  analysis  of  consumers’  environmental  values  and  media  consumption  habits  is  relevant.  It  follows  by  defining  research  objectives  and  methodologies.  Results  are  presented  in  three  sections:  demographic  trends,  environmental  values,  and  media  consumption  trends.  Finally,  I  discuss  the  implications  of  my  results  and  relate  them  back  to  the  literature.                   503.3 BACKGROUND    The  sustainability  shift  is  a  concept  present  throughout  this  thesis,  but  I  summarize  the  idea  once  more  here.  It  suggests  that  society  is  acknowledging  the  finite  capacity  of  the  planet  earth  to  withstand  environmental  degradation.  As  a  response,  sustainability  has  emerged  as  a  concept  across  the  social  spectrum.  At  its  core,  sustainability  simply  means  doing  something  better,  or  at  least  less  badly.  One  founding  definition  comes  from  the  Brundtland  Commission,  which  described  sustainability  as  meeting  “the  needs  of  the  present  without  compromising  the  ability  of  future  generations  to  meet  their  own  needs”  (United  Nations,  1987).    But  sustainability  can  mean  very  different  things  to  different  actors.  Seager  (2008)  captures  this  dilemma,  suggesting  that  traditional  methods  of  academic  inquiry  may  be  ill  suited  to  study  such  a  nebulous  concept.  It  is  worth  asking  once  more,  what  academic  would  be  best  suited  to  study  sustainability:  an  engineer,  an  ecologist,  an  economist,  or  a  political  scientist?  All  would  have  their  own  methods  and  perspectives,  none  of  which  are  necessarily  more  correct  than  the  others.  To  alleviate  this  tension,  Seager  (2008)  suggests  an  approach  to  sustainability  that  operates  conceptually  rather  than  technically,  much  like  justice  or  fairness.      Different  actors  drive  sustainability  forward  in  unique  ways.  Governments  use  incentives  and  regulations  to  promote  more  sustainable  outcomes  (Graedel  and  Allenby,  2003).  Investors  encourage  sustainability  by  limiting  their  support  to  actors  who  meet  certain  requirements  (UNEP,  2006).  Corporations  pursue  opportunities  to  create  shared  value  by  increase  profits  while  reducing  environmental  impacts  and  becoming  more  eco-­‐efficient  over  time  (Molina-­‐Azorín  et  al.,  2009).  The  marketplace  has  an  important  role  to  play  too.  Business-­‐to-­‐business  interactions  can  promote  sustainability,  with  major  buyers  forcing  their  suppliers  to  disclose  and  improve  their  environmental  footprints  (Home  Depot,  2011;  Wal-­‐Mart,  2011).  Consumers  can  drive  sustainability,  as  well,  leveraging  both  their  buying  power  and  preference  for  more  sustainable  products  (D'Souza  et  al.,  2007).  Some  suggest  (Diamantopoulos  et  al.,  2003)  that  consumers  have  a  willingness  to  change  their  consumption  habits  and  pay  a  premium  for  what  they  consider  to  be  more  sustainable  products,  but  the  extent  to  which  they  are  willing  to  change  is  difficult  to  measure  as  demographic  variables  do  not  consistently  predict  green  consumption  patterns.  The  role  that  consumers  play  in  motivating  more  sustainable  outcomes  is  the  major  reason  why  I  want  to  investigate  the  environmental  values  of  media  consumers.                   51Media,  much  like  sustainability,  is  undergoing  a  transformation.  A  shift  is  happening  from  paper  products  to  digital  alternatives.  But  it  is  an  uneven  and  complex  shift,  and  there  is  no  single  metric  that  can  capture  the  diversity  in  the  scope,  volume,  or  intensity  of  this  transition.  A  huge  variety  of  products  previously  available  only  on  paper  –  newspapers,  books,  catalogues,  magazines,  and  so  on  –  can  now  be  consumed  across  a  system  of  Internet  connected  digital  devices  like  laptops,  tablets,  and  smartphones.  This  shift  is  not  necessarily  zero-­‐sum;  an  increase  in  digital  media  consumption  does  not  necessitate  a  decrease  in  paper  media  consumption,  but  certain  product  categories  –  newspapers  and  magazines  come  to  mind  –  are  more  vulnerable  to  disruption  than  others.      Since  the  turn  of  the  millennium  and  the  advent  of  more  sophisticated  and  affordable  ICT,  a  host  of  media  shifts  have  taken  place.  Major  magazines  have  gone  bust,  no  longer  able  to  generate  sufficient  revenues  (Media  Life  Magazine,  2014).  Magazines  that  have  survived  have  had  to  supplement  their  revenue  with  digital  sales,  or  receive  subsidies  from  major  corporate  backers  (Pew  Research  Journalism  Project,  2013).  Revenues  from  classified  sections  in  newspapers  have  imploded,  as  free  services  like  Craigslist  emerge,  offering  hyper-­‐local  advertising  at  little  to  no  cost  (Edmonds,  2013).  Paper  mail  volumes  have  undergone  dramatic  changes,  as  well:  less  and  less  personal  correspondence  is  being  delivered  by  mail,  and  companies  increasingly  prefer  electronic  invoices.  As  a  result,  mail  delivery  volumes  have  returned  to  levels  not  seen  since  the  early  1980s  (United  States  Postal  Service,  2014).      Others  have  researched  this  shift,  finding  evidence  that  consumers  are  rapidly  embracing  new  media  platforms  (Duggan  and  Smith,  2013;  Bell  et  al.,  2013).  Chyi  and  Lee  (2013)  and  Chan-­‐Olmsted  et  al.  (2013)  both  conducted  surveys  of  digital  media  consumption  in  order  to  better  gauge  the  preferences,  usage,  and  intent  of  media  consumers.  Researchers  have  also  made  efforts  to  connect  media  consumption  with  environmental  issues.  Holbert  et  al.  (2003)  use  media  consumption  as  a  means  for  testing  environmental  values,  connecting  patterns  of  television  viewing  with  environmental  beliefs.  Hansen  (2010)  connects  environmental  values  with  media  in  general,  suggesting  that  how  consumers  choose  to  digest  media  has  an  important  influence  over  their  environmental  behaviour.  This  body  of  research  suggests  that  connecting  media  consumption  with  environmental  values  in  consumers  is  a  suitable  backdrop  for  exploring  sustainability  and  media.                 523.4 OBJECTIVES  Given  the  two  shifts  I  have  identified  –  the  media  shift  and  the  sustainability  shift  –  I  want  to  understand  the  interactions  between  environmental  values  and  media  consumption.  To  that  end,  I  set  out  in  this  chapter  to:  • Examine  the  media  habits  of  segments  with  different  environmental  values.  • Examine  the  environmental  values  of  different  media  segments.  • Identify  any  patterns  or  significant  trends  that  contextualizes  discussions  of  the  environmental  footprint  of  media,  and  strengthens  our  understanding  of  sustainability.    3.5 METHODOLOGY  I  delivered  a  web-­‐based  survey  to  1,435  consumers  in  North  America:  half  in  Canada,  half  in  the  United  States,  after  attaining  ethics  approval  (UBC  BREB  certificate  #H09-­‐03036).  The  survey  sample  mirrored  demographic  trends  from  national  censuses,  with  both  males  and  females  from  representative  age  groups  were  polled.  The  respondents  were  derived  from  a  survey  panel  operated  by  Market  Tools,  a  survey  provider  that  audits  and  verifies  the  age,  location,  gender,  and  income  of  survey  respondents.  As  the  survey  was  made  available  to  respondents,  controls  were  put  in  place  to  ensure  that  our  sample  mirrored  data  from  the  latest  national  censuses  from  both  Canada  and  the  United  States.  Market  Tools  verifies  the  information  provided  by  respondents  using  a  variety  of  means,  including  telephone  interviews  and  credit  score  inquiries.  It  took  the  sample  approximately  a  week  to  respond.  There  are  inherent  biases  that  are  introduced  by  choosing  an  online  survey  to  assess  questions  related  to  paper  and  pixels.  Couper  (2000)  suggests  the  proliferation  of  online  surveys  may  impact  the  quality  of  responses,  as  respondents  are  saturated  with  potential  surveys  to  fill  out  and  therefore  make  choices  based  on  personal  interest  or  entertainment  value.  There  is  a  risk  that  over-­‐surveying  can  diminish  the  quality  of  responses,  and  that  the  only  people  filling  out  online  surveys  are  those  who  are  web-­‐savvy  to  begin  with  (Couper,  2000).  However,  the  online  survey  method  allowed  us  to  target  very  specific  demographic  groups,  strengthening  the  findings  without  imposing  prohibitive  costs.    Our  survey  was  designed  after  reviewing  similar  surveys  on  media  consumption  and  environmental  variables  (Holbert  et  al.,  2003;  Dou  et  al.,  2006;  Dutta-­‐Bergman,  2004).  I  identified  questions  that  reveal  information  about  both  the  paper  and  digital  media  consumption  habits.                   53The  survey  itself  consisted  of  three  sections  (for  a  complete  version  of  the  survey  questions,  see  Appendix  B).  The  first  investigated  basic  demographic  trends,  the  second  explored  the  environmental  values  of  consumers,  and  the  third  measured  the  media  consumption  habits  of  the  segments.  The  section  on  demographic  trends  asked  simple  questions  about  age,  gender,  income,  location  and  similar  variables.      To  study  environmental  values,  I  utilized  the  widely  adopted  new  ecological  paradigm  (NEP)  scale  in  order  to  categorize  the  respondents.  Each  respondent  received  a  score  that  indicates  how  anthro-­‐centric  or  eco-­‐centric  he  or  she  is.  Originally  developed  by  Liere  &  Dunlap  (1980),  and  then  later  revised  by  Dunlap  et  al.  (2000),  the  NEP  scale  is  a  tool  frequently  employed  in  survey  methodologies  to  gauge  perceptions  of  sustainability  (see  Cordani  et  al.,  2003;  Lundmark,  2007;  Rideout  et  al.,  2005;  Schultz,  2001).  I  asked  consumers  to  answer  the  15  questions  in  Table  2  and  calculated  their  NEP  score  accordingly.  Like  all  studies  that  employ  the  NEP  scale,  the  eight  odd-­‐numbered  questions  were  worded  in  such  a  way  that  agreement  suggests  a  proecological  worldview  (a  world  view  biased  towards  the  environment).  In  contrast,  disagreement  with  the  seven  even-­‐numbered  questions  suggests  a  proecological  worldview.  The  median  NEP  score  of  the  sample  was  52.  I  then  divided  the  sample  in  half  with  717  respondents  with  an  NEP  score  of  52  or  higher  labeled  NEP  Eco.  Another  718  residents,  with  NEP  scores  of  52  or  lower,  were  labeled  NEP  Anthro.  This  basis  for  this  categorization  is  the  likelihood  that  the  NEP  Eco  segment  prioritized  environmental  issues  when  compared  to  the  NEP  Anthro  segment.                 54Table  2:  The  new  ecological  paradigm  1.  We  are  approaching  the  limit  of  the  number  of  people  the  earth  can  support    7.  Plants  and  animals  have  as  much  right  as  humans  to  exist   13.  The  balance  of  nature  is  very  delicate  and  easily  upset  2.  Humans  have  the  right  to  modify  the  natural  environment  to  suit  their  needs  8.  The  balance  of  nature  is  strong  enough  to  cope  with  the  impacts  of  modern  industrial  nations    14.  Humans  will  eventually  learn  enough  about  how  nature  works  to  be  able  to  control  it  3.  When  humans  interfere  with  nature  it  often  produces  disastrous  consequences  9.  Despite  our  special  abilities  humans  are  still  subject  to  the  laws  of  nature  15.  If  things  continue  on  their  present  course,  we  will  soon  experience  a  major  ecological  catastrophe  4.  Human  ingenuity  will  insure  that  we  do  NOT  make  the  earth  unlivable   10.  The  so-­‐called  ecological  crisis  facing  humankind  has  been  greatly  exaggerated      5.  Humans  are  severely  abusing  the  environment   11.  The  earth  is  like  a  spaceship  with  very  limited  room  and  resources      6.  The  earth  has  plenty  of  natural  resources  if  we  just  learn  how  to  develop  them  12.  Humans  were  meant  to  rule  over  the  rest  of  nature        The  section  on  media  consisted  of  a  variety  of  queries  on  consumption  habits  and  technology  ownership.  I  also  asked  the  sample  to  describe  how  their  media  consumption  habits  have  changed.  I  wanted  to  explore  whether  those  who  use  more  or  less  digital  media  have  different  environmental  values.  Because  digital  media  is  often  touted  as  a  ‘green’  alternative,  I  hypothesized  that  the  Digital  Heavy  segment  would  have  stronger  environmental  values.  To  do  so,  I  devised  a  metric  for  indicating  just  how  tech-­‐savvy  a  respondent  might  be.  The  objective  was  to  create  two  groups  –  ‘Digital  Light’  and  ‘Digital  Heavy’  –  that  represented  different  groups  of  media  consumers.  I  asked  the  respondents  several  questions,  which  when  taken  in  aggregate,  indicate  how  much  or  how  little  digital  media  a  consumer  uses  in  their  lives.  For  each  question  answered  in  the  affirmative,  a  consumer’s  digital  use  score  increased.  The  questions  used  are  summarized  in  Table  3.                   55Table  3:  Digital  use  score  How  long  have  you  been  using  the  Internet?    How  many  computers  are  used  in  your  household?    How  many  mobile  devices  (cellphone,  smartphone,  etc.)  are  used  in  your  household?  Do  you  own  any  of  the  following  devices:  Digital  Camera  Digital  Video  Camera  DVD  Player  BluRay  Player  Flat  Panel  Television  MP3  Player    Desktop  Computer  Laptop  Computer  Tablet  Computer  Cellphone  Smartphone    E-­‐Reader  If  you  get  your  news  online,  which  of  the  following  sources  do  you  use  (check  all  that  apply)?  Dedicated  news  sites  (e.g.  CNN.com,  etc.)  National  newspaper  websites  Regional  newspaper  websites  Local  newspaper  websites   News  aggregators  (e.g.  Google  News,  Yahoo  News,  etc.)  Blog  Please  rate  how  frequently  you  use  the  listed  online  services  with  the  following  scale:  Never,  Seldom,  Sometimes,  Often,  Always.  Instant  Messaging  Email  Social  Networking  Blogging  Research  Reading  the  news  Watching  videos  Online  banking  Online  shopping    The  digital  use  score  was  calculated  by  adding  up  every  instance  a  consumer  answered  in  the  affirmative  to  a  question  that  would  indicate  digital  media  consumption.  For  example,  when  I  asked,  “How  long  have  you  been  using  the  internet?”  and  the  respondent  provided  a  number  between  1  (relating  to  a  period  of  0  to  6  months)  and  5  (relating  to  a  period  of  6  years  or  more).  The  longer  the  respondent  had  been  using  the  Internet,  the  higher  their  digital  use  score  would  be.  The  same  method  was  used  for  questions  on  computers  and  mobile  devices  in  the  household.  I  asked,  “Do  you  own  any  of  the  following  devices?”  and  for  each  device  that  a  respondent  owned,  their  digital  score  would  increase  by  one  point.  It  follows  that  the  more  digital  devices  owned  by  a  respondent,  the  higher  their  digital  use  score.  I  applied  the  same  logic  and  method  to  questions  on  Online  News  Sources  and  Online  Services.  The  more  a  respondent  said  they  used  various  digital  media,  the  higher  their  digital  use  score.  While  inexact,  this  method  allowed  us  to  categorize  respondents  who  do  not  use  digital  media  as  frequently  as  others.  After  calculating  the  digital  use  score,  I  identified  the  median  score  and  divided  the  sample  into  two  segments.      Once  all  the  necessary  data  had  been  gathered,  I  applied  statistical  methods  where  appropriate.  For  questions  on  consumption  of  different  digital  media  and  paper  media  sources,  I  conducted  a  standard  t-­‐test  to  determine  statistical  significance  between  means,  with  α=  0.05.  For  the  questions  on  the  use  of  various  online  services,  I  conducted  a  z-­‐test,  in  order  to  ascertain  statistical  significance  between  proportions  with  α=  0.05.                 56  While  a  statistical  test  would  have  been  appropriate  for  the  questions  on  willingness  to  pay  for  media,  because  the  scale  was  not  an  interval  scale,  I  was  unable  to  compute  means  that  could  be  tested.  In  these  cases,  I  compared  varying  levels  of  willingness  to  pay  with  the  following  method:  the  percentage  of  respondents  unwilling  to  pay  was  subtracted  from  the  percentage  of  those  willing  to  pay,  arriving  at  a  percentage  that  could  be  either  positive  or  negative.  The  more  unwilling  to  pay  a  certain  segment  was,  the  lower  the  percentage  would  be.  I  then  converted  the  percentages  from  all  segments  and  across  all  categories  into  numerical  scale  between  1.0  (which  would  imply  100%  willing  to  pay)  and  –1.0  (implying  100%  unwilling  to  pay).  I  plotted  the  results  accordingly.    3.6 RESULTS  I  present  the  results  in  three  parts:  demographic  information,  environmental  values,  and  media  consumption  trends.  In  each  section,  I  have  presented  the  results  using  the  following  segments:  NEP  Eco,  NEP  Anthro,  Digital  Light,  Digital  Heavy.  In  some  instances,  I  aggregate  results  for  all  respondents.    3.6.1 Demographic  Trends  Our  survey  sample  (n=1,435)  had  the  following  demographic  attributes.  The  gender  breakdown  of  the  sample  set  is  described  in  Figure  5.  The  sample  was  almost  evenly  split  between  males  and  females  (49.5%  male,  50.5%  female).  When  comparing  media  consumption  patterns,  males  were  more  likely  to  be  Digital  Heavy  than  Digital  Light,  but  by  a  very  small  percentage  (2.4%).  On  environmental  variables  males  were  more  likely  to  be  NEP  Anthro  by  a  margin  of  6.3%.  Figure  5:  Gender  breakdown  of  respondents  by  segment    49.5%  46.6%  52.9%  48.3%  50.7%  50.5%  53.4%  47.1%  51.7%  49.3%  All  NEP  Eco  NEP  Anthro  Digital  Light  Digital  Heavy  Male   Female               57I  also  asked  respondents  to  provide  their  age.  The  results  are  presented  in  Figure  6.  The  Digital  Heavy  segment  tended  to  be  younger  (mean  age:  41.7  years)  than  the  Digital  Light  segment  (mean  age:  52.1  years).  The  NEP  Anthro  and  eco  segments  had  similar  mean  ages  (respectively  46.6  and  46.8  years).  These  figures  are  similar  to  the  mean  for  the  entire  sample  (46.8  years).    Figure  6:  Age  breakdown  of  respondents  by  segment      I  asked  respondents  to  describe  their  educational  achievements,  and  the  results  can  be  found  in  Figure  7.  I  found  that  the  NEP  Anthro  and  NEP  Eco  segments  had  very  similar  educational  profiles.  The  Digital  Light  and  Digital  Heavy  segments  did  exhibit  some  differences.  The  Digital  Heavy  segment  was  the  group  most  likely  to  have  attained  a  Bachelor  degree  or  higher,  and  the  least  likely  to  have  not  completed  high  school.    0.0%   10.0%   20.0%   30.0%   40.0%   50.0%   60.0%   70.0%   80.0%   90.0%   100.0%  All  NEP  Eco  NEP  Anthro  Digital  Light  Digital  Heavy  All   NEP  Eco   NEP  Anthro   Digital  Light   Digital  Heavy  >  19   2.0%   1.7%   2.5%   1.5%   2.8%  20  -­‐  29   15.5%   14.0%   17.1%   9.2%   21.7%  30  -­‐  39   20.5%   20.4%   20.8%   14.7%   26.1%  40  -­‐  49   19.1%   19.7%   18.7%   19.3%   18.3%  50  -­‐  59   18.5%   21.8%   15.0%   19.9%   17.4%  60  -­‐  69   12.2%   11.7%   12.6%   15.8%   8.8%  70  -­‐  79   9.9%   8.8%   10.5%   15.3%   4.5%  <  80   2.4%   1.8%   2.8%   4.2%   0.4%               58Figure  7:  Education  breakdown  of  respondents  by  segment      Respondents  were  asked  to  describe  their  household  income.  The  NEP  Anthro  and  NEP  Eco  segments  reported  similar  income  levels  that  were  also  inline  with  trends  for  the  sample  as  a  whole.  The  Digital  Heavy  segment  reported  the  highest  income  levels  overall,  with  51.54%  of  respondents  reporting  a  household  income  of  $50,000  or  above  compared  to  35.61%  for  Digital  Light  respondents.    0.0%   25.0%   50.0%   75.0%   100.0%  All  NEP  Eco  NEP  Anthro  Digital  Light  Digital  Heavy  All   NEP  Eco   NEP  Anthro   Digital  Light   Digital  Heavy  Less  than  High  School   2.7%   2.1%   3.2%   4.2%   1.3%  High  School  /  GED   20.3%   19.3%   21.4%   26.8%   13.8%  Some  College   39.0%   38.8%   39.2%   39.4%   39.1%  Bachelor's  Degree   25.4%   27.4%   23.3%   18.6%   31.1%  Master's  Degree   9.5%   9.4%   9.8%   8.4%   11.2%  Doctoral  Degree   0.8%   0.8%   0.7%   0.8%   0.6%  Professional  Degree   2.3%   2.2%   2.4%   1.8%   2.9%               59Figure  8:  Income  breakdown  of  respondents  by  segment      I  asked  respondents  to  describe  the  area  in  which  they  live,  with  Urban,  Suburban,  and  Rural  as  options.  For  the  entire  sample,  33.8%  were  urban,  41.6%  were  suburban,  and  24.5%  were  rural.  The  NEP  Anthro  and  NEP  Eco  segments  answered  similarly,  but  the  Digital  Heavy  segment  was  more  likely  to  be  urban  (34.5%)  or  suburban  (44.3%)  when  compared  to  the  Digital  Light  segment  (33.2%  urban;  39.2%  suburban).    3.6.2 Environmental  Values  Our  results  on  environmental  values  are  presented  in  Figure  9,  showing  the  frequency  distribution  of  NEP  Scores  for  the  Digital  Light  and  Digital  Heavy  segments.  The  Digital  Heavy  segment  tends  to  have  higher  NEP  scores,  suggesting  that  the  segment  has  stronger  environmental  values.  But  the  difference  is  slight:  the  mean  NEP  score  of  the  Digital  Heavy  segment  is  52.4,  while  the  Digital  Light  segment  has  a  0.0%   25.0%   50.0%   75.0%   100.0%  All  NEP  Eco  NEP  Anthro  Digital  Light  Digital  Heavy  All   NEP  Eco   NEP  Anthro   Digital  Light   Digital  Heavy  Under  $10,000   3.1%   2.5%   3.8%   4.2%   2.0%  $10,000  to  $19,999   7.7%   8.4%   6.8%   9.1%   6.7%  $20,000  to  $29,999   10.8%   10.9%   10.6%   12.2%   9.9%  $30,000  to  $39,999   11.1%   11.6%   10.5%   12.4%   9.9%  $40,000  to    $49,999   11.7%   10.8%   12.7%   11.9%   11.7%  $50,000  to  $74,999   21.1%   22.6%   19.8%   19.0%   22.9%  $75,000  to  $99,999   11.3%   10.5%   12.2%   8.4%   13.5%  $100,000  to  $150,000   7.7%   7.1%   8.2%   5.9%   9.6%  Over  $150,000   4.0%   4.5%   3.6%   2.4%   5.4%  Would  rather  not  say   11.4%   11.2%   11.7%   14.7%   8.2%               60mean  score  of  51.7.  A  test  reveals  a  p-­‐value  of  0.161,  suggesting  the  difference  in  means  is  not  of  statistical  significance.    Figure  9:  Frequency  distribution  of  NEP  scores  by  digital  use  segment            I  asked  the  following:  “Many  environmental  issues  involve  difficult  trade-­‐offs  with  the  economy.  Which  of  the  following  statements  best  describes  your  view?”  and  presented  respondents  with  five  options.  The  results  are  summarized  in  Figure  10.  The  NEP  Eco  segment  had  the  highest  percentage  of  respondents  who  thought  the  environment  should  be  given  the  highest  priority  (13.0%),  while  the  NEP  Anthro  segment  had  the  smallest  number  of  respondents  (8.4%)  who  felt  the  environment  should  be  a  priority.  The  NEP  Anthro  segment  was  also  the  most  likely  to  prioritize  the  economy.  The  Digital  Heavy  segment  had  stronger  environmental  values  than  its  Digital  Light  counterpart:  62.3%  felt  that  the  environment  was  more  important  than  the  economy,  compared  to  56.0%  of  the  Digital  Light  segment.    0  10  20  30  40  50  60  70  15  18  21  24  27  30  33  36  39  42  45  48  51  54  57  60  63  66  69  72  75  #  of  respondents  NEP  score  Digital  Light   Digital  Heavy  Median  Line  NEP  Anthro   NEP  Eco               61Figure  10:  Environmental  trade  offs  breakdown  of  respondents  by  segment      Our  final  environmental  question  asked  respondents,  “Do  you  believe  that  I  have  a  responsibility  to  look  out  for  the  interests  of  future  generations,  even  if  it  means  making  ourselves  worse  off?”  The  Digital  Heavy  and  Light  segments  produced  almost  identical  answers.  The  NEP  segments  responded  in  a  predictable  fashion:  only  4.6%  of  the  NEP  Eco  segment  felt  that  present  needs  outweighed  future  needs,  compared  to  19.8%  of  the  NEP  Anthro  segment.    Figure  11:  Environmental  outlook  breakdown  of  respondents  by  segment    All  NEP  Eco  NEP  Anthro  Digital  Light  Digital  Heavy  (a)  The  highest  priority  should  be  given  to  protec?ng  the  environment,  even  if  it  hurts  the  economy.  (b)  Both  the  environment  and  the  economy  are  important,  but  the  environment  should  come  first.  (c)  Both  the  environment  and  the  economy  are  important,  but  the  economy  should  come  first.  (d)  The  highest  priority  should  be  given  to  economic  considera?ons  such  as  jobs,  even  if  it  hurts  the  environment.  (e)  No  opinion.  87.7%  95.4%  80.2%  87.8%  88.1%  12.3%  4.6%  19.8%  12.2%  11.9%  All  NEP  Eco  NEP  Anthro  Digital  Light  Digital  Heavy  Yes  -­‐  Responsible  for  future  genera?ons  at  our  own  expense.    No  -­‐  Not  responsible  for  future  genera?ons  at  our  own  expense.                 623.6.3 Media  Consumption  Trends  In  this  section,  I  describe  the  media  consumption  trends,  organized  by  the  four  defined  segments.  I  began  by  asking  users  to  describe  the  type  of  Internet  connection  they  use  at  home.  In  all  but  one  of  the  categories,  80%  or  more  of  respondents  reported  that  they  had  DSL  (digital  subscriber  line)  or  cable  broadband.  The  only  exception  was  the  Digital  Light  segment:  76.1%  of  respondents  said  they  had  DSL  or  cable.  In  contrast,  the  Digital  Heavy  segment  had  87.6%  of  respondents  reporting  DSL  or  cable  broadband  at  home.      I  followed  by  asking  the  respondents  about  the  online  news  sources  they  consult,  with  full  results  summarized  in  Figure  12  and  Figure  13.  When  comparing  the  NEP  segments,  the  only  category  to  exhibit  a  statistically  significant  difference  was  news  aggregators,  where  NEP  Eco  used  that  service  more.  In  the  digital  categories,  Digital  Heavy  used  all  online  services  in  statistically  significant  numbers.  Of  note,  a  large  number  of  respondents  in  both  segments  reported  not  reading  news  online  at  all.    Figure  12:  Digital  media  sources  –  NEP  segments    0.0%  10.0%  20.0%  30.0%  40.0%  50.0%  Dedicated  news  sites    (e.g.  CNN.com)  Na?onal  newspaper  websites  Regional  newspaper  websites  Local  newspaper  websites  News  aggregators  (e.g.  Google  News)  *  Blogs   Don’t  read  news  online  *  significant  at  α  =  0.05  NEP  Eco   NEP  Anthro               63Figure  13:  Digital  media  sources  –  digital  segments      I  also  asked  the  respondents  to  report  on  the  types  of  paper  media  that  they  consume,  summarized  in  Figure  14  and  Figure  15.  I  found  that  the  NEP  Eco  segment  was  more  likely  to  consume  local  newspapers,  magazines,  and  books.  The  other  paper  news  sources  showed  no  significant  variation  between  NEP  Eco  and  NEP  Anthro.  When  I  compared  the  paper  media  habits  of  Digital  Heavy  and  Light,  I  found  statistically  significant  differences  in  several  categories:  local  newspapers,  hobbyist  or  niche  magazines,  and  books  (purchased  or  loaned).  In  each  of  those  categories,  Digital  Heavy  consumed  more  paper-­‐based  media.    0.0%  10.0%  20.0%  30.0%  40.0%  50.0%  60.0%  Dedicated  news  sites    (e.g.  CNN.com)  *  Na?onal  newspaper  websites  *  Regional  newspaper  websites  *  Local  newspaper  websites  *  News  aggregators  (e.g.  Google  News)    *  Blogs  *   Don’t  read  news  online  *  *  significant  at  α  =  0.05  Dig  Light   Dig  Heavy               64Figure  14:  Paper  media  sources  –  NEP  segments      Figure  15:  Paper  media  sources  –  digital  segments      0.0%  20.0%  40.0%  60.0%  80.0%  Na?onal  newspapers  Regional  newspapers  Local  newspapers  *  News  magazines  Entertainment  magazines  Hobbyist  or  niche  magazines  *  Books  (purchased)  *  Books  (loaned  from  library)  *  Don’t  read  paper  based  media  *  significant  at  α  =  0.05  NEP  Eco   NEP  Anthro  0.0%  20.0%  40.0%  60.0%  80.0%  Na?onal  newspapers  Regional  newspapers  Local  newspapers  *  News  magazines  Entertainment  magazines  Hobbyist  or  niche  magazines  *  Books  (purchased)  *  Books  (loaned  from  library)  *  Don’t  read  paper  based  media  *  significant  at  α  =  0.05  Dig  Light   Dig  Heavy               65I  asked  the  respondents  to  give  a  sense  of  how  often  they  use  a  variety  of  online  services,  describing  their  usage  on  a  scale  of  one  to  five,  one  being  never  and  five  being  often.  I  then  took  the  mean  for  each  category  and  have  summarized  the  results  in  Figure  16  and  Figure  17.  For  each  category,  a  paired  t-­‐test  was  conducted  to  compare  the  average  response  of  the  segments.  With  the  NEP  segments,  I  found  statistically  significant  differences  in  the  usage  rates  of  email,  blogging,  and  research.  In  the  digital  segments,  the  Digital  Heavy  segment  had  a  statistically  significantly  higher  mean  in  all  categories.  For  each  category,  the  average  response  of  each  segment  is  shown  on  a  scale  of  1.0  to  5.0.  Respondents  used  the  following  scale:  1  equals  never,  2  equals  seldom,  3  equals  sometimes,  4  equals  often,  and  5  equals  always.      Figure  16:  Online  services  –  NEP  segments      2.6  4.7  2.8  1.6  3.7  3.2  2.7  3.4  2.8  2.5  4.6  2.7  1.7  3.4  3.1  2.6  3.4  2.8  1.0  2.0  3.0  4.0  5.0  Instant  messaging  Email*   Social  networking  Blogging*   Research*   Read  news   Watch  videos  Banking   Shopping  *  denotes  p  value  <  0.025  NEP  Eco   NEP  Anthro               66Figure  17:  Online  services  –  digital  segments      I  asked  respondents  to  describe  time  spent  on  various  activities:  browsing  the  Internet,  reading  a  newspaper,  reading  a  magazine,  and  reading  a  book.  The  findings  are  summarized  by  segment  in  Table  4.  I  observed  that  the  NEP  segments  had  similar  media  consumption  habits.  Across  the  24  categories  (four  media  activities  and  six  possible  responses  on  time  spent),  any  differences  were  slight,  with  all  the  responses  within  ±  6%  of  each  other.  I  also  found  that  in  17  of  the  24  categories,  the  NEP  Anthro  segment  reported  spending  more  time  with  media,  although  by  only  a  small  percentage  (between  0.1%  and  3.8%).  When  I  compared  the  Digital  Light  and  Digital  Heavy  segments,  I  found  that  Digital  Heavy  respondents  tended  to  spend  more  time  consuming  media  in  all  categories.  Digital  Light  respondents  were  more  likely  to  spend  no  time  reading  a  newspaper,  magazine,  or  book  during  the  day.        1.9  4.5  2.0  1.2  3.1  2.6  2.1  2.8  2.3  3.1  4.8  3.5  2.1  4.0  3.8  3.3  4.0  3.3  1.0  2.0  3.0  4.0  5.0  Instant  messaging*  Email*   Social  networking*  Blogging*   Research*   Read  news*   Watch  videos*  Banking*   Shopping*  *  denotes  p  value  <  0.025  Dig  Light   Dig  Heavy               67Table  4:  Media  activities     Browsing  the  Internet   Reading  a  newspaper   Reading  a  magazine   Reading  a  book       NEP     Eco   Anthro   Eco   Anthro   Eco   Anthro   Eco   Anthro  None   0.6%   2.0%   23.0%   25.0%   35.9%   34.1%   24.9%   28.6%  Less  than  1  hour   20.7%   23.5%   59.1%   52.9%   53.8%   48.5%   30.7%   33.2%  1  to  2  hours   32.8%   29.9%   15.4%   16.1%   8.8%   12.3%   30.6%   24.3%  2  to  4  hours   23.9%   26.1%   2.1%   4.3%   1.3%   2.9%   10.6%   10.2%  4  to  6  hours   13.3%   9.6%   0.3%   0.8%   0.3%   1.1%   2.0%   2.2%  6  or  more  hours   8.8%   8.9%   0.1%   0.8%   0.0%   1.1%   1.3%   1.4%                    Digital     Light   Heavy   Light   Heavy   Light   Heavy   Light   Heavy  None   2.5%   0.0%   29.6%   19.0%   43.9%   26.5%   33.8%   20.0%  Less  than  1  hour   33.7%   10.6%   52.2%   59.2%   44.8%   56.8%   28.1%   35.2%  1  to  2  hours   35.3%   27.5%   15.8%   15.9%   9.6%   11.6%   26.4%   28.6%  2  to  4  hours   16.6%   33.0%   2.1%   4.1%   1.5%   2.7%   8.7%   12.3%  4  to  6  hours   6.6%   16.6%   0.1%   1.0%   0.0%   1.4%   2.0%   2.4%  6  or  more  hours   5.3%   12.3%   0.1%   0.8%   0.1%   1.0%   1.1%   1.5%    I  investigated  willingness  to  pay  for  different  media  categories.  Results  are  shown  for  the  NEP  segments  (Figure  18)  and  the  digital  segments  (Figure  19).  For  each  product  category,  I  asked  respondents  whether  or  not  they  would  be  willing  to  pay  to  obtain  a  product.  I  then  calculated  a  score  for  each  category  by  adding  +1  for  each  Yes  response,  0  for  each  Maybe,  and  -­‐1  for  each  No.  The  final  score  represents  an  aggregate  level  of  willingness  to  pay  for  a  product.  The  NEP  segments  behaved  similarly.  But  comparing  the  digital  segments  revealed  that  Digital  Heavy  respondents  were  more  willing  to  pay  for  all  product  categories,  particularly  printed  magazines  and  printed  books.                 68Figure  18:  Willingness  to  pay  –  NEP  segments      Figure  19:  Willingness  to  pay  –  digital  segments      -­‐1.0  -­‐0.8  -­‐0.5  -­‐0.3  0.0  0.3  0.5  0.8  1.0  Na?onal  newspaper  Regional  newspaper  Local  newspaper  Printed  magazine  Printed  book  Online  newspaper  Online  magazine  Online  blogs   Electronic  books  <-­‐-­‐-­‐        Less  Willing                              More  Willing        -­‐-­‐-­‐>  NEP  Eco   NEP  Anthro  -­‐1.0  -­‐0.8  -­‐0.5  -­‐0.3  0.0  0.3  0.5  0.8  1.0  Na?onal  newspaper  Regional  newspaper  Local  newspaper  Printed  magazine  Printed  book  Online  newspaper  Online  magazine  Online  blogs   Electronic  books  <-­‐-­‐-­‐        Less  Willing                                          re  Willing        -­‐-­‐-­‐>  Digital  Light   Digital  Heavy               69I  asked  respondents  about  their  perceived  credibility  of  different  media  sources:  printed  newspapers,  printed  magazines,  printed  books,  online  newspapers,  online  magazines,  online  news  aggregators,  and  blogs.  Respondents  ranked  each  source  on  a  scale  that  ranged  from  Not  Credible  to  Somewhat  Credible  to  Credible  to  Very  Credible.  Across  all  segments,  the  most  credible  news  sources  were  printed:  64.2%  of  printed  books  were  considered  Credible  or  Very  Credible  followed  by  61.3%  of  printed  newspapers.  Of  digital  media,  online  newspapers  ranked  the  highest  (51.2%  by  the  same  measurement),  while  online  news  aggregators  (33.4%)  and  blogs  (9.6%)  were  considered  the  least  credible  media  types.      Respondents  also  reported  on  perceived  effectiveness  of  various  advertising  types:    Newspapers  display     Online  banner    Newspapers  classifieds   Online  pop-­‐up    Newspapers  flyer  Insert   Online  text-­‐based    Magazine  display     Online  classified  (e.g.  Craigslist)  I  asked  them  to  rank  the  effectiveness  of  each  on  a  scale  that  ranged  from  Not  Effective  to  Somewhat  Effective  to  Effective  to  Very  Effective.  I  summed  the  latter  two  groups  and  ranked  the  categories.  The  survey  group  identified  newspaper  flyer  inserts,  newspaper  classifieds  and  magazine  display  advertisements  as  the  most  effective.  Online  advertisements,  in  contrast,  were  considered  less  effective  than  any  paper-­‐based  counterpart.  Online  pop-­‐ups,  banners,  and  text-­‐based  ads  were  the  least  effective:  respectively,  70.4%,  44.7%,  and  42.2%  found  them  completely  ineffective.      Finally,  I  asked  respondents  to  gauge  how  their  consumption  patterns  change  over  time  across  a  variety  of  product  categories.  I  measured  respondents  perceived  consumption  levels  in  the  past  (five  years  ago),  present  (today),  and  future  (five  years  from  today).  The  findings  are  presented  in  Figure  20  through  Figure  25.  There  are  trends  that  emerge  across  all  the  figures.  The  first  is  of  paper  media’s  decline  across  all  categories,  with  only  the  library  serving  as  the  exception  (all  segments  reported  the  noble  aspiration  of  visiting  the  library  more  frequently  in  the  future).  The  other  is  increasing  digital  media  consumption.  In  all  categories  across  all  segments,  respondents  expected  to  consume  more  digital  media.  And  in  every  category,  NEP  Eco  outpaced  NEP  Anthro  in  adopting  digital  media,  although  by  a  very  small  margin.  Comparing  the  Digital  Light  and  Digital  Heavy  segments  revealed  similar  trends:  paper  media  in  decline  and  digital  media  on  the  rise,  but  Digital  Heavy  respondents  reported  a  more  rapid  shift  towards  digital  media.                 70Figure  20:  Media  habits:  newspaper  or  magazine  delivered  to  home  NEP  Eco   NEP  Anthro      Digital  Light   Digital  Heavy        Figure  21:  Media  habits:  books  purchased  from  bookstore  NEP  Eco   NEP  Anthro      Digital  Light   Digital  Heavy                   71Figure  22:  Media  habits:  books  borrowed  from  library  NEP  Eco   NEP  Anthro      Digital  Light   Digital  Heavy        Figure  23:  Media  habits:  reading  news  on  PC  or  laptop  computer  NEP  Eco   NEP  Anthro      Digital  Light   Digital  Heavy                   72Figure  24:  Media  habits:  read  news  on  smartphone  or  mobile  device  NEP  Eco   NEP  Anthro      Digital  Light   Digital  Heavy        Figure  25:  Media  habits:  books  purchased  for  electronic  reader  NEP  Eco   NEP  Anthro      Digital  Light   Digital  Heavy                   733.7 DISCUSSION  I  sought  to  answer  two  questions:  what  are  the  media  habits  of  consumers  with  different  environmental  values?  And  what  are  the  environmental  values  of  different  kinds  of  media  consumers?  I  established  four  segments  of  respondents:  NEP  Eco  and  NEP  Anthro,  and  Digital  Light  and  Digital  Heavy.  I  compared  each  pair  of  segments,  looking  for  patterns  or  trends  that  could  contextualize  knowledge  of  media  consumption  and  the  environment.      The  NEP  segments  did  not  differ  significantly  on  demographic  variables:  age,  income,  and  gender  were  all  closely  aligned.  The  NEP  segments  also  exhibited  predictable  environmental  values:  the  NEP  Eco  segment  answered  questions  about  the  environment  with  stronger  proecological  bias  than  the  NEP  Anthro  segment.  These  findings  were  in  line  with  previous  research  that  uses  the  NEP  (Cordani  et  al.,  2003;  Lundmark,  2007;  Rideout  et  al.,  2005;  Schultz,  2001).  But  when  I  measured  media  consumption,  the  NEP  segments  followed  similar  patterns.  Rarely  did  responses  differ  by  significant  amounts,  with  one  outlier:  the  NEP  Anthro  segment  apparently  spends  more  time  blogging,  an  anomaly  amongst  otherwise  consistent  media  habits.      The  digital  segments  shared  similar  environmental  values.  Their  average  NEP  scores  were  not  statistically  different,  although  the  Digital  Heavy  segment  showed  a  slight  bias  towards  pro-­‐ecological  views  across  all  questions.  Where  the  digital  segments  differed  was  in  demographic  criteria.  The  Digital  Heavy  segment  was  younger,  more  urban,  and  had  higher  income  and  education  levels.    When  I  compared  the  media  consumption  habits  of  the  Digital  Heavy  and  Digital  Light  segments,  a  few  results  were  noteworthy.  While  the  Digital  Heavy  segment  predictably  reported  consuming  more  digital  media,  it  also  had  a  stronger  relationship  with  paper  media:  the  segment  reported  consuming  more  local  newspapers,  niche  magazines,  and  purchased  and  borrowed  more  books.  Digital  Heavy  respondents  reported  spending  more  time  with  most  forms  of  paper  media,  and  were  also  more  likely  to  indicate  a  willingness  to  pay  for  paper  products,  printed  books  in  particular.  I  also  noticed  that  all  segments  reported  noticeable  differences  in  how  they  perceived  paper  and  digital  media.  Paper  media  sources  were  reported  to  be  more  credible  sources  of  information  and  paper-­‐based  advertising  was  seen  as  more  effective  than  digital  equivalents.  The  four  segments  also  all  expressed  a  stronger  willingness  to  pay  for  paper  media.                   74And  yet,  I  also  observed  all  of  the  segments  reporting  a  shift  away  from  paper  media,  in  line  with  trends  observed  by  others  (Duggan  and  Smith,  2013;  Bell  et  al.,  2013).  Newspapers  and  magazines  were  products  that  the  literature  identified  as  particularly  vulnerable  to  disruption  (Media  Life  Magazine,  2014;  Pew  Research  Journalism  Project,  2013).  And  all  of  the  survey  segments  reported  using  more  in  the  past  than  they  do  today,  and  an  intention  to  consume  less  in  the  future,  verifying  what  the  literature  has  suggested.  All  segments  also  reported  consuming  more  books  (either  purchased  or  borrowed)  in  the  past  than  they  do  today.  They  also  all  reported  aspiring  towards  more  book  consumption  in  the  future.      More  dramatic  than  reported  changes  in  paper  consumption,  however,  were  the  consistent  responses  from  all  segments  on  changing  digital  media  habits.  All  segments  reported  that  in  the  future  they  would  consume  more  news  online,  use  mobile  devices  more  often,  and  even  start  using  an  electronic  reader.  They  also  reported  consuming  more  types  of  digital  media  today  than  they  did  in  the  past.  These  observations  are  in  line  with  other  surveys  of  consumers  (Chyi  and  Lee,  2013;  Chan-­‐Olmsted  et  al.,  2013)  that  have  identified  a  shift  in  media  consumption  habits  from  paper  to  digital  sources.  This  digital  shift  underlies  another  trend:  however  they  feel  about  the  environment,  consumers  are  shifting  towards  digital  media.  And  despite  what  people  consumers  think  about  paper  media,  they  are  also  anticipating  a  shift  towards  digital.  It  follows,  then,  that  environmental  values  have  little  influence  over  media  consumption.  So  in  the  midst  of  a  media  shift,  the  environment  seems  to  be  playing  little  or  no  role  at  all.      Our  results  also  revealed  some  features  of  the  NEP  score.  While  it  was  a  reliable  indicator  of  environmental  values  –  as  the  NEP  segments  strongly  correlated  with  pro-­‐ecological  views  in  other  environmental  questions  –  the  two  NEP  segments  showed  similar  demographic  patterns.  The  NEP  scale  is,  no  doubt,  a  valid  indicator  of  environmental  values,  befitting  a  methodology  that  has  been  extensively  used  by  other  researchers  (Cordani  et  al.,  2003;  Lundmark,  2007;  Rideout  et  al.,  2005;  Schultz,  2001).  But  the  fact  that  the  NEP  segments  did  not  have  significantly  different  demographic  variables  was  a  puzzle.  Some  research  (Grossman  and  Krueger,  1995)  has  suggested  that  income  and  the  environment  are  related,  but  the  results  do  not  support  the  idea  that  higher  income  and  education  achievement  results  in  stronger  environmental  values,  although  I  did  not  attempt  to  explicitly  test  that  hypothesis.                   75Our  research  establishes  that  media  consumption  is  shifting  from  paper  to  digital  media.  In  the  midst  of  that  shift,  I  found  two  key  trends.  Demographic  variables  –  like  age,  income,  and  education  –  tend  to  influence  media  consumption,  but  not  environmental  values.  Environmental  values  do  not  seem  to  influence  media  consumption  habits.  If  there  is  no  connection  between  environmental  values  and  media  consumption,  it  could  be  that  consumers  who  are  concerned  about  the  environment  aren’t  applying  those  values  to  media  consumption.  Because  media  is  omnipotent  and  transient  (digital  appears  virtually;  paper  can  be  delivered,  and  disposed  of  when  no  longer  needed),  perhaps  the  materiality  of  media  footprints  eludes  consumers’  concern.      This  disconnect  should  be  a  source  of  concern.  The  environmental  footprint  of  paper  was  improved  by  the  concerns  of  consumers:  efforts  to  use  better  inks,  save  important  forests,  and  avoid  harmful  bleaches  were  all  driven  by  consumer’s  worries.  Paper  media  has  been  the  recipient  of  environmental  scrutiny.  But  I  have  established,  and  the  literature  has  confirmed  that  paper  media  is  in  decline  (Duggan  and  Smith,  2013;  Duggan  and  Brenner,  2013).  New  digital  platforms  are  proliferating  rapidly  and  media  is  increasingly  virtual  and  dispersed  (Bell  et  al.,  2013).  The  imperative  of  environmental  sustainability  suggests  that  environmental  values  should  be  applied  to  all  media  types  to  ensure  that  environmental  impacts  are  documented  and  reduced.      3.8 CONCLUSIONS  In  this  chapter,  I  had  three  objectives:  to  examine  media  habits  of  different  environmental  segments;  to  examine  the  environmental  values  of  different  media  segments;  and  to  identify  patterns  in  these  segments  that  might  deepen  an  understanding  of  sustainability  and  media.  I  found  that  media  consumption  patterns  had  little  influence  over  environmental  values:  people  consuming  large  amounts  of  paper  media  or  large  amounts  of  digital  media  did  not  vary  much  in  their  ecological  bias.  I  also  found  that  stronger  or  weaker  environmental  values  did  not  consume  media  in  significantly  different  ways.  My  interpretation  of  these  results  suggests  that  media  consumption  is  a  consumptive  choice  that  seems  to  have  escaped  the  environmental  scrutiny.  Media,  both  paper  and  digital,  has  characteristics  that  make  it  intangible  and  transient.  Consumers,  are  a  result,  may  be  disconnected  from  media  footprints,  playing  a  weaker  role  in  driving  sustainable  media  than  they  otherwise  could.                 76Chapter  4: The  Case  of  Life  Cycle  Comparisons:  Efforts  to  Measure  Paper  and  Digital  Media8    4.1 ABSTRACT  The  consumption  of  the  written  word  is  changing,  as  media  transitions  from  paper  products  to  digital  alternatives.    I  reviewed  the  life  cycle  assessment  (LCA)  research  literature  that  compared  the  environmental  footprint  of  digital  and  paper  media.  To  validate  the  role  of  context  in  influencing  LCA  results,  I  assessed  LCAs  that  did  not  compare  paper  and  print,  but  focused  on  a  product  or  component  that  is  part  of  the  Information  and  Communication  Technology  (ICT)  sector.    Using  a  framework  that  identifies  problems  in  LCA  conduct,  I  assessed  whether  the  comparative  LCAs  were  accurate  expressions  of  the  environmental  footprints  of  paper  and  print.  I  hypothesized  that  the  differences  between  the  product  systems  that  produce  paper  and  digital  media  weaken  LCA’s  ability  to  compare  environmental  footprints.  I  also  hypothesized  that  the  characteristics  of  ICT  as  an  industrial  sector  weaken  LCA  as  an  environmental  assessment  methodology.  I  found  that  existing  comparative  LCAs  offered  problematic  comparisons  of  paper  and  digital  media  for  two  reasons  –  the  stark  material  differences  between  ICT  products  and  paper  products,  and  the  unique  characteristics  of  the  ICT  sector.  I  suggested  that  the  context  of  the  ICT  sector,  best  captured  by  the  concept  of  “Moore’s  Law”,  will  continuously  impede  the  ability  of  the  LCA  methodology  to  measure  ICT  products.    4.2 INTRODUCTION  The  consumption  of  the  written  word  is  changing.  Newspapers,  magazines,  books  and  other  paper  products  are  being  replaced  by  a  complex  system  of  interconnected  electronic  devices.  The  nature  and  pace  of  this  transition  is  uneven.  Some  paper  media  products  and  publishers  may  survive,  while  others  will  adapt  or  disappear.  Given  the  importance  of  sustainability  –  broadly  defined  as  activities  that  do  not  compromise  the  well-­‐being  of  future  generations  (United  Nations,  1987)  –  it  is  worth  considering  the  implications  of  a  shift  from  paper  to  digital  media  from  an  environmental  perspective.      The  Internet  and  Information  and  Communication  Technologies  (ICT)  are  transforming  the  profile  of  the  global  economy  and  impacting  the  environment.  ICT  includes  technologies  such  as  desktop  and  laptop  computers,  smartphones,  e-­‐readers,  software,  peripherals,  and  connections  to  the  Internet  that  fulfill                                                                                                                            8  A  version  of  this  chapter  was  peer-­‐reviewed  and  published  as  follows:  J.  G.  Bull  and  R.  A.  Kozak.  (2014).  Comparative  life  cycle  assessments:  The  case  of  paper  and  digital  media.  Environmental  Impact  Assessment  Review.  February  (1),  10-­‐18.               77information  processing  and  communications  functions.  The  impact  of  ICT  on  the  global  economy  is  complex.  Berkhout  and  Hertin  (2004,  p.903)  studied  the  direct,  indirect,  and  structural  impacts  of  the  ICT  sector.  They  found  that  the  sector  and  its  impacts  are  “complex,  interdependent,  deeply  uncertain  and  scale-­‐dependent.”  Hilty  et  al.  (2006,  p.1618)  worried  that  “there  is  some  risk  that  ICT  will  become  counterproductive  with  regard  to  environmental  sustainability.”  They  encouraged  a  systematic  view  of  ICT  to  ensure  its  application  is  used  in  support  of  sustainable  development.  Williams  (2011,  p.354)  argues  for  a  broad  view  on  the  impacts  of  ICT,  suggesting  that  the  “energetically  expensive  manufacturing  process,  and  the  increasing  proliferation  of  devices  needs  to  be  taken  into  account.”  Andrae  and  Anderson  (2010)  found  that  not  all  LCAs  of  ICT  products  are  created  equally.  They  found  desktop  and  laptop  LCAs  to  be  the  least  consistent  of  the  consumer  products  that  they  examined,  rooted  in  subjective  choices  and  different  system  boundaries  and  lifetimes.  Malmodin  et  al.  (2010)  found  that,  in  2007,  the  ICT  sector  produced  1.3%  of  global  greenhouse  gas  emissions  and  used  3.9%  of  global  electricity.  Given  the  growth  of  the  ICT  sector  since  2007,  this  figure  has  likely  increased.      4.3 BACKGROUND  Given  the  emergence  of  ICT  and  its  potential  to  disrupt  various  sectors  of  the  economy,  researchers  have  studied  the  potential  environmental  trade-­‐offs.  Researchers  have  compared  traditional  and  web-­‐based  retailing  (Edwards  et  al.,  2010),  working  at  the  office  or  at  home  (Mokhtarian  et  al.,  1995),  different  music  delivery  methods  (Weber  et  al.,  2010),  and  paper-­‐based  telephone  directories  and  online  equivalents  (Zurkirch  &  Reichart,  2002).  But  one  of  the  most  frequently  considered  transitions  is  that  from  paper  to  digital  media.  Products  such  as  invoices,  telephone  directories,  textbooks,  office  paper,  magazines,  and  newspapers  all  have  digital  alternatives.    The  environmental  impact  of  these  trade-­‐offs  has  most  commonly  been  measured  by  means  of  a  life  cycle  assessment  (LCA),  a  rigorously  defined  and  transparent  methodology  for  quantifying  environmental  burdens  associated  with  the  creation,  use,  and  disposal  of  products  and  systems.    LCA  is  rooted  in  efforts  to  compare  products,  with  a  seminal  study  conducted  in  1969  that  examined  the  differences  between  various  beverage  containers  (LeVan,  1995).  The  tool  was  extended  to  other  comparisons  contrasting,  for  example,  paper  and  plastic  bags,  cloth  and  disposable  diapers,  or  steel  beams  and  dimensional  lumber.  All  of  these  comparisons  are  trade-­‐offs  between  two  product  systems  that  can  be  defined  in  a  straightforward  way.  The  trade-­‐off  between  paper  and  digital  media,  however,  is  more  complex.  Digital  products  can  replace  paper  consumption,  but  this  is  not  their  only  function.               78Researchers  have  also  found  that  increased  digital  media  consumption  does  not  necessarily  reduce  paper  media  consumption  (Sellen  and  Harper,  2001).    These  compounding  factors  suggest  that  the  ICT  sector  may  strain  LCA’s  ability  to  make  meaningful  comparisons.      LCA  excels  in  considering  trade-­‐offs  between  product  systems  that  can  be  clearly  defined  and  in  comparing  products  that  are  discrete  substitutes.  A  paper  or  plastic  bag  is  a  straightforward  consumptive  choice.  The  same  cannot  be  said  for  paper  or  digital  media.  Studies  suggest  that  the  more  discrete  and  precise  the  trade-­‐off  considered,  the  more  effective  the  LCA  is  as  a  tool  of  environmental  assessment  (Gaudreault  et  al.,  2007a,  2007b).  Earlier  research  has  also  shown  that  the  LCA  is  constrained  in  its  ability  to  compare  at  all,  with  Finnveden  (2000,  p.299)  suggesting  that,  “it  can  in  general  not  be  shown  that  one  product  is  environmentally  preferable  to  another,  even  if  this  happens  to  be  the  case.”  Should  process  x  or  y  be  employed  to  minimize  environmental  footprint?  Would  product  a  or  b  have  a  smaller  environmental  footprint?  These  are  the  questions  that  LCAs  can  and  should  answer.  But  sometimes  the  relationships  between  x  and  y  are  enormously  complex.  Managing  this  complexity  imposes  unavoidable  uncertainty  and  a  resulting  series  of  assumptions.  How  does  the  LCA  perform  in  assessing  a  complex  consumptive  choice?  Further,  how  does  the  LCA  perform  when  complexity  lies  not  just  in  a  lack  of  a  discrete  trade-­‐off,  but  in  the  fundamental  character  of  one  of  the  subjects  being  studied?      4.4 OBJECTIVES  Given  the  complex  trade  offs  between  paper  and  digital  media,  I  hypothesize  that  the  differences  between  the  product  systems  that  produce  different  media  undermine  the  LCA’s  ability  to  accurately  compare  environmental  footprints.  I  also  hypothesize  that  the  characteristics  of  ICT  as  an  industrial  sector  weaken  LCA  as  an  environmental  assessment  methodology.  To  test  this  latter  hypothesis,  I  also  examined  LCAs  that  looked  exclusively  at  ICT  products.      This  chapter  answers  these  questions  by  reviewing  comparative  LCAs  that  have  examined  paper  and  digital  media.  By  doing  so,  I  aim  to  elucidate  the  strengths  and  weaknesses  of  LCA  as  a  comparative  tool.  Further,  I  seek  to  strengthen  my  understanding  of  the  role  that  context  plays  in  LCAs,  and  in  this  particular  case,  the  context  of  the  ICT  sector.  I  focus  on  ICT  throughout  the  chapter  because  I  believe  this  sector  warrants  particular  scrutiny.  It  is  disruptive  to  many  aspects  of  the  global  economy  and  several  industrial  sectors  beyond  paper-­‐based  media,  with  direct,  indirect  and  behavioural  effects.  It  is               79changing  rapidly,  with  new  products,  processes  and  devices  emerging.  Forestry  and  paper  production  are  not  without  environmental  impacts,  with  land-­‐use  change,  emissions  from  production,  and  the  creation  of  waste  as  prominent  examples.  (For  a  more  detailed  review  of  issues  associated  with  paper  LCAs  see  Gaudreault  et  al.,  2007a,  2007b).    4.5 METHODOLOGY  I  describe  an  analytical  framework  for  assessing  problems  in  the  conduct  of  an  LCA,  and  the  methods  for  selecting  comparative  LCAs  and  validating  the  results  through  ICT  LCAs.  I  organize  the  results  around  problems  identified  in  an  analytical  framework.  I  describe  the  problem,  examine  how  it  is  addressed  in  comparative  LCAs,  and  then  assess  the  approach  of  ICT  LCAs.    An  LCA  is  a  four-­‐step  tool  designed  to  estimate  the  potential  environmental  impact  of  a  product,  process,  or  system  (ISO,  2006).  These  four  steps,  and  the  challenges  that  occur  at  each  stage,  are  summarized  by  Reap  et  al.  (2008a,  2008b).  The  four  stages  of  conducting  an  LCA  are  goal  and  scope  definition,  inventory  analysis,  impact  assessment,  and  interpretation.  Along  these  four  stages,  the  authors  suggest  six  challenges  that  are  of  particular  concern:  functional  unit  definition,  boundary  selection,  allocation,  spatial  variation,  local  environmental  uniqueness  and  data  availability/quality.  These  six  challenges  structured  the  review  of  comparative  LCAs  that  examine  paper  and  digital  media.  Reap  et  al.  (2008a,  2008b)  identified  these  challenges  as  most  important  because  they  had  significant  influence  over  study  results  and  adequate  solutions  are  available  to  ameliorate  impacts.  I  describe  the  specifics  of  each  challenge  in  the  Results  section,  followed  by  the  analysis  of  comparative  LCAs  and  ICT  LCAs.      I  chose  studies  that  specifically  examined  potential  trade-­‐offs  between  paper  and  digital  media  for  several  reasons.  Paper  and  digital  media  are  very  distinct  product  systems,  and  I  wanted  to  gauge  the  robustness  of  the  LCA  when  comparing  such  different  units  of  analysis.  Further,  the  ICT  sector  is  a  dynamic  and  growing  industry  that  has  disrupted  many  existing  industrial  sectors.  The  idea  of  the  “paperless  office”  was  held  up  as  an  environmentally  preferable  future  (Sellen  and  Harper,  2001).  The  phrase  “please  consider  the  environment  before  printing  this  email”  is  often  appended  to  emails.  It  suggests  that  printing  on  paper  is  bad  for  the  environment,  while  sending  an  email  is  innocuous.  I  wanted  to  understand  whether  this  assumption  that  digital  media  is  preferable  is  supported  by  academic  research.  I  searched  academic  databases  for  peer-­‐reviewed  literature  on  the  subject,  but  also               80considered  publically  available  technical  reports  and  white  papers.  In  the  end,  this  left  us  with  seven  studies  which  I  summarized  using  the  six  key  challenges  in  Reap  et  al.’s  (2008a,  2008b)  analytical  framework.  The  studies  reviewed  are  listed  below.  • Deetman,  S.  &  Odegard,  I.  (2009)  Scanning  Life  Cycle  Assessment  of  Printed  and  E-­‐paper  Documents  based  on  the  iRex  Digital  Reader.  [Online]  Available  from:  http://media.leidenuniv.nl/legacy/irex-­‐dr1000-­‐lca-­‐scan-­‐final-­‐report.pdf.  [Accessed:  9th  June  2014]  • Enroth,  M.  (2009)  Environmental  impact  of  printed  and  electronic  teaching  aids,  a  screening  study  focusing  on  fossil  carbon  dioxide  emissions.  Advance  in  Printing  and  Media  Technology.36.  p.  1-­‐9.  • Gard,  D.  L.  &  Keoleian,  G.  A.  (2003)  Digital  versus  print.  Energy  performance  in  the  selection  and  use  of  scholarly  journals.  Journal  of  Industrial  Ecology.  6  (2).  p.  115–132.  • Hischier,  R.,  Wäger,  P.  &  Gauglhofer,  J.  (2005)  Does  WEEE  recycling  make  sense  from  an  environmental  perspective?  The  environmental  impacts  of  the  Swiss  take-­‐back  and  recycling  system  for  waste  electrical  and  electronic  equipment  (WEEE).  Environmental  Impact  Assessment  Review.  25  (5).  p.  525-­‐539.  • Kozak,  G.  L.  &  Keoleian,  G.  A.  (2003)  Printed  Scholarly  Books  and  E-­‐book  Reading  Devices:  A  Comparative  Life  Cycle  Assessment  of  Two  Book  Options.  Electronics  and  the  Environment,  2003.  IEEE  International  Symposium  on.  p.  291-­‐296.    • Moberg,  Å.,  Borggren,  C.  &  Finnveden,  G.  (2011)  Books  from  an  environmental  perspective-­‐Part  2:  e-­‐books  as  an  alternative  to  paper  books.  International  Journal  of  Life  Cycle  Assessment.  16  (3).  p.  238-­‐246.  • Moberg,  Å.,  Johansson,  M.,  Finnveden,  G.  &  Jonsson,  A.  (2010)  Printed  and  tablet  e-­‐paper  newspaper  from  an  environmental  perspective  —  A  screening  life  cycle  assessment.  Environmental  Impact    Assessment  Review.  30  (3).  p.  177-­‐191.    • Toffel,  M.  W.  &  Horvath,  A.  (2004)  Environmental  implications  of  wireless  technologies.  News  delivery  and  business  meetings.  Environmental  Science  and  Technology.  38  (11).  p.  2961–2970.  Reap  et  al.’s  (2008a,  2008b)  framework  provided  the  concepts  and  theory  necessary  to  identify  whether  inaccuracies  in  LCAs  result  from  a  failure  to  implement  the  LCA  methodology  appropriately.  But  the  framework  cannot  identify  whether  there  are  problems  outside  the  scope  of  LCA  methodology  that  influence  results.  To  identify  the  potential  of  context  to  influence  comparative  LCA  results,  I  triangulated  the  results  by  also  using  ICT  LCAs  that  did  not  attempt  to  compare  one  product  to  another.  The  underlying  logic  is  that  if  the  same  problems  are  identified  in  both  comparative  LCA  s  and  ICT  LCAs,  I  can  better  gauge  the  role  of  context  in  influencing  LCA  results.      The  key  findings  of  each  LCA  are  presented  in  Table  5.  In  every  comparative  LCA,  it  is  found  that  digital  media  is  preferable  to  paper  media.  However,  in  many  of  the  studies  the  authors  suggest  that  there  is  insufficient  data  to  express  much  confidence  in  the  study  findings.  Each  study  expresses  its  results  in  different  units,  from  energy  used  to  global  warming  potential  to  impact  indicators  specific  to  the  LCA               81model  being  employed.  Nonetheless,  a  consistent  trend  of  finding  digital  media  preferable  while  also  noting  the  extent  of  uncertainty  and  assumption  in  the  digital  lifecycle  is  present.      Table  5:  Key  findings  of  comparative  LCAs  Study   Functional  Unit  Deetman  &  Odegard   The  authors  concluded  that  the  iRex  reader  performed  better  than  printers  and  paper  in  supplying  information  for  the  office  worker.  They  indicated  that  using  LWC  paper,  after  3000  single-­‐sided  prints  the  e-­‐reader  was  preferable.  Using  woodfree  uncoated  paper,  5000  prints  were  required  before  the  e-­‐reader  was  a  preferable  option    Enroth   The  study  found  that  the  impact  on  global  warming  of  a  web-­‐based  electronic  teaching  aid  was  approximately  10  times  higher  than  the  environmental  impact  of  a  printed  textbook.    Gard  &  Keoleian   The  study  found  that  digital  media  consumption  varied  between  4.10  and  216  MJ  per  functional  unit.  Paper  media  consumption  varied  between  0.55  and  525  MJ  per  functional  unit.      Hischier  &  Reichart   The  study  found  that  reading  the  news  on  the  Internet  causes  more  impact  than  viewing  it  on  a  TV,  but  reading  on  the  Internet  causes  less  of  an  impact  than  a  physical  newspaper.    They  concluded  that  comparing  multifunctional  products  using  strict  functional  units  did  not  reflect  environmental  impacts.    Kozak   The  e-­‐reader  was  found  to  have  environmental  impact  preferable  to  paper  media  in  most  scenarios,  with  the  only  exception  being  when  a  paper  text  book  is  re-­‐used  at  least  4  times.    Moberg  et  al.  (2010)   The  printed  newspaper  in  general  had  a  higher  energy  use,  higher  emissions  of  gases  contributing  to  climate  change  and  several  other  impact  categories  than  the  tablet  e-­‐paper  newspaper.  It  was  concluded  that  tablet  e-­‐paper  has  the  potential  to  decrease  the  environmental  impact  of  newspaper  consumption    Moberg  et  al.  (2011)   The  study  found  that  when  the  e-­‐book  was  compared  with  a  paper  book,  the  number  of  books  read  on  the  e-­‐book  reader  during  its  lifetime  was  crucial  in  evaluating  environmental  performance  compared  with  paper  books.  The  results  indicated  that  there  are  impact  categories  and  circumstances  where  paper  books  are  preferable  to  e-­‐books  from  an  environmental  perspective  and  vice  versa.    Toffel  &  Horvath   The  authors  compared  reading  a  newspaper  to  reading  the  same  information  on  a  PDA.  They  concluded  that  reading  the  news  on  the  PDA  results  in  the  release  of  23-­‐140  times  less  C02,  several  orders  of  magnitude  less  NOX  and  SOX,  and  the  use  of  26-­‐67  times  less  water.                         82Below  is  a  list  of  eight  non-­‐comparative  LCAs  from  the  ICT  sector  that  I  also  reviewed  for  this  chapter:  • Choi,  B.,  Shin,  H.  S.,  Lee,  S.  Y.,  Hur,  T.  (2006)  Life  Cycle  Assessment  of  a  Personal  Computer  and  its  Effective  Recycling  Rate.  International  Journal  of  Life  Cycle  Assessment.  11  (2).  p.  122-­‐128.  • Duan,  H.,  Eugster,  M.,  Hischier,  R.,  Streciher-­‐Porte,  M.  &  Li,  J.  (2009)  Life  cycle  assessment  study  of  a  Chinese  desktop  personal  computer.  Science  of  the  Total  Environment.  407  (5).  p.  1755-­‐1764.  • Durucan,  S.,  Korre,  A.  &  Munoz-­‐Melendez,  G.  (2006)  Mining  life  cycle  modelling:  a  cradle-­‐to-­‐gate  approach  to  environmental  management  in  the  minerals  industry.  Journal  of  Cleaner  Production.  14(12-­‐13).  p.  1057-­‐1070.  • Frey,  S.  D.,  Harrison,  D.  J.  &  Billett,  E.  (2006)  Ecological  footprint  analysis  applied  to  mobile  phones.  Journal  of  Industrial  Ecology.  10  (1-­‐2).  p.  199-­‐216.  • Liu,  C.  H.,  Lin,  S.  J.  &  Lewis,  C.  (2010)  Life  cycle  assessment  of  DRAM  in  Taiwan’s  semiconductor  industry.  Journal  of  Cleaner  Production.  18  (5).  p.  419-­‐425.    • Lu,  L.  T.,  Wernick,  I.  K.,  Hsiao,  T.  Y.,  Yu,  Y.  H.,  Yang,  Y.  M.  &  Ma,  H.  W.  Balancing  the  life  cycle  impacts  of  notebook  computers:  Taiwan's  experience.  Resources,  Conservation  and  Recycling.  48  (1).  p.  13-­‐25.  • Plepys,  A.  (2004)  The  environmental  impacts  of  electronics.  Going  beyond  the  walls  of  semiconductor  fabs.  Institute  of  Electrical  and  Electronics  Engineers.  p.  159-­‐165.  • Scharnhorst,  W.  (2006)  Life  Cycle  Assessment  of  Mobile  Telephone  Networks,  with  Focus  on  the  End-­‐of-­‐Life  Phase.  International  Journal  of  Life  Cycle  Assessment.  11  (4).  p.  290-­‐291.  • Williams,  E.  D.,  Ayres,  R.  U.  &  Heller,  M.  (2002)  The  1.7  Kilogram  Microchip:  Energy  and  Material  Use  in  the  Production  of  Semiconductor  Devices.  Environmental  Science  &  Technology.  36  (24).  p.  5504-­‐5510.  4.6 RESULTS  I  detail  the  results  of  the  review  of  comparative  LCAs  and  non-­‐comparative  ICT  LCAs  below.  The  results  are  organized  according  to  the  analytical  framework  developed  by  Reap  et  al.  (2008a,  2008b)  that  identifies  the  six  key  challenges  in  the  conduct  of  the  LCA.  However,  because  of  their  similarity,  spatial  variation  and  environmental  uniqueness  are  discussed  in  tandem  in  section  4.6.4.  I  summarize  the  LCAs  reviewed  in  aggregate,  identifying  and  detailing  indicative  results  when  appropriate.      4.6.1 Functional  Unit  Definition  The  functional  unit  is  a  measure  of  the  performance  of  the  functional  outputs  of  the  product  system  (ISO,  2006).  It  is  designed  to  provide  a  reference  to  which  inputs  and  outputs  are  related,  and  a  basis  for  comparability  of  LCA  results.  Challenges  emerge  when  trying  to  incorporate  products  that  have  multiple  functions.  According  to  Reap  et  al.  (2008a,  p.292),  “not  identifying,  decomposing,  specifying  and/or  prioritizing  these  [multiple]  functions  appropriately  with  respect  to  a  study’s  goal  and  scope  might  yield  a  functional  unit  that  fails  to  reflect  reality  well”.  Non-­‐quantifiable  or  difficult-­‐to-­‐quantify  functions  can  also  be  a  source  of  error.                 83Table  6  summarizes  the  functional  unit  in  the  comparative  LCAs  reviewed,  with  a  mix  of  single-­‐function  and  multi-­‐functional  units.  The  paper  product  was  consistently  considered  a  single-­‐function  product  (although  in  some  instances,  such  and  Enroth  (2010)  and  Kozak  (2003),  the  author  modeled  the  ability  to  reuse  the  product).  Some  ICT  devices  were  considered  single-­‐function,  such  as  the  e-­‐readers  modeled  in  Deetman  and  Odegard  (2009),  Moberg  et  al.  (2010),  and  Kozak  (2003).  Other  ICT  devices,  such  as  the  PC  (personal  computer)  modeled  in  Gard  and  Keoleian  (2002)  and  Hischier  and  Reichart  (2003)  and  the  PDA  (personal  digital  assistant)  modeled  by  Toffel  &  Horvath  (2004)  were  multifunctional.  In  order  to  specify  multi-­‐functionality,  the  studies  relied  on  average  data  for  consumer  behaviour.  For  example,  Hischier  and  Reichart  (2003)  used  survey  data  from  Switzerland  to  quantify  average  Internet-­‐use  time.  Most  of  the  studies  oriented  their  functional  unit  around  a  product  rather  than  a  service.  The  exceptions  were  Hischier  and  Reichart’s  (2003)  second  and  third  scenarios,  which  modeled  the  service  of  being  “well  informed”.  However,  they  only  considered  one  device,  a  home  PC,  as  necessary  to  staying  well-­‐informed  using  ICT  products,  an  assumption  that  may  not  reflect  the  reality  of  digital  media  consumption.    Another  consistent  characteristic  was  the  modeling  of  discrete  functional  units.  The  trade-­‐off  between  one  product  and  another  was  always  absolute,  viewing  the  paper  and  ICT  products  as  substitutes,  rather  than  potential  complements.      Table  6:  Summary  of  functional  units  of  the  comparative  LCAs  reviewed  Study   Functional  Unit  Deetman  &  Odegard   One  year  of  office  paper  use,  using  either  an  e-­‐reader  or  a  laser  printer  with  paper  Enroth   The  use  of  a  teaching  aid  (either  a  textbook  or  a  computer)  over  the  course  of  five  years  by  5,000  students  per  year.  Gard  &  Keoleian   Reading  a  scientific  journal  article  in  a  traditional  or  digital  library  system.    Hischier  &  Reichart   1. Reading  of  a  500-­‐word  article  (either  online  or  in  a  newspaper).  2. Staying  “up-­‐to-­‐date”  by  watching  either  25  minutes  of  TV,  10  minutes  online  or  reading  43%  of  a  newspaper.  3. Daily  media  consumption,  either  110  minutes  of  TV,  74  minutes  on  the  Internet  or  an  entire  newspaper.  Kozak   40  scholarly  textbooks  500  pages  in  length,  either  purchased  in  physical  form  or  downloaded  to  an  e-­‐reader.  Moberg  et  al.  (2010)   The  consumption  of  a  newspaper  during  one  year,  either  in  physical  form  or  on  an  e-­‐reader.  Moberg  et  al.  (2011)   One  specific  book  bought  and  ready  by  one  person,  either  as  a  360  page  hardcover  novel  or  a  1.5  MB  PDF  file.  Toffel  &  Horvath   One  year’s  worth  of  the  New  York  Times,  either  delivered  in  Berkley  California  or  wirelessly  to  a  PDA.                     84The  functional  unit  was  not  a  source  of  much  discussion  or  concern  in  the  ICT  LCAs.  The  authors  did  not  attempt  to  allocate  burdens  between  various  activities,  thus  avoiding  many  of  the  challenges  associated  with  functional  unit  definition.  This  approach  avoids  many  of  the  challenges  discussed  above,  such  as  how  to  appropriately  model  consumer  behaviour.  With  the  exception  of  Hischier  &  Reichart’s  (2004)  attempt  to  model  being  “well  informed”,  the  LCAs  reviewed  picked  a  single  product  (be  it  a  PC,  a  mobile  phone  or  a  component  inside  an  ICT  device)  and  designed  their  LCA  accordingly.        4.6.2 Boundary  Selection  Boundary  selection  determines  which  processes  and  activities  are  to  be  included  in  an  LCA.  The  LCA  scientist  must  balance  the  desire  to  produce  objective,  scientific  and  repeatable  LCAs  with  the  constraints  of  resources  and  time  to  carry  out  an  LCA.  Implicit  in  boundary  selection  is  the  definition  of  cut-­‐off  criteria,  which  involves  deciding  whether  a  particular  flow  of  mass  or  energy  should  be  included  in  the  LCA.    Commenting  on  cut-­‐off  criteria,  Reap  et  al.  (2008a,  p.293)  summarizing  Suh  et  al.  (2004)  suggest:  1. “There  is  no  theoretical  or  empirical  basis  that  guarantees  that  a  small  mass  or  energy  contribution  will  always  result  in  negligible  environmental  impacts.  2. Some  input  flows  bypass  the  product  system  and  do  not  contribute  mass  or  energy  content  to  the  product.  3. Environmental  impacts  by  inputs  from  service  sectors  cannot  be  properly  judged  on  the  basis  of  mass  and  energy.    4. While  the  individual  inputs  and  outputs  cutoff  may  be  insignificant,  their  total  sum  might  change  the  results  considerably.”      Ideally,  an  LCA  would  establish  system  boundaries  after  reviewing  existing  data  and  determining  that  particular  flows  are  not  significant  enough  to  merit  inclusion.  However,  if  the  data  exists  to  make  this  determination,  it  is  not  clear  why  the  LCA  should  not  include  this  data  in  the  first  place.  This  is  what  Reap  et  al.  (2008a)  consider  a  paradox  in  the  conduct  of  an  LCA.      Boundary  selection  can  pose  additional  challenges.  Input-­‐Output  (IO)  analysis,  where  the  burdens  of  a  system  are  estimated  by  measuring  energy  and  mass  inputs  and  outputs  and  then  allocating  these  estimates  to  a  unit  of  product,  has  become  a  popular  method  of  defining  system  boundaries  in  LCAs.  Reap  et  al.  (2008a,  p.294)  referencing  Lenzen  (2000)  found  that  this  method  can  be  problematic  when  studying  the  estimated  fossil  fuel  consumption  associated  with  commodities  extraction.  Using  the  IO  method,  an  error  range  of  9%  to  100%  was  identified  when  comparing  the  results  of  an  IO  method  with               85estimates  derived  using  field  measurements.  It  is  safe  to  assume  that  similar  problems  would  be  found  when  comparing  IO  estimates  with  activities  besides  commodities  extraction.      The  comparative  LCAs  reviewed  were  screening  LCAs,  designed  to  use  readily  available  data  to  identify  the  most  important  stages  and  processes  in  the  lifecycle  of  a  system  or  product.  Few  measured  data  were  used,  with  a  reliance  on  secondary  sources,  proxy  data,  average  data  and  life  cycle  inventory  (LCI)  databases.  This  impacted  the  definition  of  system  boundaries,  as  authors  could  be  broad  and  inclusive  since  little  data  was  actually  being  measured.  The  only  author  to  explicitly  consider  cut-­‐off  criteria  was  Kozak  (2003).  He  chose  to  not  model  the  impact  of  any  material  that  constituted  less  than  1%  of  the  mass  of  the  functional  unit.  All  of  the  LCAs  refrained  from  modeling  “content  creation”  as  this  was  assumed  to  be  a  variable  common  to  both  paper-­‐based  and  digital  media  systems.    In  the  comparative  LCAs,  defining  boundaries  in  the  ICT  sector  was  a  particular  challenge  due  to  limited  data  availability.  Only  Deetman  and  Odegard  (2009)  and  Moberg  et  al.  (2010)  had  access  to  actual  data  on  the  ICT  product  being  considered.  However,  these  data  were  considered  confidential  and  were  not  disclosed.  Kozak  (2003)  and  Toffel  and  Horvath  (2004)  demonstrated  just  how  acute  the  challenge  of  ICT  data  can  be  when  broad  boundaries  are  set.  They  were  forced  to  use  equivalency  ratios  from  LCAs  for  entirely  different  products  to  describe  the  functional  unit  being  used.  They  found  an  existing  LCA,  took  the  impact  to  mass  ratio  for  a  particular  product  and  applied  this  ratio  to  an  entirely  different  product.  The  actual  energy  and  materials  embodied  in  the  ICT  device  was  understood  only  by  proxy.  None  of  the  studies  explicitly  modeled  the  abiotic  resource  depletion  associated  with  ICT  products.  Given  the  presence  of  hundreds  of  different  metals,  rare  earth  minerals  and  chemicals  (Lau  et  al.,  2002)  in  an  ICT  device,  there  was  a  potential  for  “significant  insignificants”  (a  concept  used  by  Reap  et  al.  (2008a,  p.299)  to  describe  the  potential  environmental  impacts  of  small  flows  of  energy  or  mass)  in  all  of  the  comparative  LCAs.      Other  challenges  in  boundary  selection  occurred  in  the  use  stage.  For  a  paper  product,  the  inclusion  of  personal  transportation  to  a  retail  outlet  had  a  potentially  significant,  but  highly  variable,  influence  over  study  findings  (Kozak,  2003;  Moberg  et  al.,  2010).  For  ICT  products,  boundary  selection  was  complicated  by  efforts  to  model  the  Internet  backbone  (cables,  routers,  switches,  and  data  centers)  that  delivers  digital  media.  Some  studies  (Deetman  and  Odegard,  2009)  refrained  from  attributing  any  impacts  to  the  Internet.  Gard  and  Keolian  (2002)  recognized  that  that  digital  media  utilized  the  Internet,  but  file  sizes               86are  tiny  relative  to  overall  volume  handled  and,  consequently,  they  did  not  include  the  Internet  in  their  LCA.  Both  Moberg  et  al.  (2010)  and  Enroth  (2010)  attempted  to  model  the  energy  and  material  requirements  of  the  Internet.  However,  they  expressed  these  figures  with  hesitation,  citing  their  reliance  on  Economic  Input-­‐Output  (EIO)  data  that  are  dated  and  may  not  accurately  reflect  the  potential  impacts  of  Internet  infrastructure.      Modeling  the  downstream  impacts  of  the  products  considered  was  another  challenge.  The  end-­‐of-­‐life  (EOL)  management  of  both  paper  products  and  ICT  was  modeled  in  several  ways.  Some  studies,  like  Toffel  and  Horvath  (2004),  did  not  model  EOL  at  all.  Others,  like  Kozak  (2003),  assumed  that  paper  products  would  not  be  recycled,  but  instead  be  held  in  perpetuity,  while  ICT  products  would  be  disposed  of  in  a  local  landfill.  Several  studies  (Deetman  and  Odegard,  2009;  Hischier  and  Reichart,  2003;  Moberg  et  al.,  2010)  assumed  that  ICT  EOL  would  occur  within  the  regulatory  confines  of  the  European  Waste  Electrical  and  Electronic  Equipment  (WEEE)  Directive  and  the  modeled  EOL  impacts  are  representative  only  of  waste-­‐handling  in  Europe  under  ideal  conditions.  Hischier  and  Reichart  (2003)  assume  that  95%  of  the  ICT  equipment  is  recycled  or  incinerated,  with  the  rest  going  to  a  landfill.  The  others  assume  100%  recovery  of  ICT  equipment  for  either  recycling  or  incineration.        The  ICT  LCAs  reviewed  had  similar  challenges  with  boundary  selection.  The  complexity  of  ICT  required  system  boundaries  to  be  broad  and  inclusive.  As  with  the  comparative  LCAs,  the  raw  material  requirements  of  the  ICT  product  were  often  not  modeled  due  to  a  lack  of  available  data  (Choi  et  al.,  2006;  Frey,  2006).  While  aggregate  volumes  of  a  particular  element,  such  as  copper,  were  available,  the  supply  chain  and  footprint  of  each  particular  element  were  never  actually  modeled.    Choi  et  al.  (2006)  demonstrated  how  an  ICT  LCA  includes  a  complex  network  of  pre-­‐manufacturing  activities.  However,  they  were  only  able  to  model  the  basic  raw  material  requirements  of  these  activities.  Any  impacts  associated  with  processing  these  materials  were  unaccounted  for.  Further,  Choi  et  al.  (2006)  suggested  that  several  components  in  the  pre-­‐manufacturing  stage  were  not  modeled  due  to  a  lack  of  data.  But  they  also  found  that  the  pre-­‐manufacturing  stage  was  the  source  of  the  majority  of  the  environmental  impacts  in  the  manufacture  of  a  desktop  computer.        4.6.3 Allocation  Allocation  is  the  “procedure  of  appropriately  allocating  the  environmental  burdens  of  a  multi-­‐functional  process”  (Reap  et  al.,  2008a,  p.  296.)  The  question  of  how  to  apportion  the  burdens  of  multi-­‐functional               87processes  to  single  products  or  functions  is  considered  one  of  the  classical  methodological  problems  in  LCA  science  (Rusell,  2005).  Examples  of  multifunctional  processes  include  landfills  and  oil  refineries:  facilities  that  are  connected  to  many  kinds  of  products  and  functional  units.  Many  solutions  to  the  allocation  problem  have  been  suggested.  These  include:  sub-­‐dividing  a  process  into  sub-­‐processes  to  create  more  explicit  connections;  allocating  based  off  of  physical  relationships  in  cases  where  subdivision  is  not  possible;  allocating  based  off  of  non-­‐physical  relationships  if  physical  relationships  cannot  be  proven;  and  expanding  the  product  system  under  consideration  in  order  to  avoid  allocation  problems  altogether  (Reap  et  al.,  2008a,  2008b).  Each  solution,  however,  creates  its  own  set  of  problems.  Subdivision  may  not  produce  sub-­‐processes  that  actually  relate  to  the  function  unit.  Expanding  the  product  system  is  difficult  as  it  imposes  additional  data  requirements  on  the  life  cycle  inventory  (LCI).  Non-­‐physical  relationships,  such  as  energy,  mass,  volumes  and  economic  value,  are  often  used  to  assist  in  allocation  procedures.  Reap  et  al.,  (2008a,  p.298)  suggest  that  these  relationships,  “have  generally  been  discredited  for  lack  of  justification  [and]  despite  these  warnings  from  the  LCA  community,  non-­‐causal  relationships  seem  to  be  the  predominant  allocation  method  used  in  LCI  practice.”    The  problem  of  allocation  procedures  was  given  some  attention  in  the  comparative  LCAs  that  I  reviewed.  The  consumption  of  digital  media  requires  many  processes  that  are  multi-­‐functional.  Computers,  with  some  exceptions  like  e-­‐readers,  are  multi-­‐functional.  The  infrastructure  that  delivers  digital  media  is  also  multifunctional.  The  supply  chains  that  manufacture  computers,  from  mines  to  petrochemical  facilities  to  component  factories,  are  all  multi-­‐functional  processes.  Toffel  and  Horvath  (2004)  used  energy  consumption  of  the  communications  equipment  sector  to  determine  the  environmental  impact  of  manufacturing  a  mobile  phone,  an  example  of  using  a  non-­‐physical  relationship.  In  the  other  comparative  LCAs,  allocation  procedures  were  only  discussed  with  regards  to  paper  media,  not  digital.      In  the  ICT  LCAs  reviewed,  only  two  discussed  the  problem  of  allocation.  Choi  et  al.  (2006,  p.125)  suggest  that  a  “skillful  recycling  expert”  was  needed  to  allocate  various  waste  flows  associated  with  a  PC.  Allocation  procedures  were  required  to  model  the  packaging  of  a  PC,  the  construction  of  the  steel  frame  that  houses  a  PC  and  the  assembly  of  a  PC.  All  of  these  allocations  were  made  using  volume  and  weight,  relationships  identified  by  Reap  et  al.  (2009a)  as  problematic.  Durucan  et  al.  (2006)  reviewed  the  conduct  of  LCAs  in  the  mining  industry,  which  supplies  many  of  the  raw  materials  present  in  digital               88products.  They  found  that,  with  regards  to  aluminum  and  steel,  “very  little  or  no  emphasis  has  been  placed  on  the  extraction  of  the  mineral  ore  and  the  consequent  waste  handling  aspects  of  the  industry  in  relation  to  the  allocation  of  environmental  burdens.”  (Durucan  et  al.,  2006  p.1058)    4.6.4 Spatial  Variation  and  Environmental  Uniqueness  Impacts  generated  over  a  product’s  life  cycle  can  have  effects  at  the  local,  regional,  or  global  level.  Site-­‐generic  LCAs  lack  spatial  information  and  assume  globally  homogenous  effects,  which  can  embed  inaccuracies  in  LCA  results.  Environmental  impacts  like  acidification  and  eutrophication,  for  example,  can  vary  by  a  magnitude  of  three  due  to  local  meteorological  variations  (Huijbreghts  et  al.,  2001;  Potting  et  al.,  1998).  Hellweg  (2001)  suggests  that  geological  conditions  and  geographic  location  can  impact  groundwater  contamination  from  landfills  by  four  orders  of  magnitude.  Land  use  is  also  highly  dependent  on  spatial  variation.  Infrastructure  that  supports  a  product’s  life  cycle  can  occupy  land  of  varying  ecological  productivity  and  can  indirectly  change  local  meteorological  and  hydrological  patterns  (through,  for  example,  changing  run-­‐off  patterns  of  precipitation).  As  stated  by  Canals  et  al.  (2006),  “land  use  impacts  are  highly  dependent  on  the  conditions  where  they  occur.”    Reap  et  al.  (2008b,  p.378)  summarizes:  “Each  environment  affected  by  resource  extraction  or  pollution  is,  to  a  greater  or  lesser  extent,  unique.  As  a  result,  each  local  environment  is  uniquely  sensitive  to  the  stresses  placed  upon  it  by  a  particular  product  system’s  life  cycle.”    Issues  of  spatial  variation  and  local  environmental  uniqueness  were  given  little  attention  in  the  comparative  LCAs  reviewed.  It  should  be  remembered  that  these  were  screening  LCAs  that  relied  on  available  data.  Given  the  complexity  of  the  systems  under  consideration,  describing  a  system  with  sufficient  precision  to  account  for  spatial  variance  would  be  resource-­‐intensive,  if  not  impossible.  However,  certain  modules  in  the  comparative  LCAs  reviewed  were  more  geographically  precise.  Kozak  (2003)  and  Gard  and  Keoleian  (2002),  for  example,  modeled  specific  activities  at  the  University  of  Michigan.  They  used  the  power  bundle  for  the  region,  mapped  out  the  network  infrastructure  for  the  campus  and  assumed  personal  transportation  that  reflected  averages  for  local  geography.  Hischier  and  Reichart  (2003)  modeled  several  components  at  the  regional  level,  using  data  for  Swiss  media  consumption,  energy  consumption,  and  recycling  rates.  For  other  variables,  however,  they  relied  on  global  data  sets,  reflecting  the  nature  of  ICT  supply  chains.  These  screening  LCAs,  combined  with  the  complexity  of  ICT  supply  chains  and  the  lack  of  high-­‐quality  data  on  ICT  products,  resulted  in  major  variation  in  spatial  scale  inside  each  individual  comparative  LCA.  While  some  parts  of  the  LCAs  were  able               89to  reflect  local  variation  (electricity  bundles  being  the  most  common  example),  similar  precision  was  not  possible  for  other  variables.  None  of  the  comparative  LCAs  reviewed  discussed  whether  this  mixture  of  local,  regional,  and  global  scales  had  the  potential  to  impact  study  findings.  By  the  same  token,  the  characteristics  of  screening  comparative  LCAs  (estimated  data,  global  scale,  broadly  defined  system  boundaries)  preclude  them  from  including  local  environmental  issues.  And  again,  the  authors  did  not  discuss  the  potential  impacts  on  study  findings.      Issues  of  spatial  variation  and  local  environmental  uniqueness  followed  similar  trends  in  ICT  LCAs  as  those  found  in  comparative  LCAs.  The  complex  global  supply  chains  of  ICT  products  required  authors  to  rely  on  LCI  databases  that  often  consisted  of  national-­‐level  statistics.  The  ICT  industry  relies  on  original  equipment  manufacturers  (OEMs)  that  are  highly  specialized  and  provide  almost  all  of  the  components  inside  a  device;  tracing  and  measuring  the  impacts  of  all  of  these  suppliers  would  be  extremely  resource-­‐intensive.  This  is  not  to  suggest  that  spatial  variation  and  local  environmental  uniqueness  do  not  matter  in  ICT.  Rather,  they  are  just  attributes  that  are  very  difficult  to  measure.      One  particular  area  where  ICT  LCAs  relied  upon  global  averages  was  in  modeling  the  raw  materials  present  in  a  device.  Durucan  et  al.  (2006)  examined  why  assuming  consistent  environmental  impacts  between  commodity  supplies  is  problematic.  They  found  that  most  of  the  data  available  in  mining  LCAs  pay  very  little  attention  to  modeling  the  extraction  of  mineral  ore  or  the  waste  handling  processes  at  a  specific  site.  They  considered  available  LCA  data  on  mining  to  be  a  “largely  simplified  …  single  fact  sheet.”  (p.1058)  Given  these  uncertainties,  and  the  connection  of  the  mining  sector  with  the  pre-­‐manufacturing  stage  of  an  LCA  device,  confidence  in  ICT  LCAs  should  be  adjusted  accordingly.  The  raw  material  figures  used  are  a  reflection  of  averages  that  Durucan  et  al.  (2006)  suggest  do  not  necessarily  reflect  actual  environmental  impacts.      4.6.5 Data  Availability  and  Quality  Problems  revolving  around  data  availability  and  quality  are  numerous.  According  to  Björklund  (2002),  there  are  five  challenges  related  to  data  quality:  badly  measured  data;  data  gaps;  unrepresentative  (proxy)  data;  model  uncertainty;  and  uncertainty  about  methodological  choices.  Further,  both  data  and  models  can  fail  to  accurately  represent  the  temporal  and  spatial  scope  defined  in  an  LCA.  Some  data  is  virtually  unobservable,  in  instances  of  product  recovery  or  end-­‐of-­‐life  management.  Data  can  be  sourced  from  a  standardized  LCA  database,  but  these  data  sets  may  not  be  peer-­‐reviewed  and  finding  the               90original  data  sources  can  be  a  challenge  (Björklund,  2002).  Data  measuring  the  same  variable,  but  from  different  sources  can  arrive  at  conflicting  conclusions.  Reap  et  al.  (2008b,  p.383)  suggests  that,  “in  general,  the  literature  tends  to  agree  that  data  for  life  cycle  inventories  is  not  widely  available  nor  of  high  quality.”  Data  can  be  outdated,  compiled  at  different  times,  and  correspond  to  different  materials  produced  over  different  time  periods.      The  opportunities  for  low-­‐quality  data  to  enter  an  LCA  are,  therefore,  abundant.  And  although  one  instance  of  low-­‐quality  data  might  not  have  a  significant  impact  on  LCA  results,  many  compounded  instances  of  suspect  data  can  potentially  diminish  confidence  in  LCA  findings.  As  Bare  et  al.  (1999,  p.301)  suggest,  it  is  hard  to  know  “where  to  draw  the  line  between  sound  science  and  modeling  assumptions.”    The  comparative  LCAs  reviewed  had  significant  issues  with  data  quality  and  availability.  Two  factors  contribute  to  this:  the  broad  system  boundaries  defined  and  the  particular  challenge  of  modeling  ICT  products.  All  of  the  studies  relied  on  LCI  databases.  Moberg  et  al.  (2010),  Deetman  and  Odegard  (2009)  and  Hischier  and  Reichart  (2003)  used  the  Ecoinvent  LCI  database.  Toffel  and  Horvath  (2004)  relied  on  data  from  other  LCAs,  using  equivalency  ratios  and  the  Carnegie-­‐Mellon  EIO-­‐LCA  database,  a  source  specifically  identified  by  Reap  et  al.  (2008a,  p.294)  as  problematic.  Kozak  (2003)  and  Gard  and  Keoleian  (2003)  relied  on  the  Ecobilan  DEAM  database  for  several  modules  in  their  LCA.  In  two  instances,  Kozak  (2003)  and  Toffel  and  Horvath  (2004)  relied  on  proxy  data.  More  problematic  is  how  proxy  data  were  used  to  model  several  parts  of  a  product,  compounding  whatever  errors  the  use  of  proxy  data  might  introduce.  Given  resource  and  time  constraints,  as  well  as  the  scale  and  diversity  of  the  product  systems  being  compared,  it  is  not  surprising  that  issues  of  data  quality  and  availability  were  prominent.  However,  most  authors  did  not  comment  on  these  issues,  and  those  who  did  (Kozak,  2003;  Toffel  and  Horvath,  2004)  expressed  a  generally  high  degree  of  confidence  in  their  data.      Data  availability  was  addressed  as  a  serious  challenge  in  all  of  the  ICT  LCAs  I  reviewed.  Frey  et  al.  (2006)  and  Scharnhorst  (2006)  cited  a  paucity  of  data  in  trying  to  measure  the  environmental  footprint  of  a  mobile  phone.  Lu  et  al.  (2006),  when  measuring  a  notebook  computer  in  Taiwan,  had  to  rely  on  a  consulting  report  from  1998  (Atlantic,  1998)  that  used  voluntary  survey  data  from  private  companies  that  described  manufacturing  processes  from  the  mid-­‐1990s.  In  modeling  the  footprint  of  a  desktop  PC,  Choi  et  al.  (2006)  relied  on  66  different  databases,  either  private  databases  from  the  Simapro©  software  package  or  Korean  national  databases.  All  of  these  studies  focused  on  a  finished  product,  not  a  component.  Their  reliance  on  LCI  databases  that  are  private,  outdated,  or  populated  by  estimated  data               91is  a  source  of  concern.  However,  the  authors  were  faced  with  few  alternatives  given  the  scope  of  the  ICT  supply  chain.  Studies  that  focused  on  more  specific  components  in  the  ICT  supply  chain  (Liu  et  al.,  2010;  Plepys,  2004;  Williams  et  al.,  2002)  were  able  to  use  more  site-­‐specific  data,  which  had  higher  resolution  and  could  be  directly  attributed  to  the  products  of  the  systems  under  review.    4.7 DISCUSSION  I  reviewed  a  number  of  LCAs  that  compared  paper  and  digital  media  using  an  analytical  framework  developed  by  Reap  et  al.  (2008a,  200b).  The  findings  of  these  LCAs  were  variable  and  relied  on  several  assumptions  to  cope  with  uncertainties.  By  also  reviewing  ICT  LCAs,  I  found  that  these  uncertainties  were  not  necessarily  the  result  of  poorly  implemented  LCAs.  Instead,  the  stark  difference  between  paper  and  digital  products  weakened  the  LCA  as  a  tool  of  comparison.  Further,  using  the  LCA  to  study  digital  products  forced  LCA  practitioners  to  deal  with  the  particular  complexities  of  the  ICT  sector.  These  results  support  my  initial  hypotheses:  that  the  differences  between  the  product  systems  that  produce  paper  and  digital  media  weaken  LCA’s  ability  to  accurately  compare  environmental  footprints  and  that  the  characteristics  of  ICT  as  an  industrial  sector  weaken  LCA  as  an  environmental  assessment  methodology.  Moberg  et  al.  (2011,  p.245),  after  comparing  the  footprint  of  a  hardcover  book,  and  an  e-­‐book  suggested  that  there  is  “the  need  for  more  studies  on  a  macro  level  in  order  to  assess  the  magnitude  of  the  environmental  impacts  of  changing  media  practices.”      I  organize  the  discussion  here  in  three  parts.  I  examine  the  unavoidable  uncertainties  in  comparative  LCAs,  the  assumptions  that  result,  and  the  role  the  ICT  sector  plays  in  weakening  the  conduct  of  LCAs.    The  structure  of  this  discussion  reflects  a  concept  present  in  all  LCA  work  –uncertainties  exist  and  assumptions  are  required.      4.7.1 Uncertainties  in  Comparative  LCAs    The  main  source  of  uncertainty  in  all  of  the  LCAs  reviewed  surrounded  the  data  for  products  and  processes  in  the  ICT  sector.  Data  on  raw  material  inputs  requirements  for  ICT  devices  was  based  on  estimates  or  relied  on  aggregated  sources.  The  connections  between  a  manufactured  product,  such  as  a  semiconductor,  and  the  global  mining  sector  were  not  explicitly  discussed.  Authors  relied  on  assumptions  and  allocations  made  by  LCI  databases,  which  may  not  necessarily  capture  the  complexity  of  the  ICT  sector.  ICT  devices  have  hundreds  of  metals  and  minerals,  and  their  manufacture  requires  extremely  pure  inputs.  These  databases  sometimes  relied  on  economic  input-­‐output  (EIO)  methods,               92where  industry-­‐wide  inputs  and  outputs  are  used  to  determine  environmental  impacts.  Where  primary  data  was  available,  it  was  often  the  average  footprint  for  an  element  (e.g.  silicon)  that  may  not  reflect  the  grade  of  element  used  in  the  ICT  manufacturing  process.        Similar  uncertainty  surrounded  efforts  to  measure  the  Internet  backbone.  Only  Moberg  at  al.  (2010)  and  Enroth  (2010),  who  simply  borrowed  Moberg  et  al.’s  (2010)  findings,  attempted  to  model  the  footprint  of  the  Internet.  They  both  used  a  figure  that  was  derived  using  EIO  methods  from  the  USA  SimaPro  Input  Output  database.  These  figures  are  from  1998  and,  as  Moberg  et  al.  (2009,  p.42)  stated,  “the  results  concerning  potential  impact  of  using  telecommunication  infrastructure  are  too  uncertain  to  draw  any  real  conclusions  from.”  The  Internet  has  evolved  at  a  breakneck  pace,  with  infrastructure  like  routers,  switches  and  data  centers  constantly  being  updated.  This  infrastructure  has  increased  in  energy  efficiency,  but  may  be  characterized  by  the  rebound  effect,  where  a  particular  activity  becomes  more  efficient,  but  the  scale  of  use  increases  at  such  a  rate  that  the  overall  impact  increases  (Williams  et  al.,  2002).  In  the  comparative  LCAs  reviewed,  the  exclusion  of  the  Internet  in  some  studies  was  justified  because  of  the  sheer  volume  of  data  contrasted  with  the  miniscule  size  of  the  functional  unit  under  consideration.  This  is  a  legitimate  claim  from  the  technical  standpoint  of  conducting  an  LCA.  However,  it  does  pose  a  risk,  as  cut-­‐off  criteria  might  eliminate  a  key  variable  –  the  Internet  –  that  could  alter  study  findings.  If  the  impacts  of  Internet  infrastructure  are  not  included  in  LCAs  of  digital  media  consumption,  it  is  worth  asking  where  exactly  they  should  be  considered.      Another  source  of  uncertainty  associated  with  ICT  data  comes  from  end-­‐of-­‐life  (EOL)  management.  E-­‐waste,  which  is  waste  made  up  of  discarded,  obsolete,  or  broken  electronic  devices,  is  a  significant  global  issue  (Grossman,  2006),  but  the  seriousness  of  this  issue  is  not  evident  in  the  decisions  made  by  the  comparative  LCA  authors.  Robinson  (2009)  found  that  globally,  20  to  25  million  tonnes  of  e-­‐waste  are  generated  every  year.  Most  of  this  comes  from  Europe,  the  United  States,  and  Australasia,  but  developing  countries  will  become  major  contributors  as  their  economies  grow.  The  changing  nature  of  e-­‐waste  could  potentially  mitigate  this  problem,  as  device  miniaturisation  shrinks  the  overall  flow  of  waste  (Robinson  2009,  p.186).  But  globally  the  majority  of  E-­‐waste  is  still  disposed  of  in  a  landfill,  rather  than  properly  processed  in  a  modern  facility.        Many  of  the  comparative  LCAs  were  conducted  in  Europe,  which  has  a  strong  regulatory  framework  for  managing  e-­‐waste.  The  WEEE  Directive  (European  Commission,  2012,  Hischier  et  al.,  2005.)  forces               93suppliers  to  recover  and  properly  process  e-­‐waste,  and  compliance  is  high.  In  contrast,  only  18%  of  e-­‐waste  in  the  United  States  is  processed  domestically  (EPA,  2008).  Kozak  (2003),  however,  assumed  that  all  e-­‐waste  is  disposed  of  in  a  local  landfill.  Toffel  &  Horvath  (2004)  did  not  attempt  to  model  EOL.  Another  reason  why  EOL  was  given  little  attention  is  that  most  of  the  studies  (Hischier  and  Reichart  (2003),  Kozak  (2003),  and  Moberg  et  al.  (2010)  being  the  exceptions)  only  focus  on  issues  of  energy  use  and  global  warming  potential  (GWP).  Duan  et  al.  (2008)  when  conducting  sensitivity  analyses  around  EOL  impacts  found  that  modern  technical  systems  are  assumed,  the  environmental  impact  of  this  lifecycle  stage  are  reduced.  E-­‐waste  can  be  a  highly  toxic  and  localized  environmental  issue,  with  a  small  energy  footprint  but  a  large  environmental  footprint.  The  LCA,  as  a  tool  of  environmental  assessment,  falls  short  when  attempting  to  model  highly  localized  environmental  issues,  an  issue  discussed  extensively  by  Reap  et  al.  (2008a,  2008b).  This  focus  on  GWP  is  unsurprising,  given  the  prevalence  of  climate  change  as  an  environmental  concern.  However,  these  studies  should  not  inspire  confidence  in  understanding  the  range  of  potential  environmental  impacts  associated  with  media  consumption.  On  issues  such  as  toxicity,  abiotic  resource  depletion,  and  land  use  impacts,  to  name  a  few,  there  remains  a  high  degree  of  uncertainty.      There  are  also  uncertainties  about  the  social  impacts  of  paper  and  digital  media  consumption.  The  effects  on  human  health,  for  example,  from  manufacturing  ICT  products  have  been  found  to  be  toxic  and  at  times  carcinogenic.  Processing  e-­‐waste  is  another  social  hazard  associated  with  ICT,  as  e-­‐waste  is  often  exported  and  processed  by  underage  workers  in  jurisdictions  with  weak  environmental  controls.  I  mention  this  only  in  passing,  as  the  focus  of  the  review  is  the  environmental,  not  social,  implications  of  media  consumption.  However,  investigating  the  social  impacts  of  ICT  as  a  macro  level  is  an  important  subject  and  warrants  further  research.    4.7.2 Assumptions  in  Comparative  LCAs  The  studies  reviewed  here  should  not  be  faulted  for  having  to  estimate  several  variables  in  the  life  cycle  of  ICT  products.  However,  there  is  a  danger  that  the  LCAs  are  driven  more  by  assumptions  made  than  by  actual  data  collected,  and  that  study  results  are  presented  in  such  a  way  as  to  minimize  the  potential  uncertainties  in  the  analyses.      As  has  been  mentioned,  almost  all  of  the  studies  (Hischier  and  Reichart  (2003)  being  the  exception)  assumed  a  functional  unit  that  is  a  product,  not  a  service.  It  should  be  asked  whether  these  discrete               94product-­‐based  functional  unites  are  reflective  of  consumer  behaviour.  Deetman  and  Odegard  (2009)  modeled  an  office  worker  switching  from  printing  10,000  pages  a  year  to  using  nothing  but  an  e-­‐reader.  Kozak  (2003)  modeled  40  textbooks,  either  in  print  or  on  an  e-­‐reader.  Moberg  et  al.  (2010)  did  the  same  for  newspapers,  as  did  Toffel  and  Horvath  (2004).  In  short,  there  was  a  pattern  of  perfect  substitution.  Paper  and  ICT  products  were  seen  as  alternatives,  not  complements.  There  is  a  recognizable  advantage  for  the  modeler  in  choosing  such  a  trade-­‐off,  but  it  is  unclear  if  this  is  reflective  of  actual  consumer  behaviour.      Hischier  and  Reichart  (2004)  bucked  this  trend  and  attempted  to  model  the  environmental  impacts  of  being  “well  informed”.  They  made  assumptions  that  are  worth  looking  at  closely.  For  paper  products,  they  assumed  that  only  one  daily  newspaper  is  read.  An  alternate  and  perhaps  more  accurate  scenario  might  look  at  the  overall  print  media  consumption,  modeling  a  bundle  of  newspapers,  books,  and  magazines  in  order  to  measure  environmental  impact.  The  same  bundling  approach  would  also  provide  a  richer  analysis  of  the  impacts  of  ICT  products.  To  stay  well  informed,  many  consumers  use  a  computer  at  home,  another  computer  at  work,  a  mobile  device,  and  a  tablet  computer.  Measuring  the  aggregate  impact  of  these  ICT  products  would  provide  a  better  understanding  of  staying  informed  with  ICT  products.  Defining  how  consumers  are  digesting  digital  media  would  be  an  important  but  challenging  baseline  to  establish,  but  there  are  data  available  that  might  help  serve  that  purpose  (see,  for  example,  Bohn  and  Short  (2009)).    Comparing  the  findings  of  Enroth  (2010)  and  Kozak  (2003)  further  demonstrates  the  influence  that  assumptions  exert  on  LCA  findings.  Both  modeled  the  use  of  a  textbook,  although  Enroth’s  (2010)  textbooks  are  for  school  children,  while  Kozak  (2003)  considered  a  university  environment.  Enroth  (2010)  found  that  the  global  warming  potential  (GWP)  of  an  electronic  textbook  is  10  to  30  the  GWP  of  a  printed  textbook.  Kozak  (2003)  did  not  express  the  difference  in  GWP,  but  he  did  state  that  the  lifetime  energy  use  of  an  e-­‐reader  system  is  more  efficient  by  a  factor  of  approximately  five  (742  MJ  (megajoules)  for  the  e-­‐reader,  3,794  MJ  for  the  textbooks).  Explaining  the  difference  between  the  two  shows  how  far  assumptions  can  go  in  driving  study  results.  Enroth  (2010)  assumed  that  five  different  students  use  a  textbook  over  a  period  of  five  years.  The  study  did  not  include  the  energy  necessary  to  transport  students  to  and  from  school  in  the  footprint  of  the  textbook.  Kozak  (2003),  in  contrast,  assumed  that  a  student  purchases  a  textbook  and  retains  it  for  life.  Further,  the  study  modeled  the  energy  necessary  to  transport  the  student  to  a  bookstore  to  purchase  the  textbook.  This  personal               95transportation  constituted  almost  a  third  of  the  energy  used  in  the  lifetime  of  a  textbook.  When  Kozak  (2003)  assumed  the  textbook  could  be  resold,  and  did  not  include  the  energy  of  a  trip  to  the  bookstore  in  the  footprint  of  the  textbook,  it  only  took  three  different  students  using  the  same  textbook  for  the  energy  footprint  of  the  printed  product  to  match  that  of  the  ICT  product.    4.7.3 The  Drivers  of  Uncertainty  and  Assumptions:  The  Context  of  Moore’s  Law  The  assumptions  in  the  comparative  LCAs  reviewed  were  the  result  of  stark  differences  between  the  products  being  compared  –  paper  and  digital  media.  The  uncertainties,  however,  originated  in  the  ICT  sector  and  its  unique  complexity.  The  ICT  sector  is  defined  by  an  inexorable  pace  of  innovation.  This  process  was  first  expressed  by  Ronald  D.  Moore,  founder  of  Intel©,  who  suggested  that  transistor  density  on  a  circuit  board  would  double  every  24  months  and  that  the  cost  of  a  producing  a  transistor  would  halve  over  the  same  period.  For  over  forty  years,  Moore’s  prediction  has  held  true,  and  is  now  referred  to  as  Moore’s  Law.  This  reliable  rate  of  advancement  has  given  investors  the  confidence  needed  to  support  a  massive  research  and  development  budget  in  ICT  (Jorgenson  and  Wessner,  2004).  The  outputs  of  this  steady  cycle  innovation  disrupt  existing  industries  and  the  resulting  profits  justify  continued  investment.      Moore’s  Law  has  implications.  The  rate  of  innovation  ensures  that  technologies  and  their  associated  manufacturing  infrastructure  quickly  become  obsolete.  The  nature  of  ICT  innovation  –  higher  density  circuits  that  are  always  shrinking  in  size  –  creates  specific  environmental  challenges.  Plepys  (2004)  studied  the  environmental  impacts  of  semiconductor  manufacturing.  Semiconductors  are  the  backbone  of  ICT  and  their  environmental  impact  is  of  central  importance  to  any  assessment  of  ICT  products.  Semiconductor  manufacturing  has  rapidly  evolved,  and  the  increasing  sophistication  of  ICT  products  relies  on  component  miniaturization  and  increasingly  complex  integrated  circuits.  Put  simply,  the  ability  for  a  smartphone  today  to  out-­‐compute  a  desktop  computer  just  ten  years  old  is  due  to  advances  in  semiconductor  manufacturing  (Rupp  and  Selberherr,  2011).  These  advances  have  translated  into  sophisticated  manufacturing  processes.  At  the  time  of  writing,  Plepys  (2004)  found  that  over  200  process  steps  are  required  in  the  production  of  a  semiconductor.  Today,  that  number  is  likely  higher.  These  steps  involve  metal  disposition,  rinsing,  and  etching  the  semiconductor  with  a  variety  of  chemicals.  Purity  of  the  chemicals  employed  is  of  central  importance.  Impurities  can  be  tolerated  only  at  a  level  of  parts-­‐per-­‐billion  (Pelpys,  2004).  Plepys  (2004,  p.159)  found  that  “increasing  material  purity  requirements  may  contribute  to  shifting  the  centre  of  manufacturing-­‐related  environmental  impacts               96from  circuit  fabrication  to  raw  material  production  stages.”  He  found  that  existing  LCAs  did  not  actually  model  the  environmental  impacts  of  these  chemicals,  but  instead  used  LCI  data  for  bulk-­‐chemicals  several  grades  lower  than  the  purities  required  for  semiconductor  manufacturing.  He  concluded,  “for  this  reason  a  potentially  large  part  of  upstream  energy  consumption  remains  unaccounted.”  (p.160)      There  are  hundreds  of  components  inside  any  device,  with  hundreds  and  perhaps  thousands  of  contractors  and  sub-­‐contractors  responsible  for  different  aspects  of  an  ICT  product  (Adexa,  2000).  This  creates  a  logistical  challenge  for  simply  manufacturing  an  ICT  product.  It  presents  perhaps  an  insurmountable  challenge  for  producing  a  rigorous  LCA  of  an  ICT  product.    It  also  implies  that  part  of  the  footprint  of  an  ICT  product  is  spread  out  over  many  private  entities,  over  which  a  final  manufacturer  has  no  significant  control  or  oversight.  Recent  struggles  by  Apple©  (Fair  Labor  Association,  2012)  in  managing  the  environmental  and  human  hazards  at  their  suppliers  facilities’  in  China  highlight  this  concern.  If  a  company  with  the  resources  of  Apple©  has  difficulties  managing  their  supply  chain,  it  is  reasonable  to  assume  other  companies  experience  similar  challenges.  The  pace  of  innovation  and  the  potential  profits  associated  with  a  new  product  also  incentivizes  a  degree  of  secrecy  in  ICT  companies  (Williams  et  al.,  2002).      Moore’s  Law  has  driven  down  the  price  of  computers  to  the  extent  that  two  billion  people  currently  access  the  Internet  (Miniwatts,  2012).  New  ways  of  accessing  the  Internet  and  using  ICT  devices  continue  to  emerge,  as  tablet  computers  and  smart  phones  complement  existing  desktop  and  laptop  computers.  Part  of  every  digital  footprint  exists  in  the  “cloud”,  a  series  of  data  centers  worldwide  that  offer  computing  services  to  consumers  and  corporations.  The  energy  profiles  of  devices  themselves  are  changing.  Small  devices  use  less  power,  shifting  the  impacts  of  their  footprint  from  their  operation  to  their  construction  and  running  of  the  cloud  to  which  they  connect  (Singhal,  2005).  The  constant  stream  of  new  devices  means  old  devices  are  always  being  replaced,  creating  a  large  flow  of  e-­‐waste.  Some  suggest  that  3%  of  the  global  pool  of  electronic  components  become  obsolete  each  month  (Sandborn,  2008).  This  pattern  of  obsolescence  requires  complex  and  often  wasteful  logistical  planning  for  managers  of  electronic  systems  that  have  long  lifespans.  Because  electronic  components  are  only  produced  for  a  relatively  short  period  of  time,  systems  with  lifespans  greater  than  three  years  must  buy  excess  components  to  hedge  against  future  failures  in  10  or  15  years.    While  some  of  the  challenges  posed  by  Moore’s  law  are  only  indirectly  related  to  the  consumption  of  digital  media,  taken  as  a  whole,  these  challenges  explain  why  the  LCA  methodology  fails  to  adequately  model  the  ICT  sector.               97Moore’s  Law  and  the  rapid  pace  of  innovation,  adoption,  obsolescence  and  consumption  in  the  ICT  sector  presents  a  challenge  when  producing  meta-­‐analyses  of  LCAs.  The  time  range  of  the  LCAs  reviewed  spans  from  2003  to  2010.  Over  that  time,  the  ICT  sector  evolved.  LCAs  produced  in  2003  were  written  when  smartphones  like  the  iPhone  had  not  even  been  conceived  of  (the  first  iPhone  was  released  in  2007).  Tablet  computers  did  exist,  but  they  were  completely  unlike  modern  products  like  the  Apple  iPad,  which  was  released  in  2010  or  the  Amazon  Kindle,  released  in  2007.  When  looking  at  comparative  LCAs  that  span  over  a  decade,  the  quality  of  the  data  and  the  nature  of  the  products  being  described  have  changed.  Over  ten  years,  ICT  products  have  evolved.  Entire  new  form  factors  –  like  the  tablet  computer  and  smart  phone  –  have  emerged,  while  others  –  like  the  CRT  (cathode  ray  tube)  monitor  and  desktop  monitor  –  have  lost  previous  dominance.  At  the  same  time,  the  LCA  methodology  has  evolved.  The  LCA  achieved  ISO  standardization  in  2006,  and  several  methodological  problems  are  still  being  addressed  (Finnveden  et  al.,  2009).  This  reality  makes  it  difficult  to  compare  LCAs  from  different  time  periods.  Unfortunately,  such  comparisons  are  unavoidable  given  the  limited  supply  of  LCAs  that  compare  paper  and  digital  media  consumption.    While  the  ICT  sector  evolved,  the  background  data  that  fed  into  the  LCAs  I  reviewed  evolved  as  well.  The  earliest  studies,  Hischier  and  Reichart  (2003),  Kozak  (2003),  and  Toffel  and  Horvath  (2004)  used  multiple  data  sources  in  order  to  construct  their  LCI  databases,  as  the  existing  commercially  available  LCI  databases  could  not  support  the  desired  LCA.  As  time  progressed,  standardized  LCI  databases  improved.  Products  like  Ecoinvent  and  SimaPro  (both  LCI  databases)  have  increased  in  sophistication,  adding  data  on  specific  electronic  components  that  was  not  available  in  early  iterations.  This  suggests  that  over  time  the  underlying  LCI  databases  improved,  but  whether  these  improvements  have  kept  pace  with  the  innovation  cycle  of  ICT  is  not  clear.      4.8 CONCLUSIONS  To  model,  conduct,  and  assess  an  LCA  in  an  objective  and  repeatable  way  is  a  challenge  in  and  of  itself.  To  produce  an  LCA  using  a  functional  unit  that  divides  a  product  between  multiple  uses  introduces  a  host  of  challenges  when  allocating  environmental  burdens.  An  LCA  that  attempts  to  compare  two  completely  different  products  introduces  a  whole  set  of  other  uncertainties  and  assumptions.  To  produce  an  LCA  for  a  product  in  the  ICT  sector  means  grappling  with  one  of  the  most  complex  and  globalized  industrial  sectors.  A  comparative  LCA  of  paper  and  digital  media  is  a  culmination  of  these  challenges.                 98I  have  seen  that,  in  the  cases  reviewed,  that  the  LCA  has  drawbacks  in  modeling  ICT.  At  several  stages  in  the  supply  chain,  but  most  acutely  in  the  raw  material  acquisition  stage,  pre-­‐manufacturing,  and  Internet  backbone  stages,  the  LCAs  reviewed  used  coarse,  estimated,  and  outdated  data  to  measure  ICT  products.  Research  has  suggested  (Plepys,  2004)  that  significant  environmental  impacts  are  simply  unaccounted  for.  It  is  also  likely  that  ICT  data  poorly  represents  the  raw  material  requirements  of  a  product,  especially  in  light  of  the  number  of  raw  materials  present  and  the  potential  variation  in  raw  material  supply  chains.  Given  the  potential  for  site-­‐level  variation  in  the  extractive  industry  sector,  this  unexplored  relationship  of  ICT’s  connection  with  nonrenewable  raw  materials  is  problematic.  At  some  point  or  another,  almost  all  of  the  studies  I  reviewed  relied  on  proxy  data,  estimating  the  impact  of  a  component  or  product  by  manipulating  data  available  for  something  similar.  What  is  unclear  and,  in  my  view,  a  pressing  concern  is  the  compounding  effects  of  all  of  the  estimates,  guesses,  cutoff  criteria,  and  low-­‐resolution  data  that  characterize  lifecycle  explorations  of  ICT.      It  is  clear  that  the  scope  for  variation  between  LCA  findings,  and  inside  an  individual  LCA  is  large.  What  is  less  clear  is  where  this  variation  is  borne:  in  a  lack  of  precise  data;  in  the  nature  of  ICT  supply  chains;  in  the  challenge  of  modeling  consumer  behaviour;  in  modeling  a  multifunctional  product;  or  in  attempts  to  compare.  All  of  these  variables  should  be  considered  closely  when  reading  the  existing  literature  on  paper  and  digital  media.  These  comparisons  rely  on  averages  out  of  necessity,  not  neglect.  Indeed,  it  is  not  whether  study  findings  are  a  reflection  of  available  data,  but  whether  available  data  are  of  sufficient  quality  to  inspire  confidence  in  the  study  findings.  But  given  the  scope  for  variation  inside  comparative  LCAs,  I  am  forced  to  ask,  are  there  alternative  means  of  understanding  the  footprint  of  media  consumption  besides  the  LCA?  And  should  the  LCA  alone  inform  what  I  consider  to  be  “better”?      Looking  forward,  an  important  trend  is  that  ICT  devices  are  getting  smaller  and  more-­‐energy  efficient.  This  means  that  the  environmental  profile  of  an  ICT  device  is  shifting.  More  and  more  of  the  impacts  are  found  upstream  in  raw  material  extraction  and  production,  or  downstream  in  the  EOL  stage,  precisely  where  LCAs  are  weakest.  And  during  the  use  stage,  more  impacts  will  come  from  accessing  the  Internet.  The  footprint  of  cloud  computing,  and  the  Internet  backbone  more  generally,  is  poorly  understood.  In  short,  the  environmental  burden  of  digital  media  is  shifting  towards  areas  that  are  least  understood.  Gordon  Moore,  interviewed  in  2005  was  unintentionally  prophetic  in  a  discussion  about  his  eponymous  law,  saying,  “It  can't  continue  forever.  The  nature  of  exponentials  is  that  you  push  them  out  and  eventually  disaster  happens.”  (Dubash,  2005,  n.p.)  While  he  was  discussing  the  technical  feasibility  of               99the  perpetual  doubling  of  transistor  density,  his  comments  might  also  apply  to  the  similarly  relentless  pace  of  new  gadgets  and  methods  for  consuming  media.                   100Chapter  5: Discussion  and  Conclusions    5.1 CONCLUSIONS    It  has  been  said  a  few  times  already,  but  it  bears  repeating:  the  consumption  of  the  written  word  is  changing.  New  technologies,  business  models,  and  evolving  consumer  preferences  are  driving  a  shift  from  paper  to  digital  media.  But  this  shift  is  uneven  and  complex.  By  no  means  are  all  forms  of  paper  being  replaced  –  hygienic  uses  are  likely  to  endure  for  a  long  time  indeed  –  but  for  some  product  categories,  like  newspapers,  books,  and  magazines,  digital  media  has  represented  an  existential  threat  to  paper  alternatives.  I  call  this  “the  media  shift”.      Meanwhile,  society  is  finally  waking  up  to  the  environmental  challenges  that  we  face.  Concepts  like  climate  change,  ecological  carrying  capacity,  and  environmental  footprint  have  gained  traction.  The  impact  of  human  industry  and  consumption  on  the  planet  is  being  felt.  Some  have  suggested  that  we  are  entering  the  age  of  the  Anthropocene,  where  the  most  influential  force  on  the  planet  is  homo  sapiens  (Zalasiewicz  et  al.,  2010).  In  response,  the  concept  of  sustainability  has  emerged.  The  general  idea  is  that,  as  society  progresses,  we  should  do  better,  or  at  least  less  bad.  This  has  manifested  across  the  social  spectrum,  from  corporations  to  consumers  to  governments.  An  array  of  drivers  and  responses  to  sustainability  has  started  changing  how  we  organize  human  civilization.  I  call  this,  “the  sustainability  shift.”    It  is  no  surprise,  then,  that  the  sustainability  shift  has  intersected  its  media  counterpart.  The  impact  of  consuming  the  written  word  should  be  taken  into  account  and,  wherever  possible,  reduced.  But  do  we  know  for  certain  that  the  media  shift  is  sustainable?  Have  we  adequately  measured  and  verified  the  relative  footprints  of  paper  and  digital  media?      This  thesis  was  motivated  by  the  phrase  “please  consider  the  environment  before  printing  this  email.”  But  instead  of  deciding  in  the  moment  on  whether  to  hit  the  print  button,  I  took  a  deeper  approach.  What  does  it  mean  to  consider  the  environment?  And  why  is  it  the  printing  of  an  email,  rather  than  its  transmission,  that  is  worthy  of  consideration?  Clearly  something  is  afoot.  Other  choices  that  involve  paper  versus  digital  media  have  experienced  similar  characterizations.  For  example,  the  idea  that  a  printed  magazine  is  the  “dead  tree  edition”  of  a  digital  equivalent  is  particularly  prevalent.                   101By  exploring  how  digital  media  came  to  be  seen  as  preferable  to  paper  alternatives,  this  thesis  seeks  to  add  to  my  understanding  of  sustainability.  Many  foresters  care  about  the  environment,  perhaps  more  than  most.  The  idea  that  forest  products  are  environmentally  inferior  to  digital  alternatives  might  lead  some  foresters  to  take  umbrage.  Although  this  thesis  is  not  a  counseling  exercise  for  offended  foresters,  it  is  an  earnest  attempt  to  pause  and  consider  the  actual  environmental  impacts  of  media  consumption.      Our  consideration  of  the  environment  had  three  phases:  I  began  by  examining  a  paper  media  supply  chain  in  the  forest  products  industry.  Following  this,  I  reviewed  the  media  consumption  habits  and  environmental  values  of  consumers.  I  concluded  with  an  analysis  of  academic  efforts  to  compare  paper  and  digital  media.  This  final  chapter  weaves  these  investigations  together,  connecting  them  to  the  research  objectives,  offering  ideas  for  potential  applications  and  further  research,  and  identifying  the  strengths  and  limitations  of  my  research.  I  discuss  two  concepts  –  industrial  ecology  and  capability  maturity  –  that  offer  insights  into  the  comparison  of  paper  and  digital  media,  and  deepen  an  understanding  of  sustainability.      Earlier  in  the  thesis,  I  presented  a  flow  chart  identifying  the  key  research  objectives  and  questions.  In  Figure  26,  I  present  a  modified  version  of  this  chart,  identifying  the  key  conclusions  of  each  research  chapter  and  how  these  ideas  contributed  to  the  two  concepts  –  industrial  ecology  and  capability  maturity  –  that  I  discuss  in  this  chapter.                       102Figure  26:  Summary  of  research  chapters  and  conclusions        5.1.1 Sustainability  Perspective:  Media  Supply  Chains    Our  first  research  question  was,  “how  does  sustainability  operate  along  a  paper  media  supply  chain?”  I  found  that  the  supply  chain  is  critical  in  managing  environmental  footprints.  Every  supply  chain  is  slightly  different.  The  specific  actors,  their  location,  their  management  practices,  and  the  extent  to  which  they  collaborate  are  all  important  in  efforts  to  manage  and  reduce  environmental  footprints.  I  found  that,  for  example,  the  location  of  Catalyst  Paper  in  British  Columbia  allowed  paper  to  be  manufactured  with  clean  energy,  in  this  instance,  hydroelectric  power.  I  also  found  that  the  transportation  networks  in  British  Columbia  helped  reduce  the  footprint  of  Catalyst’s  product.  In  particular,  the  easy  access  to  sea-­‐             103based  shipping  and  rail  networks  helped  reduce  the  amount  of  inefficient  transportation  by  truck.  I  also  found  that  Catalyst,  by  locating  its  mills  in  areas  of  relative  water  abundance,  was  able  to  strike  the  right  balance  between  energy  and  water  use  in  the  paper  making  process.  All  of  these  factors  together  helped  improve  the  footprint  of  Catalyst’s  product.      But  another  trend  was  at  play  too:  Catalyst  was  trying  to  produce  a  paper  product  with  the  smallest  carbon  footprint  possible.  And  without  collaborating  with  its  supply  chain,  it  would  never  have  been  able  to  do  so.  Catalyst’s  own  staff  worked  closely  with  key  suppliers  –  including  Western  Forest  Products,  Washington  Marine  Group,  and  Burlington  Northern  Santa  Fe  Railways  –  to  optimize  the  energy  used  in  producing  and  delivering  its  product.  This  collaboration  on  environmental  issues  underscores  a  key  driver  of  sustainability:  when  “win-­‐win”  solutions  are  available,  businesses  will  work  together  in  order  to  improve  environmental  performance.  In  this  instance,  win-­‐win  solutions  were  found  when  carbon  emissions  reductions  were  sought.  Carbon  is,  for  all  intents  and  purposes,  analogous  with  energy,  and  using  energy  costs  money.  By  collaborating  together,  the  Catalyst  supply  chain  was  able  to  reduce  emissions  and  costs.      Reducing  emissions  together  was  just  the  first  step  in  supply  chain  collaboration  I  observed.  By  being  more  efficient,  I  found  that  supply  chain  actors  began  a  process  of  becoming  more  responsible.  Washington  Marine  Group,  for  example,  cites  Catalyst’s  request  for  information  on  carbon  emissions  as  a  key  driver  in  evaluating  energy  use  across  its  entire  fleet.  But  the  move  from  efficiency  to  responsibility  was  only  the  first  step.  The  interviews  with  supply  chain  actors  revealed  that,  through  close  collaboration,  the  supply  chain  was  actually  becoming  more  resilient.  When  I  spoke  to  executives  and  managers,  it  was  late  2008  and  early  2009.  The  financial  crisis  was  in  full  force,  and  almost  all  corporate  actors  were  concerned  for  their  own  well  being.  Despite  this  existential  threat,  they  suggested  that  cooperating  together  to  reduce  carbon  emissions  had  actually  strengthened  relationships  along  their  supply  chains.      Our  findings,  to  a  large  extent,  validate  the  existing  literature  on  how  corporations  respond  to  sustainability.  In  my  review  of  “the  sustainability  shift”,  I  discussed  a  variety  of  different  approaches  a  firm  can  take  to  manage  for  environmental  variables,  using  a  framework  developed  by  Auld  et  al.  (2008a),  who  identified  seven  categories  that  capture  corporate  responses  to  sustainability:  individual  firm  efforts;  individual  firm  and  NGO  agreements;  public-­‐private  partnerships;  information-­‐based               104approaches;  environmental  management  systems  (EMSs);  industry  association  codes  of  conduct;  and  non-­‐state  market-­‐driven  (NSMD)  governance  in  the  form  of  private-­‐sector  hard  laws.      Of  these  seven  categories,  my  case  study  validated  at  least  four  concepts.  I  saw  evidence  of  individual  firm  efforts  to  promote  environmental  responsibility,  as  Catalyst  did  when  it  decided  to  purchase  carbon  offsets  to  reduce  the  footprint  of  its  operations.  I  also  saw  firms  working  with  ENGOS,  in  this  instance  Washington  Marine  Group  and  WWF  Canada,  an  approach  identified  by  Auld  et  al.  (2008a)  and  Yaziji  and  Doh  (2009)  that  I  observed  in  practice.  I  observed  information-­‐based  approaches  in  action,  as  supply  chain  partners  exchanged  information  about  environmental  footprints  and  operational  logistics  in  order  to  manage  and  reduce  impacts.  Finally,  I  saw  EMSs  deployed  to  help  measure  and  gather  all  the  relevant  data  needed  to  take  meaningful  action.    This  case  study  was  a  targeted  investigation  into  one  supply  chain  on  the  west  coast  of  North  America.  But  it  revealed  a  few  important  things  about  sustainability  and  media  that  are  worth  noting:    • Not  all  media  is  created  equally:  location  and  collaboration  can  strongly  influence  the  footprint  of  a  media  product.    • Supply  chain  collaboration  is  key  to  reducing  environmental  footprints:  without  transparency  and  engagement,  companies  can  only  ever  reduce  their  own  direct  emissions.  Working  together  unlocks  opportunities  to  improve  environmental  performance.    • The  media  supply  chain  that  I  examined  demonstrated  maturity  and  capability  in  its  management  of  environmental  performance.  I  found  the  motivation,  capacity,  and  managerial  oversight  required  to  promote  environmental  sustainability.      5.1.2 Sustainability  Perspective:  Consumer  Media  Habits    Our  second  research  question  was,  “has  sustainability  impacted  consumers  media  consumption  habits?”  To  test  this  idea,  I  designed  a  survey  for  1,435  consumers  in  North  America,  with  a  sample  that  mirrored  census  results  in  Canada  and  the  United  States.  Four  segments  were  created:  Digital  Light,  Digital  Heavy,  NEP  Eco,  and  NEP  Anthro.  I  devised  a  scoring  system  to  approximate  the  digital  media  consumption  habits  of  the  digital  segments.  The  NEP  segments  were  produced  using  the  New  Ecological  Paradigm  (NEP)  (Dunlap  et  al.,  2000),  a  commonly  employed  academic  framework  for  representing  the  environmental  values  of  a  survey  sample.                   105I  found  that  consumers  are,  indeed,  shifting  from  paper  to  digital  media  sources.  Not  only  did  all  segments  report  using  more  digital  media  today  than  five  years  ago,  all  segments  expect  the  shift  to  accelerate.  This  reaffirms  the  concept  of  “the  media  shift”,  validates  findings  from  the  literature  (Duggan  and  Smith,  2013;  Duggan  and  Brenner,  2013;  Bell  et  al.,  2013),  and  gives  credence  to  my  efforts  to  consider  the  environmental  footprint  of  media.      What  I  wanted  to  investigate,  however,  was  whether  environmental  values  had  any  effect  over  media  consumption.  In  the  review  of  “the  sustainability  shift”,  I  found  that  consumers  preferred  sustainable  products  (Diamantopoulos  et  al.,  2003;  D'Souza  et  al.,  2007)  and,  in  some  instances,  are  willing  to  pay  a  premium  for  products  (Ottman,  2011).  I  surmised  that  there  would  be  some  relationship  between  environmental  values  and  media  consumption:  the  prevalence  of  the  phrase,  “please  consider  the  environment”  suggests  that  people  connect  the  environment  with  their  media  consumption  all  the  time.  But  to  my  surprise,  I  could  find  almost  no  evidence  to  suggest  that  the  two  variables  interact.  Those  with  stronger  environmental  values  had  media  consumption  habits  that  were  almost  identical  to  those  with  weaker  environmental  values.  This  suggests  that,  for  whatever  reason,  concern  for  the  environment  plays  no  role  in  media  consumption  habits.  Why  might  this  be?    Media  –  both  paper  and  digital  –  is  somewhat  transient.  Digital  media  arrives  at  the  click  of  button,  hosted  on  the  vast  network  of  ICT  that  hums  away  in  the  background  of  today’s  economy.  Digital  products  appear  and  disappear  at  the  whim  of  the  consumer,  and  there  are  few  (if  any)  moments  when  the  aggregate  impact  of  all  this  digital  media  consumption  might  occur  to  consumers.  In  many  ways,  paper  media  is  similar.  It  arrives  at  the  door,  is  consumed,  then  stored  on  a  shelf  or  tossed  it  into  a  recycling  bin.  The  impacts  of  paper  media  consumption  are  removed  from  the  consumer,  although  perhaps  less  than  its  digital  counterpart.      That  I  observed,  for  the  most  part,  no  connection  between  the  NEP  segments  and  media  consumption  is  an  interesting  contribution  to  the  literature  on  the  NEP.  Many  academics  have  used  it  to  segment  and  explore  the  habits  of  different  groups.  Ardahan  (2012)  and  Amburgey  et  al.  (2012)  found  the  NEP  a  meaningful  scale  for  exploring  the  habits  of  consumers  pursuing  recreational  activities  in  Turkey.  Velayudhan  and  Srividya  (2013)  deployed  the  scale  in  India  to  understand  the  environmental  perspectives  of  consumers  there.  Kopina  (2012)  and  Corraliza  et  al.,  (2013)  both  employed  the  NEP  scale  to  segment  and  understand  the  development  of  children.  In  short,  the  NEP  has  provided  a  reliable               106indicator  of  environmental  preferences  amongst  a  variety  of  different  segments.  But  in  my  research,  I  found  that  the  NEP  did  little  to  elucidate  the  preferences  of  different  segments  of  media  consumers.  This  reinforces  the  concept  that  there  is  weak  connectivity  between  consumers  and  media.    If  media  is  transient  and  consumers  environmental  values  don’t  influence  media  consumption  patterns,  it  follows  that  consumers  are  disconnected  from  their  own  media  footprint.  This  lack  of  connectivity  does  much  to  explain  why  the  footer,  “please  consider  the  environment  when  printing  this  email”  is  so  popular.  In  a  world  where  consumers  have  little  sway  over  the  environmental  impacts  of  entire  industrial  systems,  they  do  have  a  choice  when  it  comes  to  hitting  print.  Perhaps  then,  consumers  want  to  care  about  the  footprint  of  their  media  consumption,  but  they  are  faced  with  few  opportunities  to  do  so.  When  considering  the  environment,  how  connected  a  consumer  is  to  the  impacts  of  their  media  habits  must  be  taken  into  account.        This  case  of  consumers  revealed  three  important  things  about  media  and  the  environment:  • It  affirms  that  a  shift  from  paper  to  digital  media  is  under  way.    • It  demonstrates  no  meaningful  connection  between  environmental  values  and  media  consumption.    • It  suggests  that  the  transient  nature  of  media  leads  to  weak  connections  between  consumers  and  their  media  choices,  explaining  the  disconnect  between  environmental  values  and  media  habits.    5.1.3 Sustainability  Perspective:  Comparing  Media  Choices      Our  final  research  question  was  to  ask,  “has  sustainability  compelled  academics  to  compare  different  media  types?”  As  consumers  are  shifting  from  one  media  type  to  another,  I  assessed  academics’  comparisons  of  the  environmental  attributes  of  this  shift.  In  this  analysis,  I  focused  on  digital  media.  I  wanted  to  investigate  how  the  standard  academic  approach  of  measuring  and  comparing  environmental  footprints,  the  life  cycle  assessment  (LCA),  handles  a  comparison  between  two  very  dissimilar  products,  and  how  the  LCA  methodology  copes  with  measuring  complex  ICT  products.      I  found  that,  with  one  exception  (Enroth,  2009),  LCAs  determine  that  digital  media  is  environmentally  preferable  to  paper  alternatives.  The  one  exception  involved  a  textbook  in  elementary  school  that  is               107shared  several  times.  Only  after  several  consecutive  years  of  textbook  use  does  the  footprint  of  paper  seem  preferable  to  using  a  computer  in-­‐class.      But  more  interesting  than  the  findings  of  the  LCAs  was  everything  the  authors  did  not,  or  could  not,  discuss.  For  example,  the  rapid  expansion  of  ICT  infrastructure  and  devices  was  never  discussed.  This  established  and  global  trend  was  never  connected  –  explicitly  or  implicitly  –  to  the  footprint  of  digital  media.  Nor  was  the  nature  of  the  ICT  industry  itself  ever  considered.  ICT  is  evolving  rapidly,  best  captured  by  the  concept  of  Moore’s  Law.  Every  couple  of  years,  the  efficiency  of  ICT  doubles,  while  its  cost  halves.  This  leads  to  remarkable  technological  innovation.  It  also  leads  to  a  rapid  cycle  of  obsolescence,  as  older  ICT  is  discarded  in  favour  of  new  products.  The  fact  that  ICT  is  dependent  on  non-­‐renewable  resources  like  rare  earth  minerals,  metals,  and  plastics  was  never  mentioned,  nor  was  any  connection  between  non-­‐renewable  resource  dependency  and  digital  media  ever  established.  All  of  the  waste  that  ICT  generates  through  its  rapid  cycles  of  innovation  and  ever-­‐increasing  presence  was  never  given  serious  consideration.  While  the  LCAs  would  assume  that  ICT  waste  was  properly  disposed  at  local  facilities,  they  never  considered  the  more  likely  alternative:  that  it  would  be  shipped  overseas  to  be  dismantled  with  grave  environmental  and  human  consequences  (Grossman,  2006;  Robinson,  2009).  Moore’s  Law  also  ensures  that  the  data  being  considered  by  an  LCA  is  almost  always  out  of  date.  Collecting  life  cycle  inventory  data  is  a  slow  and  expensive  process.  With  the  ICT  sector  constantly  innovating,  it  is  no  surprise  that  data  is  often  outdated  on  digital  media;  studies  published  in  2010  were  using  estimates  on  the  footprint  of  the  Internet  from  the  late  1990s.      The  LCA  is  a  methodology  designed  to  consider  discrete  trade-­‐offs  between  one  option  and  another  (Gaudreault  et  al.,  2007a).  But  what  happens  when  it  considers  two  completely  different  media  types,  paper  and  digital?  Paper  media  is  grounded  in  the  world  of  forestry,  with  a  wide  range  of  potential  environmental  impacts.  Digital  media  is  a  virtual  product  dependent  on  a  complex  and  interconnected  global  ICT  system.  Trying  to  figure  out  exactly  what  percentage  of  that  system  should  be  allocated  towards  a  particular  digital  media  experience  is  akin  to  counting  angles  on  the  head  of  a  pin.      I  cannot  definitively  say  where  the  limitations  of  LCAs  are  borne.  Is  it  the  lack  of  precise  data;  the  nature  of  ICT;  the  challenge  of  modelling  consumer  behaviour;  or  in  the  difficulty  of  comparing  two  totally  different  things?  But  the  aggregate  effect  is  a  series  of  studies  that  reveal  little  about  the  relative  attributes  of  paper  and  digital  media.  Instead,  they  reveal  more  about  what  I  don’t  know  than  what  I  do.               108It  is  not  whether  the  LCAs  are  a  fair  and  accurate  representation  of  available  data.  It  is  whether  available  data  is  of  sufficient  quality  to  inspire  confidence  in  the  study  findings.  This  research  suggests  that  these  LCAs,  rather  than  provide  definitive  answers,  are,  in  fact,  scoping  exercises  that  do  much  to  reveal  how  little  is  actually  known.  Should  these  studies  be  relied  upon  to  determine  what  is  more  sustainable?      Our  findings  validate  research  (Plepys,  2004;  Williams  et  al.,  2002)  that  suggests  there  is  a  hidden  footprint  related  to  ICT  that  current  methods  do  not  capture.  Plepys  (2004)  identifies  chemical  purity  as  a  particular  concern,  citing  the  fact  that  LCAs  use  data  for  wholesale  grade  chemicals,  rather  than  the  ultra-­‐pure  chemicals  that  semiconductor  manufacturing  actually  requires.  Lu  et  al.  (2006)  rely  on  data  from  1998  in  order  to  measure  the  footprint  of  a  laptop  computer,  underlying  the  difficulty  of  finding  contemporaneous  data.  The  backbone  of  ICT  –  the  Internet  –  also  proved  a  challenge  to  measure:  most  of  the  LCAs  I  reviewed  refrained  from  measuring  it,  or  if  they  did,  expressed  little  confidence  in  their  findings:  “the  results  concerning  potential  impact  of  using  telecommunication  infrastructure  are  too  uncertain  to  draw  any  real  conclusions  from”  (Moberg  et  al.,  2010,  p.42).  Studies  measuring  the  footprint  of  digital  products,  therefore,  mostly  ignore,  out  of  necessity  not  neglect,  the  entire  backbone  of  ICT  infrastructure  that  actually  makes  digital  media  possible.  Without  the  Internet,  digital  media  would  not  be  able  to  be  delivered  with  such  speed  and  at  such  volume.  And  yet  studies  of  digital  media  can’t  even  measure  the  Internet’s  footprint.      Relating  back  to  the  objectives  of  this  thesis,  it  is  clear  that,  when  considering  the  environment,  researchers  should  be  more  ambitious  than  simply  considering  a  single  discrete  trade-­‐off  between  one  product  and  another.  There  is  a  pressing  and  tangible  need  to  consider  the  industrial  systems  that  produce  media.  And  the  ICT  that  produces  digital  media  is  the  world’s  most  complex  and  dynamic  industrial  sector,  presenting  a  challenge  to  any  effort  to  measure  its  environmental  footprint.  The  three  key  conclusions  that  emerged  from  my  comprehensive  review  of  comparative  LCAs  were:  • LCAs  comparing  paper  and  digital  media  almost  always  find  the  latter  to  be  environmentally  preferable.  • The  assumptions  required  to  compared  two  vastly  different  products  undermine  the  focus  and  effectiveness  of  LCAs.    • The  uncertainty  introduced  by  measuring  the  ICT  sector  –  which  is  global,  complex,  and  constantly  evolving  –  makes  a  macro-­‐level  study  of  an  ICT  product  like  digital  media  particularly  difficult.                 1095.2 POTENTIAL  APPLICATIONS  AND  FUTURE  RESEARCH  Three  key  themes  have  emerged  from  this  thesis,  themes  that  when  considered  together  provide  opportunities  for  future  research  and  applications  in  the  real  world.  If  the  findings  of  Chapters  2,  3,  and  4  were  distilled  to  their  most  basic  elements  the  following  concepts  would  emerge:  • Location  and  collaboration  along  a  supply  chain  matter  in  measuring,  managing  and  improving  environmental  footprints.  • Consumers  are  detached  from  their  environmental  impacts,  unless  an  easy  decision  with  an  obvious  material  impact,  like  printing  an  email,  is  available  to  them.  • Discrete  trade-­‐offs  involving  comparisons  between  two  totally  different  products  only  partially  advances  an  understanding  of  sustainability:  there  is  a  need  to  think  systemically  as  well.    The  second  theme  is  perhaps  the  most  obvious,  but  the  least  fruitful  for  future  research.  Consumers  are  a  busy  group,  and  expecting  an  evolved  and  sophisticated  understanding  of  sustainability  is  probably  a  bit  much.  But  the  first  and  third  themes  –  the  power  of  supply  chains,  and  the  need  to  consider  industrial  systems  –  offer  powerful  concepts  and  opportunities  for  further  work.      To  better  understand  the  sustainability  of  media,  this  thesis  offers  two  concepts  for  consideration.  The  first  has  already  been  discussed:  industrial  ecology.  This  is  the  notion  that  industries  can  be  thought  of  as  systems  with  the  potential  to  minimize  waste  and  maximize  efficiency,  mimicking  nature  as  closely  as  possible.  The  second  is  a  new  concept,  but  one  that  is  already  used  in  the  world  of  business:  capability  maturity  models.  These  concepts  are  presented  in  turn,  connecting  them  to  the  literature  as  well  as  my  own  research  in  order  to  see  how  they  might  potentially  be  applied.      5.2.1 Industrial  Ecology  I  have  already  reviewed  the  concept  of  industrial  ecology  (see  Chapter  1.1.1.1.1),  but  it  is  worth  revisiting.  Given  the  LCA’s  limited  abilities  in  comparing  completely  disparate  products  (paper  and  media),  as  well  as  its  weakness  in  handling  the  complexity  of  ICT,  industrial  ecology  may  offer  a  broader,  complimentary  perspective  to  my  understanding  of  sustainability.  It  is  founded  on  the  idea  that  much  can  be  learned  from  natural  systems  to  improve  the  environmental  footprint  of  industrial  systems.  Given  that  industrial  systems  have  had  a  brief  history,  at  least  compared  to  ecosystems,  taking  a  closer  look  at  the  structures  and  successes  of  nature  can  provide  insights  into  how  to  devise  more  effective  and  sustainable  systems  (Seager,  2008).               110Figure  27  shows  three  industrial  ecologies.  The  first,  Type  I,  is  an  open  loop,  with  energy  and  materials  flowing  in,  being  processed,  and  flowing  out.  It  assumes  an  unlimited  availability  of  resources,  and  an  unlimited  ability  for  the  planet  to  absorb  the  impacts  of  industrial  activities.  Type  II  describes  most  existing  industrial  systems  as  they  occur  today.  There  are  limited  supplies  of  materials  and  energy,  and  industry  cannot  pollute  freely  without  incurring  costs.  A  completely  closed  loop  Type  III  system  is  impossible  for  industry,  as  total  resource  conservation  is  unachievable.  However,  it  is  something  that  industry  can  aspire  to.  Ecological  systems  are  Type  III  systems,  as  nature  has  only  one  input  (solar  energy),  and  waste  does  not  exist  as  every  material  is  eventually  reused  by  the  system  (Gradel  and  Allenby,  1995;  Erkman,  1997).    Figure  27:  Industrial  ecology  types        When  comparing  paper  and  digital  media,  what  might  a  framework  of  industrial  ecology  say?  Are  forestry  and  ICT  Type  I  or  Type  II  systems?  And  if  a  method  for  measuring  the  “strength”  of  a  system  could  be  devised,  would  one  system  be  stronger  than  the  other?                   111Forestry,  in  the  best  case  scenario,  depends  on  sustainably  manage  forests  that  produce  a  product  (trees)  that  can  be  renewed  in  perpetuity  (Boltz  et  al.,  2003).  The  processing  of  trees  can  be  done  with  renewable  energy,  like  biomass  or  hydro  (Bajpai,  2011),  and  the  goods  it  produces,  from  pulp  to  dimensional  lumber,  can  be  recycled  at  the  end  of  their  useful  life  (Merrild  et  al.,  2009).  In  the  best  case  scenario,  forestry  has  the  potential  to  be  a  strong  Type  II  industrial  ecology.      ICT,  on  the  other  hand,  is  more  complex.  It  depends  entirely  on  non-­‐renewable  resources  like  metals,  plastics,  and  rare-­‐earth  minerals  (Lau  et  al.,  2002).  It  is  dispersed  and  dynamic,  and  best  case  scenarios  may  be  hard  to  define.  Because  of  the  need  to  compete  on  cost  as  well  as  specialize,  very  few  ICT  firms  are  vertically  integrated  (Christiaanse  and  Kumar,  2000).  Instead,  they  rely  on  a  massive  network  of  suppliers,  contractors  and  subcontractors  (Paija,  2000),  and  enforcing  environmental  performance  under  these  conditions  is  difficult.  ICT  is  also  doing  much  more  than  simply  replacing  paper  media.  It  has  spawned  a  massive  wave  of  innovation  across  the  economy  –  from  financial  services  to  entertainment  to  car  sharing  –  that  is  adding  economic  value  and  potentially  diminishing  environmental  impact.  The  diffuse  and  disruptive  nature  of  ICT  makes  an  assessment  of  its  industrial  ecology  particularly  difficult.  ICT  uses  large  volume  of  non-­‐renewable  resources,  requires  toxic  chemicals  to  produce,  and  results  in  waste.  But  it  is  also  reshaping  the  face  of  the  global  economy.      Many  industries  have  benefited  from  the  efficiency,  information  availability,  and  new  technology  that  ICT  has  provided  (Yi  and  Thomas,  2007).  ICT  itself  has  become  more  efficient,  just  as  Moore’s  Law  would  suggest.  Devices  are  getting  smaller  and  smaller,  using  fewer  raw  materials  as  a  result  (Williams  et  al.,  2002).  Products  like  computer  monitors  and  televisions  used  to  be  large  and  heavy,  full  of  glass  and  lead  and  other  materials.  Today,  they  are  light  and  made  mostly  of  plastics.  In  a  way,  ICT  is  now  doing  more  with  less.  The  forest  sector  itself  has  also  benefited  greatly  from  ICT  in  developing  more  efficient  supply  chains  and  industrial  processes  (Kollberg,  2005).  But  some  argue  that  any  resource  efficiency  that  has  been  unlocked  by  ICT  is  more  than  offset  by  the  increase  in  economic  activity,  as  classic  example  of  Jevon’s  Paradox  (Polimeni,  2009).  So,  there  is  a  tension  in  understanding  ICT:  on  the  one  hand,  it  is  getting  increasingly  more  efficient.  On  the  other,  it  unlocks  more  and  more  economic  growth  with  commensurate  environmental  impacts.  This  paradox  is  not  a  question  to  be  answered,  but  a  feature  of  ICT  to  be  acknowledged.                   112The  potential  of  industrial  ecology  is  in  helping  economic  actors  take  a  deeper  and  more  contextual  assessment  of  their  own  environmental  behaviours.  Beyond  simply  measuring  productivity  and  efficiency,  industrial  ecology  offers  a  concept  to  aspire  towards:  maximize  renewability  and  minimize  waste.  This  concept  does  not  preclude  other  methods  of  environmental  assessment  from  being  used.  In  fact,  given  that  its  strengths  are  conceptual  rather  than  quantitative,  other  methods  should  be  encouraged.  But  as  I  demonstrated  in  Chapter  4,  when  the  same  environmental  tool  (in  this  instance  the  LCA)  is  used  again  and  again  to  answer  basically  the  same  question,  key  parts  of  the  puzzle  can  be  missed.      In  the  case  of  paper  or  digital  media,  the  context  of  paper  and  ICT  media  was  simply  ignored.  Forestry’s  potential  strengths  –  renewability,  recyclability,  and  strong  best  case  scenarios  –  went  unrecognized  by  any  over  the  research.  And  ICT’s  inexorable  pace  of  innovation,  complexity,  resource  consumption,  and  growth  eluded  all  of  the  LCAs  consideration.  This  is  not  to  say  that  every  LCA  should  also  contain  a  lengthy  discourse  on  the  conceptual  and  contextual  aspects  of  sustainability  surrounding  a  particular  decision,  but  somewhere  in  the  academic  landscape  such  review  should  take  place.  This  dissertation  has  acted  as  such  a  review,  but  more  than  that,  has  identified  a  need  and  opportunity  to  do  so  at  a  larger  and  deeper  scale.      Industrial  Ecology,  and  the  core  concepts  it  contains,  offers  the  opportunity  to  strengthen  considerations  of  the  environment.  If  environmental  assessments  are  not  considered  in  isolation,  but  instead  framed  as  part  of  the  larger  ecology  of  an  industry  or  economy,  there  is  an  opportunity  to  promote  truly  positive  outcomes.  Relying  exclusively  on  limited  and  narrow  methods  of  assessment  carries  the  risk  only  passively  promoting  sustainable  outcomes,  ignoring  the  opportunity  to  conduct  holistic  assessments  that  consider  how  industries  organize  themselves.  The  ICT  industry  could  use  industrial  ecology  as  an  aspirational  tool,  promoting  the  increased  use  of  renewable  and  recyclable  materials.  But  in  order  to  do  so,  it  will  need  to  evolve  and  strengthen  its  supply  chains’  ability  to  manage  environmental  criteria.      5.2.2 Capability  Maturity  Models  Capability  maturity  models  (CMM)  are  a  concept  that  emerged  from  the  world  of  software  development  (Humphrey,  1989).  Institutions  all  over  the  world  test  the  resilience  of  their  processes  and  procedures  using  CMM  (Paulk,  1993;  Paulk,  1995).  Governments,  consulting  agencies,  and  procurement               113departments  use  CMM  to  improve  their  performance  (Fraser  et  al.,  2002).  The  idea  is  that,  over  time,  policies  and  procedures  can  be  developed  that  increase  the  capability  of  a  particular  business  function.  At  the  outset,  it  may  begin  with  simply  defining  necessary  steps.  But  over  time,  metrics  may  be  developed.  These  metrics,  once  fully  understood  can  lead  to  the  active  optimization  of  a  particular  process.    There  is  an  opportunity  to  apply  this  concept  to  media  supply  chains,  or  any  other  industrial  sector.  In  the  supply  chain  case  study  (Chapter  2),  I  saw  the  power  of  specific  actors  working  together  to  manage  environmental  performance.  The  more  they  collaborated,  the  more  efficient,  responsible,  and  resilient  their  supply  chain  became.  In  a  way,  it  became  more  capable  and  mature.  I  also  saw  (Chapter  4)  that  in  the  ICT  industry,  an  evolving  and  complex  industrial  system  makes  it  difficult  to  measure  environmental  performance.  The  question  is  how  might  the  idea  capability  maturity  be  used  along  media  supply  chains?    Consider  the  perspective  of  a  major  media  buyer,  like  a  global  retailer  with  a  catalogue  distributed  worldwide.  They  may  see  a  material  risk  associated  with  their  media  purchases  and  decide  to  minimize  that  risk.  This  means  that  they  want  to  avoid  sourcing  from  irresponsible  suppliers.  If  that  retailer  were  found  to  have  used  illegal  fibre  from  a  high  conservation  value  forest,  it  would  damage  its  reputation  and  brand.  So  the  buyer  would  insist  only  on  buying  certified  fibre,  the  first  step  in  developing  capability  maturity  along  its  supply  chain.  The  next  step  might  be  asking  for  enterprise-­‐level  reporting  from  all  its  suppliers  on  energy  use  and  carbon  emissions,  along  with  a  plan  to  reduce  environmental  impacts.  The  buyer  could  go  further,  and  insist  not  only  on  enterprise-­‐level  reporting,  but  site-­‐level  (e.g.  a  pulp  mill)  or  product  level  (e.g.  a  tonne  of  paper)  data.  The  final  step  would  be  to  insist  that  all  suppliers  audit  and  verify  their  product  level  data,  bringing  in  third-­‐party  verification  where  appropriate.  Verified  product-­‐level  data  with  third-­‐party  input  would  indicate  a  very  mature  and  capable  supply  chain  indeed.    The  same  thought  experiment  could  be  applied  to  the  ICT  sector.  How  would  that  media  buyer  ensure  its  digital  presence  has  the  best  environmental  performance  possible?  They  would  start  by  contacting  the  data  center  operators  that  host  its  website  and  digital  applications.  The  same  enterprise  to  site  to  product  level  data  rubric  could  be  applied.  Insisting  on  verified  data  could  also  be  an  option.  But  ICT  is  unlike  forestry.  The  networks  are  much  more  diffuse,  and  simply  talking  to  data  center  operators  by  no  means  implies  that  the  environmental  impact  of  digital  media  has  been  measured  and  managed  to  the               114fullest  extent  possible.  Further,  forestry  is  an  industry  that  has  withstood  public  scrutiny  for  a  long  time.  Environmentalists  have  long  been  engaged  in  promoting  sustainable  forestry,  while  also  publicly  criticizing  companies  that  engage  in  practices  that  are  damaging  to  the  environment.  As  a  result,  the  forest  products  industry  has  had  to  develop  certification  schemes  and  transparent  supply  chain  practices  in  order  to  secure  its  social  license  to  operate.      The  ICT  industry  has  only  recently  to  endure  this  kind  of  scrutiny.  As  work  on  this  dissertation  unfolded,  new  legislation  in  the  United  States  emerged  asking  companies  to  report  on  so-­‐called  “conflict  minerals”  in  their  supply  chain.  With  the  passage  of  the  Dodd-­‐Frank  Wall  Street  Reform  Act,  a  series  of  rules  was  introduced  to  ensure  that  ICT  companies  disclose  their  use  of  these  minerals  (SEC,  2012).  Conflict  minerals  are  sourced  from  the  Congo  Basin,  and  include:  columbite-­‐tantalite,  used  in  the  production  of  capacitors;  cassiterite,  an  ore  used  to  produce  tin,  essential  for  the  production  of  solder  on  the  circuit  boards  of  electronics;  wolframite,  a  source  of  tungsten  which  is  used  in  the  vibrator  function  on  cell  phones;  and  gold.  Research  suggests  (Ochoa  and  Keenan,  2011)  that  the  dispersed  and  complex  global  supply  chain  for  metals  and  minerals  will  make  it  difficult  for  ICT  companies  to  distinguish  a  legitimate  source  of  a  mineral  form  one  that  is  prohibited.  ICT  companies  are  working  (Motorola,  2014)  to  identify  legal  suppliers  and  are  working  together  to  jointly  improve  their  procurement  practices.  But  progress  has  been  slow,  with  a  recent  report  suggesting  that  most  companies  are  not  adequately  investigating  their  supply  chains  to  identify  illegal  sources  (Michaels,  2014).  Intel,  a  prominent  American  semiconductor  manufacturer,  has  met  significant  resistance  in  its  efforts  to  weed  out  conflict  minerals  (King,  2014).      The  difficulty  in  sourcing  just  a  few  minerals  demonstrates  the  difficulties  of  managing  environmental  variables  in  the  ICT  sector.  Given  the  massive  number  of  minerals  and  metals  present  in  ICT  products  (Lau  et  al.,  2002)  and  the  increasing  complexity  of  the  processes  and  chemicals  used  (Plepys,  2004),  it  follows  that  the  ICT  industry  is  going  to  have  a  difficult  time  developing  both  capacity  and  maturity  in  managing  environmental  footprints  along  its  supply  chain.  But  these  are  questions  that  the  LCAs  reviews  in  Chapter  4  neglected  to  ask.  In  the  rush  to  determine  what  was  better  –  paper  or  digital  –  the  authors  ignored  the  context  of  ICT.  But  given  what  I  learned  in  Chapter  2  when  studying  a  paper  media  supply  chain,  ignoring  the  context  has  consequences:  specifics,  location,  and  collaboration  all  matter.  And  along  paper  media  supply  chains,  there  is  demonstrated  capacity  and  maturity  for  grappling  with  environmental  issues.                 115  To  return  to  the  example  of  a  media  buyer:  by  using  a  common  framework  along  both  paper  and  ICT  supply  chains,  a  buyer  might  discover  that  one  industry  is  more  capable  than  the  other  at  handling  environmental  issues.  Should  this  and  does  this  impact  a  consideration  of  the  relative  footprint  of  paper  and  digital  media?  I  have  seen  throughout  the  literature  (Gauthier,  2005;  Auld  et  al.,  2008a;  Dauvergne  and  Lister,  2012)  that  companies  play  a  particularly  critical  role  in  promoting  sustainability.  Good  environmental  performance  is  dependent  on  the  responsible  and  transparent  behaviour  of  private  actors.  This  suggests  that  considerations  of  the  environment  should  include  an  assessment  of  the  capability  maturity  of  a  particular  supply  chain  or  industry.  Further  developing  the  concept  of  capability  maturity  models  is  a  promising  field  of  research.    5.3 STRENGTHS  AND  LIMITATIONS    The  strength  of  this  research  is  its  willingness  to  take  diverse  research  and  connect  it  together  in  order  to  advance  a  definition  of  sustainability.  It  identifies  the  role  of  supply  chains,  consumers,  and  life  cycle  comparisons  that  consider  the  environmental  impacts  of  media.  By  connecting  case  studies,  consumer  surveys,  and  comparative  analyses  conducted  by  others,  it  advances  the  understanding  of  what  sustainable  media  means.      But  this  thesis  is  just  one  in  an  effort  to  expand  and  refine  theories  of  sustainability.  It  is  clear  that  “doing  better”  (or  “doing  less  bad”)  is  a  complex  concept  that  can  mean  different  things  to  different  people.  Industry,  consumers,  and  academics  all  have  their  own  unique  contributions  that  have  to  be  taken  into  account.  Industry  operates  in  a  profit-­‐driven  world  and  is  often  loathe  taking  steps  to  improve  environmental  performance  unless  there  are  tangible  benefits  to  the  bottom  line.  Consumers,  despite  all  their  good  intentions,  do  not  have  the  time,  resources,  or  skills  to  fully  assess  the  sustainability  of  their  decisions.  And  when  that  decision  involves  two  wildly  different  products  –  paper  and  digital  media  –  it  is  not  reasonable  to  expect  a  particularly  sophisticated  decision-­‐making  process.  Academics  are  compelled  by  scientific  methods  and  the  need  for  testable  hypotheses.  They  develop  standardized  and  rigorous  methodologies  like  the  LCA  that  are  transparent  and  driven  by  data.  But  when  data  is  scarce  and  uncertainties  and  assumptions  arise,  academic  methods  can  fail  to  see  the  forest  for  the  trees.  In  this  instance,  the  LCAs  I  reviewed  did  not  grapple  with  the  complex  context  of  ICT  and  comparative  media  choices.  The  strength  of  this  thesis  is  to  acknowledge  the  shortcomings  of  different  actors,  but  not  be  swayed  away  from  asking  big  and  difficult  questions.                 116  The  limitations  of  this  research,  however,  are  significant.  While  it  combines  three  focused  and  rigorous  scientific  studies,  connecting  them  together  involves  a  willingness  to  think  conceptually.  But  concepts  of  sustainability  are  nebulous  and  difficult  to  test.  It  is  worth  revisiting  Seager’s  (2008,  p.447)  definition:  “Sustainability  might  best  be  defined  as  an  ethical  concept  that  things  should  be  better  in  the  future  than  they  are  at  present.  Like  other  ethical  concepts  such  as  fairness  or  justice,  sustainability  is  best  interpreted  conceptually  rather  than  technically.”  The  limitation,  then,  is  that  I  am  studying  an  objective  and  scientific  issue  –  environmental  impacts  of  media  consumption  –  but  I  am  advancing  an  ethical  concept.  A  more  focused  effort  –  for  example,  studying  one  part  of  the  ICT  lifecycle  and  trying  to  advance  data  availability  –  could  enjoy  additional  scientific  rigour.  But  it  would  be  primarily  a  technical  effort  and  would  not  achieve  as  much  in  advancing  definitions  of  sustainability  or  providing  concepts  that  might  help  compare  the  attributes  of  industrial  systems.    Instead,  this  dissertation  is  best  considered  as  a  guide  for  more  focused  investigations  into  media  footprints.  While  additional  research  into  environmental  footprints  of  media  consumption  are  quite  necessary,  this  thesis  offers  a  broad  and  contextual  overview  that  can  ground  environmental  investigations  in  a  broader  understanding  of  what  it  means  to  be  sustainable.      In  the  investigation  of  the  paper  supply  chain,  I  acknowledge  that  there  are  limitations  to  the  methods  employed  in  this  research.  Any  extrapolation  from  a  specific  case  study  to  a  more  general  population  should  be  treated  with  caution.  That  said,  I  did  triangulate  the  interview  findings  to  strengthen  their  validity  and,  in  so  doing,  developed  a  framework  that  potentially  extends  beyond  this  case  study.  The  concepts  of  carbon  as  a  catalyst  for  deeper  integration  between  supply  chain  partners,  and  carbon  as  a  starting  point  for  a  transition  from  efficient  to  responsible  to  resilient  supply  chains  has  been  validated  for  this  case  study  of  the  paper  and  print  sectors.  However,  I  cannot  infer  that  these  patterns  will  hold  true  in  all  supply  chains,  although  they  are  likely  to  manifest  in  some.    In  my  investigation  of  media  consumers,  there  are  potential  weaknesses  as  well.  The  NEP  scoring  system,  often  used  to  identify  the  environmental  values  of  a  sample,  is  less  commonly  employed  to  create  two  different  segments.  In  my  research,  the  mean  NEP  score  of  the  two  segments  was  very  similar,  and  the  overall  environmental  values  of  the  two  may  not  have  differed  much.  To  create  the               117Digital  Light  and  Digital  Heavy  segments  I  had  to  devise  my  own  scale  that  has  not  benefitted  from  the  rigor  of  multiple  academic  applications.  And  finally,  in  my  comprehensive  review  of  comparative  LCAs,  the  level  of  detail  and  methodological  background  provided  by  LCA  authors  limited  the  depth  of  comparison  I  was  able  to  undertake.      5.4 FINAL  THOUGHTS  This  thesis  set  out  to  consider  the  environmental  footprint  of  media.  I  reviewed  the  important  role  of  supply  chains  in  improving  environmental  performance.  The  habits  and  values  of  media  consumers  were  established,  showing  a  clear  shift  towards  digital  media,  but  with  little  regard  for  the  environmental  implications.  Finally,  academic  efforts  to  compare  paper  and  digital  media  were  found  wanting,  as  they  failed  to  consider  the  complexity  of  the  comparison  being  made  or  the  industrial  systems  they  were  measuring.  I  have  found  that  considering  the  environment  is  much  more  difficult  than  deciding  whether  to  hit  print  or  not.  It  involves  big,  ethical  concepts  like  sustainability,  as  well  as  complex  and  dynamic  systems  like  ICT  and  forestry.  I  discussed  two  concepts  –  industrial  ecology  and  capability  maturity  models  –  that  should  be  further  developed  by  future  research,  ensuring  that  the  context  of  media  consumption  is  always  considered  in  parallel  with  specific  media  choices.      But  with  all  that  said,  please  consider  the  environment  before  printing  this  thesis.             118Bibliography  Adexa.  (2000)  The  Electronics  Supply  Chain:  Winning  in  a  virtual  environment.  [Online]  Available  from:  http://www.adexa.com/pdf/electronics.pdf.  [Accessed:  8th  June  2014]  Agriculture  Today.  (2010)  Carbon  Assumption  Blown.  [Online]  Available  from:  http://www.dpi.nsw.gov.au/archive/agriculture-­‐today-­‐stories/ag-­‐today-­‐archives/december-­‐2010/carbon-­‐assumption-­‐blown.  [Accessed:  31st  May  2014]  Amburgey,  J.  W.,  and    Thoman,  D.  B.  (2012).  Dimensionality  of  the  New  Ecological  Paradigm  Issues  of  Factor  Structure  and  Measurement.  Environment  and  Behavior,  44(2),  235-­‐256.  Andonova,  L.  (2010).  Public-­‐Private  Partnerships  for  the  Earth:  Politics  and  Patterns  of  Hybrid  Authority  in  the  Multilateral  System.  Global  Environmental  Politics,  10(2),  pp.25-­‐53.  Andrae,  A.  S.  G.  &  Anderson,  O.  (2010)  Life  cycle  assessments  of  consumer  electronics  –  are  they  consistent?  International  Journal  of    Life  Cycle  Assessment.  15  (8).  p.  827–836.  Ardahan,  F.  (2012).  The  new  ecological  paradigm  (NEP  scale)  score  of  Turkish  mountaineers/rock  climbers,  cyclists  and  trekkers.  Energy  Education  Science  and  Technology  Part  B-­‐Social  and  Educational  Studies,  4(4),  2035-­‐2046.  Ashby,  L.  (2011)  Extension's  progress  in  paperless  revolution:  balancing  digital  and  paper.  Journal  of  Extension.  [Online]  49  (1).  Available  from:  http://www.joe.org/joe/2011february/pdf/JOE_v49_1comm2.pdf.  [Accessed:  1st  June  2014]  Athanasiou,  T.  (1996).  The  age  of  greenwashing—.  Capitalism  Nature  Socialism,  7(1),  pp.1-­‐36.  Atlantic  Consulting.  (1998)  LCA  Study  of  the  Product  Group  Personal  Computers  in  the  EU  Ecolabel  Scheme.    London:  Atlantic  Consulting.  Auld,  G.,  Bernstein,  S.  &  Cashore,  B.  (2008a)  The  new  corporate  social  responsibility.  Annual  Review  of  Environment  and  Resources.  33  (1).  p.  413-­‐435.  Auld,  G.,  Gulbrandsen,  L.  &  McDermott,  C.  (2008b)  Certification  schemes  and  the  impacts  on  forests  and  forestry.  Annual  Review  of  Environment  and  Resources.  33  (1).  p.  187-­‐211.  Australian  Government.  (2008)  Carbon  Pollution  Reduction  Scheme  –  Summary  of  Preferred  Positions.  [Online]  Available  from:  http://www.aacb.org.au/files/Greenpaper%20Fed%20Govt%20Summary%202008.pdf.  [Accessed:  14th  February  2009]  AVC.  (2014)  How  big  is  the  tablet  market?  [Online]  Available  from:  http://avc.com/2014/04/how-­‐big-­‐is-­‐the-­‐tablet-­‐market/.  [Accessed:  30th  May  2014]               119Bajpai,  P.  (2011)  Environmentally  friendly  production    of  pulp  and  paper.  [Online]  New  Jersey:  John  Wiley  &  Sons,  Inc.  Available  from:  http://books.google.ca/books?id=EcGmVX8lYGUC&lpg=PT72&dq=environmental%20impacts%20of%20paper%20mills&pg=PT72#v=onepage&q=environmental%20impacts%20of%20paper%20mills&f=false.  [Accessed:  31st  May  2014]  Bare,  J.,  Pennington,  D.  W.,  Udo  de  Haes,  H.  A.  (1994)  Life  cycle  impact  assessment  sophistication.  International  Journal  of  Life  Cycle  Assessment.  4  (5).  p.  299–306.  BBC  News.  (2014)  Cloud  business  may  be  more  efficient  but  is  it  greener?  [Online]  Available  from:  http://www.bbc.com/news/business-­‐27405810.  [Accessed:  28th  May  2014]  Bell,  C.,  Fausset,  C.,  Farmer,  S.,  Nguyen,  J.,  Harley,  L.,  &  Fain,  W.  B.  (2013).  Examining  social  media  use  among  older  adults.  In  Proceedings  of  the  24th  ACM  Conference  on  Hypertext  and  Social  Media  (158-­‐163).  Benjaafar,  S.,  Li,  Y.  and  Daskin,  M.  (2013).  Carbon  Footprint  and  the  Management  of  Supply  Chains:  Insights  From  Simple  Models.  IEEE  Trans.  Automat.  Sci.  Eng.,  10(1),  pp.99-­‐116.  Berg,  S.  &  Lindholm,  E.  (2005)  Energy  use  and  environmental  impacts  of  forest  operations  in  Sweden.  Journal  of  Cleaner  Production.  13  (3).  p.  33-­‐42.    Berkhout,  F.  &  Hertin,  J.  (2004)  De-­‐materialising  and  re-­‐materialising:  digital  technologies  and  the  environment.  Futures.  36  (8).  p.  903-­‐920.  Bernstein,  S.  and  Cashore,  B.  (2007).  Can  non-­‐state  global  governance  be  legitimate?  An  analytical  framework.  Regulation  &  Governance,  1(4),  pp.347-­‐371.  Björklund,  A.  E.  (2002)  Survey  of  approaches  to  improve  reliability  in  LCA.  International  Journal  of  Life  Cycle  Assessment.  7  (2).  p.  64–72.  Bohn,  R.  E.  &  Short,  J.  E.  (2009)  How  Much  Information?  2009  Report  on  American  Consumers.  San  Diego:  Global  Information  Industry  Center.  Boltz,  F.,  Holmes,  T.  P.  &  Carter,  D.  R.  (2003)  Economic  and  environmental  impacts  of  conventional  and  reduced-­‐impact  logging  in  tropical  South  America:  a  comparative  review.  Forest  Policy  and  Economics.  5  (1).  p.  69-­‐81.    Bondy,  K.,  Matten,  D.  and  Moon,  J.  (2004).  The  Adoption  of  Voluntary  Codes  of  Conduct  in  MNCs:  A  Three-­‐Country  Comparative  Study.  Business  &  Society  Review,  109(4),  pp.449-­‐477.                 120Borggren,  C.,  Mohberg,  A.  &  Finnveden,  G.  (2011)  Books  from  an  environmental  perspective  -­‐  part  1:  environmental  impacts  of  paper  books  sold  in  traditional  and  internet  bookshops.  The  International  Journal  of  Life  Cycle  Assessment.  16  (2).  p.  138  Brown,  H.  S.,  de  Jong,  M.  &  Lessidrenska,  T.  (2009)  The  rise  of  the  Global  Reporting  Initiative:  a  case  of  institutional  entrepreneurship.  Environmental  Politics.  18  (2).  p.  182-­‐200.  Bumpus,  A.  (2014).  Firm  responses  to  a  carbon  price:  corporate  decision  making  under  British  Columbia's  carbon  tax.  Climate  Policy,  pp.1-­‐19.  Business  Insider.  (2012)  The  future  of  digital.  [Online]  Available  from:  http://www.businessinsider.com/future-­‐of-­‐digital-­‐slides-­‐2012-­‐11?op=1.  [Accessed:  30th  May  2014]  Canada  Revenue  Agency.  (2014)  Classes  of  depreciable  property.  [Online]  Available  from:  http://www.cra-­‐arc.gc.ca/tx/bsnss/tpcs/slprtnr/rprtng/cptl/dprcbl-­‐eng.html#Class101.  [Accessed:  May  30th  2014]  Canals,  L.  M.  et  al.  (2007)  Key  elements  in  a  framework  for  land  use  impact  assessment  within  LCA.  International  Journal  of  Life  Cycle  Assessment.  12  (1).  p.  5-­‐15.  Cashore,  B.,  Egan,  E.,  Auld,  G.  and  Newsom,  D.  (2007).  Revising  Theories  of  Nonstate  Market-­‐Driven  (NSMD)  Governance:  Lessons  from  the  Finnish  Forest  Certification  Experience.  Global  Environmental  Politics,  7(1),  pp.1-­‐44.  Cashore,  B.,  van  Kooten,  G.,  Vertinsky,  I.,  Auld,  G.  and  Affolderbach,  J.  (2005).  Private  or  self-­‐regulation?  A  comparative  study  of  forest  certification  choices  in  Canada,  the  United  States  and  Germany.  Forest  Policy  and  Economics,  7(1),  pp.53-­‐69.  Carbon  Disclosure  Project.  (2008)  Carbon  Disclosure  Project  Report  2008  Global  500.  [Online]  Available  from:  https://www.cdproject.net/CDPResults/67_329_143_CDP%20Global%20500%20Report%202008.pdf.  [Accessed:  20th  October  2009]  Carbon  Disclosure  Project.  (2010)  Carbon  Disclosure  Project  2010  Global  500  Report.  [Online]  Available  from:  https://www.cdproject.net/CDPResults/CDP-­‐2010-­‐G500.pdf.  [Accessed:  21st  April  2011]  Carlberg,  G.  E.  &  Stuthridge,  T.  R.  (1996)  Environmental  fate  and  distribution  of  substances.  In:  Servos,  M.R.,  Munkittrick,  K.R.,  Carey,  J.H.  &  Kraak,  G.J.V.D  (eds).  Environmental  fate  and  effects  of  pulp  and  paper  mill  effluents.  Delray  Beach,  Florida:  St  Lucie  Press.    CBC  News.  (2013a)  Apple's  Chinese  factories  improved  working  conditions.  [Online]  Available  from:  http://www.cbc.ca/news/business/apple-­‐s-­‐chinese-­‐factories-­‐improved-­‐working-­‐conditions-­‐1.2461718.  [Accessed:  27th  May  2014]  CBC  News.  (2013b)  Planned  obsolescence  pervasive  in  digital  age.  [Online]  Available  from:  http://www.cbc.ca/news/technology/planned-­‐obsolescence-­‐pervasive-­‐in-­‐digital-­‐age-­‐1.2479281.  [Accessed:  30th  May  2014]               121Chan-­‐Olmsted,  S.,  Rim,  H.,  and  Zerba,  A.  (2013).  Mobile  News  Adoption  among  Young  Adults  Examining  the  Roles  of  Perceptions,  News  Consumption,  and  Media  Usage.  Journalism  &  Mass  Communication  Quarterly,  90(1),  126-­‐147.  Choi,  B.,  Shin,  H.  S.,  Lee,  S.  Y.,  Hur,  T.  (2006)  Life  Cycle  Assessment  of  a  Personal  Computer  and  its  Effective  Recycling  Rate.  International  Journal  of  Life  Cycle  Assessment.  11  (2).  p.  122-­‐128.  Christiaanse,  E.  and  Kumar,  K.  (2000).  ICT-­‐enabled  coordination  of  dynamic  supply  webs.  International  Journal  of  Physical  Distribution  &  Logistics  Management,  [online]  30(3),  pp.268-­‐285.  Available  at:  http://dx.doi.org/10.1108/09600030010326019  [Accessed  3  Jul.  2014].  Chyi,  H.  I.,  and  Lee,  A.  M.  (2013).  Online  News  Consumption:  A  structural  model  linking  preference,  use,  and  paying  intent.  Digital  Journalism,  1(2),  194-­‐211.  Cisco.  (2014)  Cisco  visual  networking  index:  global  mobile  data  traffic  forecast  update,  2013-­‐2018.  [Online]  Available  from:  http://www.cisco.com/c/en/us/solutions/collateral/service-­‐provider/visual-­‐networking-­‐index-­‐vni/white_paper_c11-­‐520862.pdf.  [Accessed:  28th  May  2014]  Commission  of  the  European  Communities  (2003).  Forest  Law  Enforcement,  Governance  and  Trade  (FLEGT):  Proposal  for  an  EU  Action  Plan.  Brussels:  European  Parliament.  1-­‐26.  Corraliza,  J.  A.,  Collado,  S.,  and  Bethelmy,  L.  (2013).  Spanish  Version  of  the  New  Ecological  Paradigm  Scale  for  Children.  The  Spanish  Journal  of  Psychology,  16(27).  Costanza,  R.  &  Mageau,  M.  (1999)  What  is  a  healthy  ecosystem?.  Aquatic  Ecology.  33  (1).  p.  105–115.  Crals,  E.  and  Vereeck,  L.  (2005).  The  affordability  of  sustainable  entrepreneurship  certification  for  SMEs.  International  Journal  of  Sustainable  Development  &  World  Ecology,  12(2),  pp.173-­‐183.  Creswell,  J.  W.  (1998)  Qualitative  inquiry  and  research  design:  Choosing  among  five  traditions.  Thousand  Oaks,  CA:  SAGE  publications,  Inc.  Dahlsrud,  A.  (2008).  How  corporate  social  responsibility  is  defined:  an  analysis  of  37  definitions.  Corporate  Social  Responsibility  and  Environmental  Management,  15(1),  pp.1-­‐13.  Dauvergne,  P.  and  Lister,  J.  (2012).  Big  brand  sustainability:  Governance  prospects  and  environmental  limits.  Global  Environmental  Change,  22(1),  pp.36-­‐45.  Deetman,  S.  &  Odegard,  I.  (2009)  Scanning  Life  Cycle  Assessment  of  Printed  and  E-­‐paper  Documents  based  on  the  iRex  Digital  Reader.  [Online]  Available  from:  http://media.leidenuniv.nl/legacy/irex-­‐dr1000-­‐lca-­‐scan-­‐final-­‐report.pdf.  [Accessed:  9th  June  2014]  Dexter,  L.  A.  (1970)  Elite  and  Specialized  Interviewing.  Evanston,  IL:  Northwestern  University  Press.                   122Diamantopoulos,  A.,  Schlegelmilch,  B.,  Sinkovics,  R.  and  Bohlen,  G.  (2003).  Can  socio-­‐demographics  still  play  a  role  in  profiling  green  consumers?  A  review  of  the  evidence  and  an  empirical  investigation.  Journal  of  Business  Research,  56(6),  pp.465-­‐480.  Doh,  J.  and  Guay,  T.  (2004).  Globalization  and  Corporate  Social  Responsibility:  How  Non-­‐Governmental  Organizations  Influence  Labor  and  Environmental  Codes  of  Conduct.  Management  and  International  Review,  pp.7-­‐29.  Duan,  H.,  Eugster,  M.,  Hischier,  R.,  Streciher-­‐Porte,  M.  &  Li,  J.  (2009)  Life  cycle  assessment  study  of  a  Chinese  desktop  personal  computer.  Science  of  the  Total  Environment.  407  (5).  p.  1755-­‐1764.  Dubash  M.  (2005)  Moore’s  Law  is  dead,  says  Gordon  More.  [Online]Available  from:  http://news.techworld.com/operating-­‐systems/3477/moores-­‐law-­‐is-­‐dead-­‐says-­‐gordon-­‐moore/  Information.    [Accessed:  2nd  October  2012]  Duggan,  M.,  and    Brenner,  J.  (2013).  The  demographics  of  social  media  users,  2012    Washington,  DC:  Pew  Research  Center's  Internet  and  American  Life  Project.  Duggan,  M.,  and  Smith,  A.  (2013).  Social  media  update  2013.  Washington,  DC:  Pew  Research  Center's  Internet  and  American  Life  Project.  Durucan,  S.,  Korre,  A.  &  Munoz-­‐Melendez,  G.  (2006)  Mining  life  cycle  modelling:  a  cradle-­‐to-­‐gate  approach  to  environmental  management  in  the  minerals  industry.  Journal  of  Cleaner  Production.  14(12-­‐13).  p.  1057-­‐1070.  Driedger,  R.  (2001).  From  Cradle  to  Grave:  Extended  Producer  Responsibility  for  Household  Hazardous  Wastes  in  British  Columbia.  Journal  of  Industrial  Ecology,  5(2),  pp.89-­‐102.  D'Souza,  C.,  Taghian,  M.  and  Khosla,  R.  (2007).  Examination  of  environmental  beliefs  and  its  impact  on  the  influence  of  price,  quality  and  demographic  characteristics  with  respect  to  green  purchase  intention.  Journal  of  Targeting,  Measurement  and  Analysis  for  Marketing,  15(2),  pp.69-­‐78.  EDF.  (2011)  Why  I  Work  with  Walmart.  Environmental  Defense  Fund.  [Online]  Available  from:  http://business.edf.org/projects/walmart  [Accessed:  25th  May  2011]  Edmonds,  R.  (2013)  Newspaper:  by  the  numbers.  [Online]  Available  from:  http://stateofthemedia.org/2013/newspapers-­‐stabilizing-­‐but-­‐still-­‐threatened/newspapers-­‐by-­‐the-­‐numbers/.  [Accessed:  1st  June  2014]  Edwards,  J.  B.,  Mckinnon,  A.  C.  &  Cullinane,  S.  L.  (2010)  Comparative  analysis  of  the  carbon  footprints  of  conventional  and  online  retailing:  A  “  last  mile  ”  perspective.  International  Journal  of  Physical  Distribution  &  logistics  Management.  40  (1-­‐2).  p.  103-­‐123.  Ehrenfeld,  J.  &  Gertler,  N.  (1997)  Industrial  ecology  in  practice:  The  evolution  of  interdependence  at  Kalundborg.  Journal  of  Industrial  Ecology.  1  (1).  p.  67-­‐79.               123Engelhaupt,  E.  (2008)  Do  food  miles  matter?.  Environmental  Science  &  Technology.  42  (10).  p.  3482-­‐3906.  Enroth,  M.  (2009)  Environmental  impact  of  printed  and  electronic  teaching  aids,  a  screening  study  focusing  on  fossil  carbon  dioxide  emissions.  Advance  in  Printing  and  Media  Technology.36.  p.  1-­‐9.  EPA  Office  of  Solid  Waste.  (2008)  Electronics  Waste  Management  in  the  United  States.  [Online]  Available  from:  http://www.epa.gov/osw/conserve/materials/ecycling/docs/app-­‐1.pdf.  [Accessed:  2nd  October  2012]  EPA.  (1996)  EPA  1994  Toxics  Release  Inventory  Public  Data  Release  Appendix  A:  Questions  and  Answers  XIII  33/50  Program.  [Online]  Available  from:  http://www.mapcruzin.com/scruztri/docs/qa_3350.htm  [Accessed:  21  April  2011]  EPN.  (2013)  I  are  not  anti-­‐paper,  I  are  anti-­‐low-­‐utility-­‐paper.  [Online]  Available  from:  http://www.environmentalpaper.eu/2013/03/27/we-­‐are-­‐not-­‐anti-­‐paper-­‐we-­‐are-­‐anti-­‐low-­‐utility-­‐paper/  [Accessed  14  June  2014]  Epstein-­‐Reeves,  J.  (2011)  The  Do  Well  Do  Good  Public  Opinion  Survey  on  Corporate  Social  Responsibility.  [Online]  Available  from:  http://dowelldogood.net/wp-­‐content/uploads/2011/03/DWDG_CSR_Final.pdf  [Accessed:  21st  April  2011]  Erkman,  S.  (1997)  Industrial  ecology:  an  historical  view.  Journal  of  Cleaner  Production.  5  (1-­‐2).  p.  1-­‐10.  Europa.  (2005)  Questions  &  Answers  on  Emissions  Trading  and  National  Allocation  Plans.  [Online]  Available  from:  http://europa.eu/rapid/pressReleasesAction.do?reference=MEMO/05/84&format=HTML&aged=1&language=EN&guiLanguage=en.  [Accessed:  25  May  2011]  European  Commission.  (2012)  Waste  Electrical  and  Electronic  Equipment  Directive.  [Online]  Available  from:  http://ec.europa.eu/environment/waste/weee/index_en.htm  [Accessed:  2nd  October  2012]  Fair  Labor  Association.  (2012)  Foxconn  Investigation  Report.  [Online]  Available  from:  http://www.fairlabor.org/report/foxconn-­‐investigation-­‐report.  [Accessed:  2nd  October  2012]  Ferguson,  C.  (2009)  Technology  left  behind  -­‐-­‐  the  demise  of  print  newspapers.  Against  the  Grain.  [Online]  21  (3).  Available  from:  http://docs.lib.purdue.edu/cgi/viewcontent.cgi?article=2340&context=atg.  [Accessed:  1st  June  2014]  Finnveden,  G.  (2000)  On  the  Limitations  of  Life  Cycle  Assessment  and  Environmental  Systems  Analysis  Tools  in  General.  International  Journal  of  Life  Cycle  Assessment.  5  (4).  p.  229-­‐238.  Finnveden,  G.,  Hauschild,  M.,  Ekvall,  T.,  Guinée,  J.,  Heijungs,  R.,  Hellweg,  S.,  Koehler,  A.,  Pennington,  D.  &  Suh  S.  (2009)  Recent  developments  in  Life  Cycle  Assessment.  Journal  of  Environmental  Management.  91  (1).  p.  1-­‐21.               124Flatz,  A.,  Serck-­‐Hanssen,  L.  and  Tucker-­‐Bassin,  E.  (2001).  The  Dow  Jones  Sustainability  Group  Index:  The  first  worldwide  sustainability  index.  Sustainable  Banking:  The  Greening  of  Finance,  pp.222-­‐233.  Forest  Ethics.  (n.d.)  Landmark  Great  Bear  Rainforest  Agreement  –  A  World-­‐Leading  Model.  [Online]  Available  from:  http://wewillnotbesilenced.ca/great-­‐bear-­‐rainforest-­‐successes.  [Accessed  25th  May  2011]  FPAC.  (2010)  Canadian  Forest  Industry  and  Environmental  Groups  Sign  World’s  Largest  Conservation  Agreement.  [Online]  Available  from:  http://www.fpac.ca/index.php/en/press-­‐releases-­‐full/canadian-­‐forest-­‐industry-­‐and-­‐environmental-­‐groups-­‐sign-­‐worlds-­‐largest-­‐/.  [Accessed:  21st  September  2010]  Fraser,  P.,  Moultrie,  J.,  and  Gregory,  M.  (2002).  The  use  of  maturity  models/grids  as  a  tool  in  assessing  product  development  capability.  In  Engineering  Management  Conference,  IEEE  International  2002.  Vol.  1,  pp.  244-­‐249.  Frey,  S.  D.,  Harrison,  D.  J.  &  Billett,  E.  (2006)  Ecological  footprint  analysis  applied  to  mobile  phones.  Journal  of  Industrial  Ecology.  10  (1-­‐2).  p.  199-­‐216.  Friedman,  M.  (1970)  The  Social  Responsibility  of  Business  is  to  Increase  its  Profits.  [Online]  Available  from:  http://www.colorado.edu/studentgroups/libertarians/issues/friedman-­‐soc-­‐resp-­‐business.html  [Accessed:  25th  May  2011]  Gaines,  S.E.  (1991)  Polluter-­‐Pays  Principle:  From  Economic  Equity  to  Environmental  Ethos.  Texas  International  Law  Journal.  26  (463)  Gauthier,  C.  (2005).  Measuring  Corporate  Social  and  Environmental  Performance:  The  Extended  Life-­‐Cycle  Assessment.  Journal  of  Business  Ethics,  59(1-­‐2),  pp.199-­‐206.  Galloway,  B.  J.,  Munkittrick,  K.  R.,  Curry,  A.,  Wood,  C.  &  Dunn,  S.  (2004)  Identifying  a  suitable  fish  species  for  monitoring  multiple  effluents  in  the  upper  saint  john  river,  Canada.  In:  Borton,  D.L.,  Hall,  T.J.,  Fisher,  R.P.  &  Thomas,  J.F.  (eds).  Pulp  &  paper  mill  effluent  environmental  fate  &  effects.  Lancaster,  PA:  DEStech  Publications,  Inc.  Gard,  D.  L.  &  Keoleian,  G.  A.  (2003)  Digital  versus  print.  Energy  performance  in  the  selection  and  use  of  scholarly  journals.  Journal  of  Industrial  Ecology.  6  (2).  p.  115–132.  Garnaut.  (2008)  Garnaut  Climate  Change  Review.  [Online]  Available  from:  http://www.garnautreview.org.au/.  [Accessed:  25th  May  2011]  Gaudreault,  C.,  Samson,  R.  &  Stuard,  P.R.  (2007)  Life-­‐cycle  thinking  in  the  pulp  and  paper  industry,  Part  I:  Current  practices  and  most  promising  avenues.  Tappi  Journal.  6  (7).  p.  25-­‐31.  Gaudreault,  C.,  Samson,  R.  &  Stuard,  P.R.  (2007)  Life-­‐cycle  thinking  in  the  pulp  and  paper  industry,  part  2:  LCA  studies  and  opportunities  for  development.  Tappi  Journal.  6  (8).  p.  3-­‐10.               125German,  K.  (2008)  Motorola  Renew  W233.  [Online]  Available  from:  http://reviews.cnet.com/cell-­‐phones/motorola-­‐renew-­‐w233-­‐t/4505-­‐6454_7-­‐33485041.html.  [Accessed:  25th  May  2011]  Giovannucci,  D.  and  Ponte,  S.  (2005).  Standards  as  a  new  form  of  social  contract?  Sustainability  initiatives  in  the  coffee  industry.  Food  Policy,  30(3),  pp.284-­‐301.  Glaser,  B.  G.  &  Strauss,  A.  L.  (1967)  The  discovery  of  grounded  theory:  Strategies  for  qualitative  research.  New  York,  NY:  Aldine  de  Gruyter.  Goodland,  R.  and  Ledec,  G.  (1987).  Neoclassical  economics  and  principles  of  sustainable  development.  Ecological  Modelling,  38(1-­‐2),  pp.19-­‐46.  Gower,  S.  T.,  McKeon-­‐Ruediger,  A.,  Reitter,  A.,  Bradley,  M.,  Refkin,  D.,  Tollefson,  T.,  Souba  Jr.,  F.,  Taup,  A.,  Embury-­‐Williams,  L.,  Schiavone,  S.,  Weinbaur,  J.,  Janetos,  A.  C.  &  Jarvis,  R.  (2006)  Following  the  Paper  Trail:  The  Impact  of  Magazine  and  Dimensional  Lumber  Production  on  Greenhouse  Gas  Emissions:  A  Case  Study.  Washington  DC:  The  Heinz  Center.  Graedel,  T.  E.  &  Allenby,  B.  R.  (1995)  Industrial  Ecology.  New  Jersey:  Prentice  Hall  PTR.  Graedel,  T.  E.  &  Howard-­‐Grenville,  J.  (2005)  Greening  the  Industrial  Facility.  New  York:  Springer.  Greenpeace,  (2010).  Make  IT  Green:  Cloud  Computing  and  its  Contribution  to  Climate  Change.  [online]  Available  at:  http://www.greenpeace.org/usa/en/media-­‐center/reports/make-­‐it-­‐green-­‐cloud-­‐computing/  [Accessed  22  Nov.  2014].  Grossman,  E.  (2006)  High  Tech  Trash:  Digital  Devices,  Hidden  Toxics,  and  Human  Health.  Washington  DC:  Shearwater.  Hansen,  A.  (2010).  Environment,  media  and  communication.  1st  ed.  London:  Routledge.  Hansla,  A.,  Gamble,  A.,  Juliusson,  A.  and  Gärling,  T.  (2008).  Psychological  determinants  of  attitude  towards  and  willingness  to  pay  for  green  electricity.  Energy  Policy,  36(2),  pp.768-­‐774.  Hartman,  C.  and  Stafford,  E.  (1997).  Green  alliances:  building  new  business  with  environmental  groups.  Long  Range  Planning,  30(2),  pp.184-­‐149.  Hedegaard,  C.  (2008)  The  winding  and  stony  road  towards  Copenhagen.  [Online]  Available  from:  http://en.cop15.dk/about+cop15/information+for/delegates/show+article?articleid=304.  [Accessed:  14th  February  2009]  Helin,  T.,  Sokka,  L.,  Soimakallio,  S.,  Pingoud,  K.  and  Pajula,  T.  (2012).  Approaches  for  inclusion  of  forest  carbon  cycle  in  life  cycle  assessment  -­‐  a  review.  GCB  Bioenergy,  5(5),  pp.475-­‐486.  Hellweg,  S.  (2000)  Time  and  site-­‐dependent  life  cycle  assessment  of  thermal  waste  treatment  processes.  International  Journal  of  Life  Cycle  Assessment.  6  (1).  p.  46.               126Helm,  D.,  Hepburn,  C.  and  Ruta,  G.  (2012).  Trade,  climate  change,  and  the  political  game  theory  of  border  carbon  adjustments.  Oxford  Review  of  Economic  Policy,  28(2),  pp.368-­‐394.  Hetemäki,  L.  and  Nilsson,  S.  (2005).  Information  technology  and  the  forest  sector.  IUFRO  World  Series  18.  Hill,  R.,  Stephens,  D.  and  Smith,  I.  (2003).  Corporate  Social  Responsibility:  An  Examination  of  Individual  Firm  Behavior.  Business  &  Society  Review,  108(3),  pp.339-­‐364.  Hilty,  L.,  Arnfalk,  M.  P.,  Erdmann,  L.,  Goodman,  J.,  Lehmann,  M.,  Wäger,  P.  A.  (2006)  The  relevance  of  information  and  communication  technologies  for  environmental  sustainability  -­‐  A  prospective  simulation  study.  Environmental  Modeling  &  Software.  21  (11).  p.  1618-­‐1629.  Hischier,  R  &  Reichart,  I.  (2003)  Multifunctional  electronic  media  —  traditional  media.  The  problem  of  adequate  functional  unit.  International  Journal  of  Life  Cycle  Assessment.  8  (4).  p.  201–208.  Hischier,  R.,  Wäger,  P.  &  Gauglhofer,  J.  (2005)  Does  WEEE  recycling  make  sense  from  an  environmental  perspective?  The  environmental  impacts  of  the  Swiss  take-­‐back  and  recycling  system  for  waste  electrical  and  electronic  equipment  (WEEE).  Environmental  Impact  Assessment  Review.  25  (5).  p.  525-­‐539.  Holbert,  R.  L.,  Kwak,  N.,  &  Shah,  D.  V.  (2003).  Environmental  concern,  patterns  of  television  viewing,  and  pro-­‐environmental  behaviors:  Integrating  models  of  media  consumption  and  effects.  Journal  of  Broadcasting  &  Electronic  Media,  47(2),  177-­‐196.  Home  Depot.  (2011)  Wood  Purchasing  Policy.  [Online]  Available  from:  http://corporate.homedepot.com/wps/portal/Wood_Purchasing.  [Accessed:  25th  May  2011]  Houpt,  S.  (2010)  Beyond  the  Bottle:  Coke  Trumpets  its  Green  Initiatives.  [Online]  Available  from:  http://www.theglobeandmail.com/report-­‐on-­‐business/industry-­‐news/marketing/adhocracy/beyond-­‐the-­‐bottle-­‐coke-­‐trumpets-­‐its-­‐green-­‐initiatives/article1869437/.  [Accessed:  21st  April  2011]  Howlett,  M.,  Rayner,  J.  and  Tollefson,  C.  (2009).  From  government  to  governance  in  forest  planning?  Lessons  from  the  case  of  the  British  Columbia  Great  Bear  Rainforest  initiative.  Forest  Policy  and  Economics,  11(5-­‐6),  pp.383-­‐391.  Hujala,  M.  (2011).  The  role  of  information  and  communication  technologies  in  paper  consumption.  International  Journal  of  Business  Information  Systems,  7(2),  p.121.  Huijbreghts,  M.  A.,  Schöpp,  W.,  Verkuijlen,  E.,  Heijungs,  R.  &  Reijinders,  L.  (2000)  Spatially  explicit  characterization  of  acidifying  and  eutrophying  air  pollution  in  life-­‐cycle  assessment.  Journal  of  Industrial  Ecology.  4  (3).  p.  75–92.  Humphrey,  W.  S.  (1989).  Managing  the  Software  Process.  Reading,  Mass.:  Addison-­‐Wesley.               127Intergovernmental  Panel  on  Climate  Change.  (2007)  IPCC  Four  Assessment  Report.  [Online]  Available  from:  http://www.ipcc.ch/publications_and_data/publications_and_data_reports.shtm.  [Accessed:  14th  February  2009]  Internet  World  Stats.  (2012)  Internet  Users  in  the  World  Distribution  by  World  Regions  -­‐  2012  Q2.  [Online]  Available  from:  http://www.internetworldstats.com/stats.htm.  [Accessed:  2nd  October  2012]  ISO.  (2006)  ISO  14025.  Environmental  labels  and  declarations  –  Type  III  –  environmental  declarations  –  Principles  and  Procedures.  [Online]  Available  from:  http://www.iso.org/iso/catalogue_detail?csnumber=38131.  [Accessed:  12th  June  2013]  ISO.  (2006)  ISO  14040:  Environmental  management—life  cycle  assessment—principles  and  framework.  [Online]  Available  from:  http://www.iso.org/iso/catalogue_detail?csnumber=37456.  [Accessed:  13th  June  2013]  IUCN:  International  Union  for  Conservation  of  Nature.  (2008)  Developing  Carbon  Markets  for  People  and  Nature    Available  from:  http://cmsdata.iucn.org/downloads/summary_report_1549_1.pdf.  [Accessed:  14th  February  2009]  Jad,  I.  (2007).  NGOs:  between  buzzwords  and  social  movements.  Development  in  Practice,  17(4-­‐5),  pp.622-­‐629.  Jenkins,  H.  and  Yakovleva,  N.  (2006).  Corporate  social  responsibility  in  the  mining  industry:  Exploring  trends  in  social  and  environmental  disclosure.  Journal  of  Cleaner  Production,  14(3-­‐4),  pp.271-­‐284.  Jones,  S.  and  Ratnatunga,  J.  (2012).  Contemporary  issues  in  sustainability  accounting,  assurance  and  reporting.  Bingley:  Emerald  Group  Pub.  Jorgenson,  D.  &  Wessner,  C.  (2004)  Productivity  and  Cyclicality  in  Semiconductors.  Washington:  National  Academy  Press.  Kahhat,  R.,  Kim,  J.,  Xu,  M.,  Allenby,  B.,  Williams,  E.  and  Zhang,  P.  (2008).  Exploring  e-­‐waste  management  systems  in  the  United  States.  Resources,  Conservation  and  Recycling,  52(7),  pp.955-­‐964.  Khanna,  M.  and  Damon,  L.  (1999).  EPA's  Voluntary  33/50  Program:  Impact  on  Toxic  Releases  and  Economic  Performance  of  Firms.  Journal  of  Environmental  Economics  and  Management,  37(1),  pp.1-­‐25.  Khanna,  M.,  Quimio,  W.  and  Bojilova,  D.  (1998).  Toxics  Release  Information:  A  Policy  Tool  for  Environmental  Protection.  Journal  of  Environmental  Economics  and  Management,  36(3),  pp.243-­‐266.  Kharrazi,  A.,  Kraines,  S.,  Hoang,  L.  and  Yarime,  M.  (2014).  Advancing  quantification  methods  of  sustainability:  A  critical  examination  emergy,  exergy,  ecological  footprint,  and  ecological  information-­‐based  approaches.  Ecological  Indicators,  37,  pp.81-­‐89.                 128King,  I.  (2014).  Intel  Stiff-­‐Armed  Pursuing  African-­‐Mine  Conflict  Minerals.  [online]  Bloomberg.  Available  at:  http://www.bloomberg.com/news/2014-­‐05-­‐27/intel-­‐endures-­‐resistance-­‐starving-­‐africa-­‐s-­‐militia-­‐mines.html  [Accessed  3  Jul.  2014].  Kleindorfer,  P.,  Singhal,  K.  &  Van  Wassenhove,  L.  (2005)  Sustainable  Operations  Management.  Production  and  Operations  Management.  14  (4).  p.  482-­‐492.  Kolk,  A.,  Levy,  D.  and  Pinkse,  J.  (2008).  Corporate  Responses  in  an  Emerging  Climate  Regime:  The  Institutionalization  and  Commensuration  of  Carbon  Disclosure.  European  Accounting  Review,  17(4),  pp.719-­‐745.  Kollberg,  M.  (2005).  Beyond  IT  and  Productivity:  Effects  of  Digitized  Information  Flows  in  the  Logging  Industry.  PhD.  Linköping  University,  The  Institute  of  Technology.  Koltun,  P.  and  Tharumarajah,  A.  (2014).  Life  Cycle  Impact  of  Rare  Earth  Elements.  ISRN  Metallurgy,  [online]  2014.  Available  at:  http://www.hindawi.com/journals/isrn.metallurgy/2014/907536/  [Accessed  29  Jul.  2014].  Koomey,  J.  (2008)  Worldwide  electricity  used  in  data  centers.  Environmental  Research  Letters.  [Online]  3  (034008).  Available  from:  http://iopscience.iop.org/1748-­‐9326/3/3/034008/fulltext/.  [Accessed:  28th  May  2014]  Kopnina,  H.  (2012).  ‘People  are  not  plants,  but  both  need  to  grow’:  qualitative  analysis  of  the  new  ecological  paradigm  scale  for  children.  The  Environmentalist,  32(4),  394-­‐404.  Kozak,  G.  L.  &  Keoleian,  G.  A.  (2003)  Printed  Scholarly  Books  and  E-­‐book  Reading  Devices:  A  Comparative  Life  Cycle  Assessment  of  Two  Book  Options.  Electronics  and  the  Environment,  2003.  IEEE  International  Symposium  on.  p.  291-­‐296.    Kraft,  M.  (2001).  Environmental  policy  and  politics.  New  York:  Longman.  Krones,  J.  (2007)  The  Best  of  Both  Worlds:  A  Beginner’s  Guide  to  Industrial  Ecology.  MIT  Undergraduate  Research  Journal.  15.  p.  19-­‐22.  Kuher,  R.  &  Williams,  E.  (eds).  (2003)  Computers  and  the  environment:  understanding  and  managing  their  impacts.  Dordrecht,  Netherlands:  Kluwer  Academic.  Laan,  T.,  Litman  T.A.,  Steenblik,  R.  (2009).  Biofuels–At  what  cost?  Government  support  for  ethanol  and  biodiesel  in  Canada.  Winnipeg:  International  Institute  for  Sustainable  Development.  1-­‐110.  Lau,  J.,  Wong,  C.  P.,  Lee,  N.  &  Lee,  R.  (2002)  Electronics  Manufacturing:  with  lead-­‐free,  halogen-­‐free,  and  conductive-­‐adhesive  materials.  [Online]  New  York:  McGraw-­‐Hill  Companies,  Inc.  Available  from:  http://dl.acm.org/citation.cfm?id=1593670.  [Accessed:  31st  May  2014]                 129Lee,  K.  (2011).  Integrating  carbon  footprint  into  supply  chain  management:  the  case  of  Hyundai  Motor  Company  (HMC)  in  the  automobile  industry.  Journal  of  Cleaner  Production,  19(11),  pp.1216-­‐1223.  Lenzen,  M.  (2000)  Errors  in  conventional  and  input–output-­‐based  life-­‐  cycle  inventories.  Journal  of  Industrial  Ecology.  4  (4).  p.  127–148.  Levan,  S.  (1995)  Life  Cycle  Assessment :  Measuring  Environmental  Impact.  Madison,  WI:  Forest  Products  Society.    Liere,  K.D.  &  Dunlap,  R.E.  (1980).  The  Social  Bases  of  Environmental  Concern:  A  Review  of  Hypotheses,  Explanations  and  Empirical  Evidence.  Public  Opinion  Quarterly,  p.181-­‐197.  Liu,  C.  H.,  Lin,  S.  J.  &  Lewis,  C.  (2010)  Life  cycle  assessment  of  DRAM  in  Taiwan’s  semiconductor  industry.  Journal  of  Cleaner  Production.  18  (5).  p.  419-­‐425.    Liu,  G.  and  Han,  S.  (2009).  Long-­‐term  forest  management  and  timely  transfer  of  carbon  into  wood  products  help  reduce  atmospheric  carbon.  Ecological  Modelling,  220(13-­‐14),  pp.1719-­‐1723.  Lohmann,  L.  (2009).  Toward  a  different  debate  in  environmental  accounting:  The  cases  of  carbon  and  cost–benefit.  Accounting,  Organizations  and  Society,  34(3-­‐4),  pp.499-­‐534.  Lozano,  R.  (2008).  Envisioning  sustainability  three-­‐dimensionally.  Journal  of  Cleaner  Production,  16(17),  pp.1838-­‐1846.  Lu,  L.  T.,  Wernick,  I.  K.,  Hsiao,  T.  Y.,  Yu,  Y.  H.,  Yang,  Y.  M.  &  Ma,  H.  W.  Balancing  the  life  cycle  impacts  of  notebook  computers:  Taiwan's  experience.  Resources,  Conservation  and  Recycling.  48  (1).  p.  13-­‐25.  Lund-­‐Thomsen,  P.  and  Wad,  P.  (2014).  Global  Value  Chains,  Local  Economic  Organization  and  Corporate  Social  Responsibility  in  the  BRICS  Countries.  Competition  &  Change,  18(4),  pp.281-­‐290.  Luo,  X.  and  Bhattacharya,  C.  (2006).  Corporate  Social  Responsibility,  Customer  Satisfaction,  and  Market  Value.  Journal  of  Marketing,  70(4),  pp.1-­‐18.  MacKerron,  G.,  Egerton,  C.,  Gaskell,  C.,  Parpia,  A.  and  Mourato,  S.  (2009).  Willingness  to  pay  for  carbon  offset  certification  and  co-­‐benefits  among  (high-­‐)flying  young  adults  in  the  UK.  Energy  Policy,  37(4),  pp.1372-­‐1381.  Malmodin,  J.,  Moberg,  Å.,  Lundén,  D.,  Finnveden,  G.  &  Lövehagen,  N.  (2010)  Greenhouse  gas  emissions  and  operational  electricity  use  in  the  ICT  and  Entertainment  and  Media  sectors.  Journal  of  Industrial  Ecology.  14  (5).  p.  770-­‐790.  Manget,  J.,  Roche,  C.  &  Münnich,  F.  (2009)  Capturing  the  Green  Advantage  for  Consumer  Companies.  [Online]  Available  from:  http://www.bcg.com/documents/file15407.pdf.  [Accessed:  20th  September  2010]               130Mariadoss,  B.,  Tansuhaj,  P.  and  Mouri,  N.  (2011).  Marketing  capabilities  and  innovation-­‐based  strategies  for  environmental  sustainability:  An  exploratory  investigation  of  B2B  firms.  Industrial  Marketing  Management,  40(8),  pp.1305-­‐1318.  Marland,  G.  and  Schlamadinger,  B.  (1997).  Forests  for  carbon  sequestration  or  fossil  fuel  substitution?  A  sensitivity  analysis.  Biomass  and  Bioenergy,  13(6),  pp.389-­‐397.  McWilliams,  A.  and  Siegel,  D.  (2001).  Corporate  Social  Responsibility:  A  Theory  of  the  Firm  Perspective.  Academy  of  Management  Review,  26(1),  pp.117-­‐127.  Mebratu,  D.  (1998)  Sustainability  and  sustainable  development:  Historical  and  conceptual  review.  Environmental  Impact  Assessment  Review.  18  (6).  p.  493-­‐520.  Media  Life  Magazine.  (2014)  Behind  magazines'  decline  on  the  newsstand.  [Online]  Available  from:  http://www.medialifemagazine.com/behind-­‐magazines-­‐decline-­‐newsstand/.  [Accessed:  1st  June  2014]  Melnyk,  S.  (2003).  Assessing  the  impact  of  environmental  management  systems  on  corporate  and  environmental  performance.  Journal  of  Operations  Management,  21(3),  pp.329-­‐351.  Micales,  J.  A.  &  Skog,  K.  E.  (1997)  The  decomposition  of  forest  products  in  landfills.  International  Biodeterioration  &  Biodegradation.  39  (2-­‐3).  p.  145-­‐158.    Michaels,  D.  (2014).  Few  Firms  Name  Sources  in  U.S.  Conflict  Minerals  Reports.  [Online]  Bloomberg.  Available  at:  http://www.bloomberg.com/news/2014-­‐06-­‐02/few-­‐firms-­‐name-­‐sources-­‐in-­‐u-­‐s-­‐conflict-­‐minerals-­‐reports.html  [Accessed  3  Jul.  2014].  Mikkilä,  M.  and  Toppinen,  A.  (2008).  Corporate  responsibility  reporting  by  large  pulp  and  paper  companies.  Forest  Policy  and  Economics,  10(7-­‐8),  pp.500-­‐506.  MIT  Technology  Review.  (2010)  The  mobile  device  is  becoming  humankind's  primary  tool  (infographics  feature).  [Online]  Available  from:  http://www.technologyreview.com/news/421826/the-­‐mobile-­‐device-­‐is-­‐becoming-­‐humankinds-­‐primary-­‐tool-­‐infographics-­‐feature/.  [Accessed:  30th  May  2014]  Moberg,  Å.,  Borggren,  C.  &  Finnveden,  G.  (2011)  Books  from  an  environmental  perspective-­‐Part  2:  e-­‐books  as  an  alternative  to  paper  books.  International  Journal  of  Life  Cycle  Assessment.  16  (3).  p.  238-­‐246.  Moberg,  Å.,  Johansson,  M.,  Finnveden,  G.  &  Jonsson,  A.  (2010)  Printed  and  tablet  e-­‐paper  newspaper  from  an  environmental  perspective  —  A  screening  life  cycle  assessment.  Environmental  Impact    Assessment  Review.  30  (3).  p.  177-­‐191.    Mokhtarian,  P.  L.,  Handy,  S.  L.  &  Salomon,  I.  (1995)  Methodological  issues  in  the  estimation  of  the  travel,  energy,  and  air  quality  impacts  of  telecommuting.  Transportation  Research  part  A.    29  (4).  p.  283-­‐302.               131Molina-­‐Azorín,  J.F.,  Claver-­‐Cortés,  E.,  López-­‐Gamero,  M.D.  &  Tari,  J.J.  (2009)  Green  management  and  financial  performance:  a  literature  review.  Management  Decision.  47  (7).  p.  1080-­‐1100.  Moon,  W.,  Florkowski,  W.,  Bruckner,  B.  and  Schonhof,  I.  (2002).  Willingness  to  Pay  for  Environmental  Practices:  Implications  for  Eco-­‐Labeling.  Land  Economics,  78(1),  p.88.  Moore,  D.  (2014).  Pacific  Carbon  Trust  Chopped.  [online]  Available  at:  http://www.huffingtonpost.ca/2013/11/19/pacific-­‐carbon-­‐trust_n_4303711.html  [Accessed  29  Jul.  2014].  Motorola.  (2014).  Conflict  Minerals.  [Online]  Available:  http://responsibility.motorola.com/index.php/suppliers/conflictminerals/  [Accessed:  2nd  July  2014.  Naturally:Wood.  (2013)  Environmental  Product  Declarations.  [Online]  Available  from:  http://www.naturallywood.com/environmental-­‐product-­‐declarations  [Accessed:  12th  June  2013]  Neil  Adger,  W.,  Arnell,  N.  and  Tompkins,  E.  (2005).  Successful  adaptation  to  climate  change  across  scales.  Global  Environmental  Change,  15(2),  pp.77-­‐86.  Newspaper  Association  of  America.  (2013)  The  American  newspaper  media  industry  revenue  profile  2012.  [Online]  Available  from:  http://www.naa.org/Trends-­‐and-­‐Numbers/Newspaper-­‐Revenue/Newspaper-­‐Media-­‐Industry-­‐Revenue-­‐Profile-­‐2012.aspx.  [Accessed:  1st  June  2014]  Ni,  H.  and  Zeng,  E.  (2009).  Law  Enforcement  and  Global  Collaboration  are  the  Keys  to  Containing  E-­‐Waste  Tsunami  in  China.  Environmental  Science  &  Technology,  43(11),  pp.3991-­‐3994.  Nogueron,  R.  &  Hanson,  C.  (2010)  Risk  Free?  Paper  and  the  Lacey  Act.  [Online]  Available  from:  http://www.wri.org/stories/2010/11/risk-­‐free-­‐paper-­‐and-­‐lacey-­‐act  [Accessed:  21st  April  2011]  NRTEE.  (2012)  Canada’s  Opportunity:  Adopting  Life  Cycle  Approaches  for  Sustainable  Development.  [Online]  Available  from:  http://www.stakeholderforum.org/fileadmin/files/NRT-­‐Life-­‐Cycle-­‐Approaches-­‐report-­‐eng.pdf.  [Accessed:  13th  June  2014]  Obar,  J.,  Zube,  P.  and  Lampe,  C.  (2012).  Advocacy  2.0:  An  Analysis  of  How  Advocacy  Groups  in  the  United  States  Perceive  and  Use  Social  Media  as  Tools  for  Facilitating  Civic  Engagement  and  Collective  Action.  Journal  of  Information  Policy,  2,  pp.1-­‐25.  Ochoa,  C.,  and  Keenan,  P.  J.  (2011).  Regulating  information  flows,  regulating  conflict:  An  analysis  of  United  States  conflict  minerals  legislation.  Goettingen  Journal  of  International  Law.  3(129)  Ongondo,  F.O.,  Williams,  I.  D.  &  Cherrett,  T.  J.  (2011)  How  are  WEEE  doing?  A  global  review  of  the  management  of  electrical  and  electronic  wastes.  Waste  Management.  [Online]  31  (4).  p.  714-­‐730.  Available  from:  http://www.sciencedirect.com.  [Accessed:  30th  May  2014]               132OpenCongress.  (2010)  H.R.4128  –  Conflict  Minerals  Trade  Act.  OpenCongress.  [Online]  Available  from:  http://www.opencongress.org/bill/111-­‐h4128/show.  [Accessed:  21st  September  2010]  Ottman,  J.  (2011)  The  New  Rules  of  Green  Marketing:  Strategies,  Tools  and  Inspiration  for  Sustainable  Branding.  San  Francisco:  Berret-­‐Koehler  Publisher.  Overdevest,  C.  (2010)  Comparing  forest  certification  schemes:  the  case  of  ratcheting  standards  in  the  forest  sector.  Socio-­‐Economic  Review.  [Online]  8  (1).  p.  47-­‐76.  Available  from:  http://ser.oxfordjournals.org/content/8/1/47.short.  [Accessed:  30th  May  2014]  Paija,  L.  (2000).  Industrial  network  relationships  in  the  Finnish  ICT  cluster.  The  Research  Institute  of  the  Finnish  Economy.  [Online]  Available  from:  http://www1.oecd.org/science/inno/2098542.pdf  [Accessed:  2nd  July  2014]  Parguel,  B.,  Benoit-­‐Moreau,  F.  and  Larceneux,  F.  (2011).  How  Sustainability  Ratings  Might  Deter  “Greenwashing”:  A  Closer  Look  at  Ethical  Corporate  Communication.  Journal  of  Business  Ethics,  102(1),  pp.15-­‐28.  Paul,  J.  &  Campbell,  G.  (2011)  Investigating  rare  earth  element  mine  development  in  epa  region  8  and  potential  environmental  impacts.  Environmental  Protection  Agency.  [Online]  Available  from:  http://nepis.epa.gov/Exe/ZyPDF.cgi/P100FHSY.PDF?Dockey=P100FHSY.PDF.  [Accessed:  25th  May  2014]  Paulk,  M.  (1993).  Comparing  ISO  9001  and  the  Capability  Maturity  Model  for  Software.  Software  Quality  Journal.,  [Online]  2(4),  pp.245-­‐256.  Available  at:  http://dx.doi.org/10.1007/bf00403767  [Accessed  5  Jul.  2014].  Paulk,  M.  (1995).  The  capability  maturity  model.  1st  ed.  Reading,  Mass.:  Addison-­‐Wesley  Pub.  Co.  Pew  Research  Journalism  Project.  (2013)  Newsweek  by  the  numbers.  [Online]  Available  from:  http://www.journalism.org/2013/06/03/newsweek-­‐numbers/.  [Accessed:  1st  June  2014]  Plepys,  A.  (2004)  The  environmental  impacts  of  electronics.  Going  beyond  the  walls  of  semiconductor  fabs.  Institute  of  Electrical  and  Electronics  Engineers.  p.  159-­‐165.  Potting,  J.,  Schöpp,  W.,  Blok,  K.  &  Hauschild,  M.  (1998)  Site-­‐dependent  life-­‐cycle  impact  assessment  of  acidification.  Journal  of  Industrial  Ecology.  2  (2).  p.  63–87.  Puckett,  J.,  Byster,  L.,  Westervelt,  S.,  Gutierrez,  R.,  Davis,  S.,  Hussain,  A.  &    Dutta,  M.  (2002)  Exporting  Harm  The  High-­‐Tech  Trashing  of  Asia.  [Online]  Available  from:  http://www.ban.org/E-­‐waste/technotrashfinalcomp.pdf.  [Accessed:  13th  June  2014]  Rametsteiner,  E.  and  Simula,  M.  (2003).  Forest  certification,  an  instrument  to  promote  sustainable  forest  management?.  Journal  of  Environmental  Management,  67(1),  pp.87-­‐98.               133Rapport,  D.,  Bohm,  G.,  Buckingham,  D.,  Cairns,  J.,  Costanza,  R.,  Karr,  J.,  de  Kruijf,  H.,  Levins,  R.,  McMichael,  A.,  Nielsen,  N.  and  Whitford,  W.  (1999).  Ecosystem  Health:  The  Concept,  the  ISEH,  and  the  Important  Tasks  Ahead.  Ecosyst  Health,  5(2),  pp.82-­‐90.  Reap,  J.,  Roman,  F.,  Duncan,  S.  &  Bras,  B.  (2008a)  A  survey  of  unresolved  problems  in  life  cycle  assessment  part  1:  goal  and  scope  and  inventory  analysis.  International  Journal  of  Life  Cycle  Assessment.  13  (4).  p.  290–300.    Reap,  J.,  Roman,  F.,  Duncan,  S.  &  Bras,  B.  (2008b)  A  survey  of  unresolved  problems  in  life  cycle  assessment  part  2:  impact  assessment  and  interpretation.  The  International  Journal  of  Life  Cycle  Assessment.  13  (5).  p.  374–388.  Regional  Greenhouse  Gas  Initiative  (RGGI).  (2009)  Regional  Greenhouse  Gas  Initiative:  Home.  [Online]  Available  from:  http://www.rggi.org/home.  [Accessed:  25th  May  2011]  Robinson,  B.  H.  (2009)  E-­‐waste:  An  assessment  of  global  production  and  environmental  impacts.  Science  of  The  Total  Environment.  408  (2).  p.  183-­‐191.  Rupp,  K.  &  Selberherr,  S.  (2011)  The  Economic  Limit  to  Moore’s  Law.  IEEE  Transactions  on  Semiconductor  Manufacturing.  [Online]  24  (1).  p.  1-­‐4.  Available  from:  http://www.iue.tuwien.ac.at/pdf/ib_2010/JB2011_Rupp_1.pdf.  [Accessed:  2nd  October  2010]  Russel,  A.,  Ekvall,  T.  &  Baumann,  H.  (2005)  Life  cycle  assessment—  introduction  and  overview.  Journal  of  Cleaner  Production.  13  (13-­‐14).  p.  1207–1210.  Sandborn,  P.  (2008)  Trapped  on  Technology’s  Trailing  Edge.  [Online]  Available  from:  http://spectrum.ieee.org/computing/hardware/trapped-­‐on-­‐technologys-­‐trailing-­‐edge/1.  [Accessed:  2nd  October  2012]  Scharnhorst,  W.  (2006)  Life  Cycle  Assessment  of  Mobile  Telephone  Networks,  with  Focus  on  the  End-­‐of-­‐Life  Phase.  International  Journal  of  Life  Cycle  Assessment.  11  (4).  p.  290-­‐291.  Schnoor,  J.  (2014).  Responding  to  Climate  Change  with  a  Carbon  Tax.  Environmental  Science  &  Technology,  48(21),  pp.12475-­‐12476.  Schumpeter,  J.  (1942)  Capitalism,  Socialism  and  Democracy.  New  York:  Harper.  Seager,  T.  P.  (2008)  The  Sustainability  Spectrum  and  the  Sciences  of  Sustainability.  Business  Strategy  and  the  Environment.  17  (7).  p.  444-­‐453.  SEC  (U.S.  Securities  and  Exchange  Commission).  (2012).  SEC  Adopts  Rule  for  Disclosing  Use  of  Conflict  Minerals.  Available:  http://www.sec.gov/News/PressRelease/Detail/PressRelease/1365171484002.  Last  accessed  2nd  July  2014.  Sellen,  A.  J.,  Harper,  R.  H.  R.  (2001)  The  Myth  of  the  Paperless  Office.  Cambridge,  MA:  MIT  Press.               134Semenza,  J.,  Hall,  D.,  Wilson,  D.,  Bontempo,  B.,  Sailor,  D.  and  George,  L.  (2008).  Public  Perception  of  Climate  Change.  American  Journal  of  Preventive  Medicine,  35(5),  pp.479-­‐487.  Seuring,  S.  &  Muller,  M.  (2008)  From  a  literature  review  to  a  conceptual  framework  for  sustainable  supply  chain  management.  Journal  of  Cleaner  Production.  16  (15).  p.  1699-­‐1710.  Singhal,  P.  (2005)  Integrated  Product  Policy  Pilot  Project:  Life  Cycle  Environmental  Issues  of  Mobile  Phones.  [Online]  Available  from:  http://ec.europa.eu/environment/ipp/pdf/nokia_mobile_05_04.pdf.  [Accessed:  13th  June  2014]  Skog,  K.  E.  &  Nicholson,  G.  A.  (2000)  Carbon  sequestration  in  wood  and  paper  products.  In:    Joyce,  L.A.  &  Birdsey,  R.  (eds).  The  impact  of  climate  change  on  America's  forests:  a  technical  document  supporting  the  2000  USDA  forest  service  RPA  assessment.  Fort  Collins,  CO:  Rocky  Mountain  Research  Station.  SLS  Consultants,  Inc.  (2008)  The  environmental  impacts  of  the  mail:  initial  life  cycle  inventory  model  and  analysis.  [Online]  Available  from:  http://www.it.twosides.info/content/rsPDF_286.pdf.  [Accessed:  2nd  June  2014]  Srivastava,  S.  (2007)  Green  supply-­‐chain  management:  A  state-­‐of-­‐the-­‐art  literature  review.  International  Journal  of  Management  Reviews.  9  (1).  p.  53-­‐80.  Stefan,  A.  &  Paul,  L.  (2008)  Does  it  pay  to  be  green?  A  systematic  overview.  Academy  of  Management  Perspectives.    22  (4).  p.  45-­‐62.  Stanny,  E.  and  Ely,  K.  (2008).  Corporate  environmental  disclosures  about  the  effects  of  climate  change.  Corporate  Social  Responsibility  and  Environmental  Management,  15(6),  pp.338-­‐348.  Stern,  N.  (2006)  Stern  Review  Final  Report.  [Online]  Available  from:  http://webarchive.nationalarchives.gov.uk/+/http:/www.hm-­‐treasury.gov.uk/independent_reviews/stern_review_economics_climate_change/stern_review_report.cfm.  [Accessed:  14th  February  2009]  Stevenson,  A.  (2008)  AMP-­‐Ohio’s  New  Coal  Plant  is  a  Bad  Bond  Bet.  [Online]  Available  from:  http://switchboard.nrdc.org/blogs/astevenson/ampohios_new_coal_plant_is_a_b.html.    [Accessed:  14th  February  2009]    Stevis,  D.  and  Assetto,  V.  (2001).  The  international  political  economy  of  the  environment.  Boulder:  Lynne  Rienner  Publishers.  Strauss,  A.  &  Corbin,  J.  M.  (1998)  Basics  of  qualitative  research:  Techniques  and  procedures  for  developing  grounded  theory.  2nd  Ed.  Thousand  Oaks,  CA:  Sage  Publications,  Inc.  Suh,  S.  (2004)  Functions,  commodities  and  environmental  impacts  in  an  ecological–economic  model.  Ecological  Economics.  48  (4).  p.  451–467.               135Suk  Yu  Hui,  and  Kai  Hau  Yeung,  (2003).  Challenges  in  the  migration  to  4G  mobile  systems.  IEEE  Communications  Magazine,  41(12),  pp.54-­‐59.  Sustainable  Apparel  Coalition,  (2014).  Sustainable  Apparel  Coalition  -­‐  Home.  [online]  Available  at:  http://www.apparelcoalition.org/  [Accessed  29  Jul.  2014].  SwissRe  (2007).  The  economic  justification  for  implementing  restraints  on  carbon  emissions.  [Online]  Available  from:  http://www.jth.ch/thinkBank/inlinemedia/displaymedia.php?id=39.  [Accessed:  25th  May  2011]  Tang,  Z.,  Hull,  C.  and  Rothenberg,  S.  (2012).  How  Corporate  Social  Responsibility  Engagement  Strategy  Moderates  the  CSR-­‐Financial  Performance  Relationship.  Journal  of  Management  Studies,  49(7),  pp.1274-­‐1303.  Toffel,  M.  W.  &  Horvath,  A.  (2004)  Environmental  implications  of  wireless  technologies.  News  delivery  and  business  meetings.  Environmental  Science  and  Technology.  38  (11).  p.  2961–2970.  UN  Principles  for  Responsible  Investing.  (2009)  Annual  Report  of  the  PRI  Initiative  2009.  [Online]  Available  from:  http://d2m27378y09r06.cloudfront.net/viewer/?file=wp-­‐content/uploads/PRIAnnualReport09.pdf.  [Accessed:  20th  October  2009]  UNFCCC:  United  Nations  Framework  Convention  on  Climate  Change.  (2009)  Kyoto  Protocol.  [Online]  Available  from:  http://unfccc.int/kyoto_protocol/items/2830.php.  [Accessed:  13th  February  2009]  United  Nations  Environmental  Programme  (UNEP)  Finance  Initiative.  (2006)  Show  Me  The  Money:  Linking  Environmental,  Social  and  Governance  Issues  to  Company  Value.  [Online]  Available  from:  http://www.unepfi.org/fileadmin/documents/show_me_the_money.pdf.  [Accessed:  21st  April  2011]  United  Nations.  (1987)  Report  of  the  World  Commission  on  Environment  and  Development:  Our  Common  Future.  [Online]  Available  from:  http://www.un-­‐documents.net/wced-­‐ocf.htm.  [Accessed:  25th  May  2011]  United  Nations.  (2009)  Nobel-­‐winning  economist  says  UN  best  suited  to  oversee  global  economic  overhaul.  UN  News  Service.  [Online]  Available  from:  http://www.un.org/apps/news/story.asp?NewsID=31830&Cr=financial+crisis&Cr1.  [Accessed:  25th  May  2011]  United  States  Department  of  Agriculture  (USDA).  (2011)  Lacey  Act.  [Online]  Available  from:  http://www.aphis.usda.gov/plant_health/lacey_act/index.shtml.  [Accessed:  25th  May  2011]  United  States  Postal  Service.  (2014)  First-­‐class  mail  volume  since  1926.  [Online]  Available  from:  http://about.usps.com/who-­‐we-­‐are/postal-­‐history/first-­‐class-­‐mail-­‐since-­‐1926.htm.  [Accessed:  1st  June  2014]               136USCAP:  United  States  Climate  Action  Partnership.  (2009)  About  USCAP.  [Online]  Available  from:  http://www.us-­‐cap.org/about/index.asp.  [Accessed:  14th  February  2009]  Van  Marrewijk,  M.  (2003).  Concepts  and  definitions  of  CSR  and  corporate  sustainability:  Between  agency  and  communion.  Journal  of  Business  Ethics.  44  (2-­‐3).  p  95-­‐105  Velayudhan,  A.,  and  Srividya,  S.  (2013).  Environmental  perspective  among  Indians  using  new  ecological  paradigm.  Indian  Journal  of  Health  and  Wellbeing,  4(6),  1260-­‐1265.  Virtanen,  Y.  &  Nilsson,  S.  (1993)  Environmental  impacts  of  waste  paper  recycling.  London:  Routledge.  Visseren-­‐Hamakers,  I.  &  Pattberg,  P.  (2013)  I  can't  see  the  forest  for  trees:  The  environmental  impact  of  global  forest  certification  is  unknown.  GAIA.  22  (1).  p.  25-­‐28.    Wackernagel,  M.  and  Rees,  W.  (1997).  Perceptual  and  structural  barriers  to  investing  in  natural  capital:  Economics  from  an  ecological  footprint  perspective.  Ecological  Economics,  20(1),  pp.3-­‐24.  Wal-­‐Mart  (2011).  Supplier  Sustainability  Assessment.  [Online]  Available  from:  http://walmartstores.com/download/4055.pdf.  [Accessed:  25th  May  2011]  Wang,  X.,  Padgett,  J.  M.,  Powell,  J.  S.  &  Barlaz,  M.  A.  (2013)  Decomposition  of  forest  products  buried  in  landfills.  Waste  Management.  33  (11).  p.  2267-­‐2276.  Weber,  C.  L.,  Koomey,  J.  G.  &  Matthews,  H.  S.  (2010)  The  Energy  and  Climate  Change  Implications  of  Different  Music  Delivery  Methods.  Journal  of  Industrial  Ecology.  14  (5).  p.  754-­‐769.  Western  Climate  Initiative  (2009).  Western  Climate  Initiative:  Home.  [Online]  Available  from:  http://www.westernclimateinitiative.org/.  [Accessed:  25th  May  2011]  Wilk,  R.  (2001).  Greeh  Households,  Domestic  Consumers,  Environment  and  Sustainability.  Ecological  Economics,  37(3),  pp.476-­‐477.  Williams,  E.  (2011)  Environmental  effects  of  information  and  communications  technologies.  Nature.  479  (7373).  p.  354-­‐358.  Williams,  E.  D.,  Ayres,  R.  U.  &  Heller,  M.  (2002)  The  1.7  Kilogram  Microchip:  Energy  and  Material  Use  in  the  Production  of  Semiconductor  Devices.  Environmental  Science  &  Technology.  36  (24).  p.  5504-­‐5510.  Winjum,  J.,  Brown,  S.  and  Schlamadinger,  B.  (1998).  Forest  Harvests  and  Wood  Products:  Sources  and  Sinks  of  Atmospheric  Carbon  Dioxide.  Forest  Science,  44(2),  pp.272-­‐284.  Woerdman,  E.,  Arcuri,  A.  and  Clò,  S.  (2008).  Emissions  Trading  and  the  Polluter-­‐Pays  Principle:  Do  Polluters  Pay  under  Grandfathering?.  Review  of  Law  &  Economics,  4(2).                 137WWF.  (n.d.)  Climate  Savers  –  Catalyst.  [Online]  Available  from:  http://wwf.panda.org/what_we_do/how_we_work/businesses/climate/climate_savers/partner_companies/.  [Accessed:  25th  May  2011]  Yaziji,  M.  and  Doh,  J.  (2009).  NGOs  and  corporations.  Cambridge:  Cambridge  University  Press.  Yi,  L.  and  Thomas,  H.  (2007).  A  review  of  research  on  the  environmental  impact  of  e-­‐business  and  ICT.  Environment  International,  [online]  33(6),  pp.841-­‐849.  Available  at:  http://dx.doi.org/10.1016/j.envint.2007.03.015  [Accessed  3  Jul.  2014].  Yin,  R.  K.  (2009)  Case  Study  Research:  Design  and  Methods.  Thousand  Oaks,  CA:  Sage  Publications,  Inc.  York,  R.  (2006)  Ecological  paradoxes:  william  stanley  jevons  and  the  paperless  office.  Human  Ecology  Review.  [Online]  13  (2).  p.  143-­‐147.  Available  from:  http://ajph.humanecologyreview.org/pastissues/her132/york.pdf.  [Accessed:  1st  June  2014]  Yu,  J.,  Williams,  E.,  Ju,  M.  and  Yang,  Y.  (2010).  Forecasting  Global  Generation  of  Obsolete  Personal  Computers.  Environmental  Science  &  Technology,  44(9),  pp.3232-­‐3237.  Zackrisson,  M.,  Rocha,  C.,  Christiansen,  K.  and  Jarnehammar,  A.  (2008).  Stepwise  environmental  product  declarations:  ten  SME  case  studies.  Journal  of  Cleaner  Production,  16(17),  pp.1872-­‐1886.  Zalasiewicz,  J.  et  al.  (2010).  The  New  World  of  the  Anthropocene.  Environment  Science  &  Technology.  44  (7),  2228-­‐2231.  Zhang,  Y.  and  Buongiorno,  J.  (1997).  Communication  media  and  demand  for  printing  and  publishing  papers  in  the  United  States.  Forest  Science,  43(3),  pp.362-­‐377.  Zuberbühler  Associates.  (2013)  Pc  vs  tablet  vs  network  pc  vs  smartphone  sales.  [Online]  Available  from:  http://zuberbuehler-­‐associates.ch/blog/?p=885.  [Accessed:  30th  May  2014]  Zurkirch,  M.  &  Reichart,  I.  (2002)  The  Ecology  of  the  New  Economy.  Sheffield,  UK:    Greenleaf  Publishing.               138Appendices  Appendix  A      The  following  set  of  questions  was  used  to  structure  the  interviews  along  the  paper  supply  chain.  These  questions  formed  part  of  a  semi-­‐structured  interview,  and  additional  questions  or  follow-­‐ups  were  commonly  employed.       1.  Is  reducing  carbon  emissions  important  to  your  corporation?  If  so,  why?  2.  If  a  carbon  management  policy  exists,  what  has  motivated  its  development?  3.  What  has  guided  the  development  of  a  carbon  management  policy?    4.  How  do  environmental  issues  such  as  carbon,  manifest  themselves  along  your  supply  chain?  (for  example,  market  demand  for  carbon  neutral  paper,  pressure  from  NGOs)  5.  Have  your  performance  measures  and  monitoring  and  evaluation  procedures  changed  as  a  result  of  trying  to  reduce  carbon  emissions?  If  so,  how  has  this  impacted  supply  chain  management  for  your  organization?  6.  What  external  actors  (NGOs,  Government,  Corporations)  have  you  been  engaged  with  in  considering  carbon  emissions?  What  has  their  role  been?  7.  Does  collaboration  along  your  supply  chain  play  a  role  in  your  organization’s  carbon  management  policy?  8.  Has  the  pursuit  of  a  carbon  management  policy  affected  your  financial  performance?    9.  Have  you  changed  any  relationships  along  your  supply  chain  in  order  to  reduce  carbon  emissions?  If  so,  please  describe  the  impacts  of  this  change.    10.  Do  you  anticipate  changing  your  supply  chain  management  in  order  to  deal  with  government  regulations  associated  with  carbon?  11.  Where  do  you  see  your  business  in  10  years?  20  years?  How  will  this  affect  your  supply  chain?                   139  Appendix  B      The  following  survey  questions  were  delivered  through  a  website  to  our  survey  sample.      Are  you  Male  or  Female?    What  is  your  age?  •  14  to  17  •  18  to  21  •  22  to  25  •  26  to  30  •  31  to  40  •  41  to  50  •  51  to  60  •  61  or  older    What  best  describes  your  education  level?  •  Less  than  High  School  •  High  School  /  GED  •  Some  college  •  Bachelor’s  Degree  •  Master’s  Degree  •  Doctoral  Degree  •  Professional  Degree  (JD,  MD,  etc.)    What  is  your  own  yearly  income?  •  Under  $10,000  •  $10,000  to  $19,999  •  $20,000  to  $29,999  •  $30,000  to  $39,999  •  $40,000  to    $49,999  •  $50,000  to  $74,999  •  $75,000  to  $99,999  •  $100,000  to  $150,000  •  Over  $150,000  •  Would  rather  not  say    What  is  your  total  yearly  household  income,  including  all  earners  in  your  household?  •  Under  $10,000  •  $10,000  to  $19,999  •  $20,000  to  $29,999  •  $30,000  to  $39,999  •  $40,000  to    $49,999  •  $50,000  to  $74,999  •  $75,000  to  $99,999  •  $100,000  to  $150,000  •  Over  $150,000               140•  Would  rather  not  say    Which  of  the  following  best  describes  the  area  you  live  in?  •  Urban    •  Suburban  •  Rural    Which  of  the  following  categories  best  describes  your  primarily  area  of  employment  (regardless  of  your  actual  position)?  •  Homemaker  •  Retired  •  Student  •  Unemployed  •  Agriculture,  Forestry,  Fishing,  or  Hunting  •  Arts,  Entertainment,  or  Recreation  •  Broadcasting  •  Education  -­‐  College,  University,  or  Adult  •  Education  -­‐  Primary/Secondary  (K-­‐12)  •  Education  -­‐  Other  •  Construction  •  Finance  and  Insurance  •  Government  and  Public  Administration  •  Health  Care  and  Social  Assistance  •  Hotel  and  Food  Services  •  Information  -­‐  Services  and  Data  •  Information  -­‐  Other  •  Processing  •  Legal  Services  •  Manufacturing  -­‐  Computer  and  Electronics  •  Manufacturing  -­‐  Other  •  Military  •  Mining  •  Publishing  •  Real  Estate,  Rental,  or  Leasing  •  Religious  •  Retail  •  Scientific  or  Technical  Services  •  Software  •  Telecommunications  •  Transportation  and  Warehousing  •  Utilities  •  Wholesale  •  Other  (Please  specify)    Which  of  the  following  best  describes  your  role  in  industry?  •  Upper  management  •  Middle  management  •  Junior  management               141•  Administrative  staff  •  Support  staff  •  Student  •  Trained  professional  •  Skilled  laborer  •  Consultant  •  Temporary  employee  •  Researcher  •  Self-­‐employed  •  Other  (Please  specify)    The  organization  you  work  for  is  in  which  of  the  following:  •  Public  sector  •  Private  sector  •  Not-­‐for-­‐profit  •  Don't  know  •  Other  (Please  specify)    Section  Two:  Media  Consumption  How  long  have  you  been  using  the  Internet?  •  Never  used  it  •  Less  than  6  months  •  6  to  12  months  •  1  to  3  years  •  4  to  6  years  •  7  years  or  more    How  many  computers  are  used  in  your  household?  •  1  •  1  to  3  •  4  or  more    How  many  mobile  devices  are  used  in  your  household?  •  1  •  1  to  3  •  4  or  more    Do  you  own  the  following  devices:   Yes   No   Don’t  Know  Digital  Camera        Digital  Video  Camera        DVD  Player        BluRay  Player        LCD  or  Plasma  Flat  Panel  Television        MP3  Player  (such  as  Apple  IPod)        Desktop  Computer        Laptop  Computer        Netbook  Computer        Cellphone                     142Smartphone  or  Handheld  Mobile  Device  (such  as  Apple  IPhone)        E-­‐Reader  (such  as  Amazon  Kindle)          How  often  do  you  replace  the  following  devices:  >  1  year  Every  year   Every  1  to  2  years   3  years  or  more  Desktop  computer          Laptop  computer          Cellphone,  smartphone  or  handheld  device            How  frequently  do  you  access  the  web  from  the  following  places?  Daily   Weekly  Monthly   Less  than  once  a  month  Never  From  home  (including  a  home  office)            From  work            From  school            From  a  public  terminal  (e.g.  library,  cybercafé,  etc.)            From  a  mobile  device            From  other  places              What  type  of  Internet  connection  do  you  have  at  home?  •  Dial-­‐up  •  DSL  Broadband  •  Cable  Broadband  •  Don’t  know  •  Other  (Please  specify)    Do  you  use  a  mobile  device  to  connect  to  the  Internet?  •  Yes  •  No  •  Don’t  know    Where  do  you  get  your  news  online  (select  those  which  apply)?  •  Dedicated  news  sites  (e.g.  CNN.com,  etc.  )  •  National  newspaper  websites    •  Regional  newspaper  websites    •  Local  newspaper  websites  •  News  aggregators  (e.g.  Google  News,  Yahoo  News,  etc.)  •  Blogs  •  Other  (Please  specify)    What  type  of  paper-­‐based  media  do  you  consume  (select  those  which  apply)?  •  National  newspapers  •  Regional  newspapers  •  Local  newspapers  •  News  magazines  •  Entertainment  magazines  •  Hobbyist  or  niche  magazines  •  Books  (purchased)  •  Books  (loaned  from  library)               143  Please  rate  how  frequently  do  you  use  the  following  online  services  using  the  following  scale:  Never  (N),  Not  Frequently  (NF),  Somewhat  Frequently  (SF),  Frequently  (F),  Very  Frequently  (VF),  Don’t  Know  (DK):     Instant  Messaging              Email              Social  networking              Blogging  sites              Research              Reading  the  news              Watching  videos              Online  banking  /  financial  management              Online  Shopping                Please  state  how  many  hours,  on  average,  you  spend  a  day  doing  the  following  activities,  using  the  following  scale:  None   Less  than  1   1  to  2   2  to  4   4  to  6   6  or  more   Don’t  Know  Browsing  the  internet                Reading  a  newspaper                Reading  a  magazine                Reading  a  book                  Please  state  your  willingness  to  pay  for  content  on  the  following  printed  media  formats,  answer  Yes  (Y),  No  (N),  Maybe  (M)  or  Don’t  Know  (DK).  Printed  newspaper          Printed  magazine          Printed  book          Online  newspaper          Online  magazine          Online  blogs            Electronic  books            Listed  below  are  various  information  sources.  Please  state  how  credible  you  feel  the  content  and  information  is  (Not  Credible  (NC),  Somewhat  Credible  (SC),  Credible  (C),  Very  Credible  (VC),  Don’t  Know  (DK):   Printed  Newspapers            Printed  Magazines            Printed  Books            Online  Newspapers            Online  Magazines            Online  News  Aggregators            Blogs              Please  rate  how  effective  you  find  the  following  types  of  advertising,  based  on  this  scale:  Not  Effective  (NA),  Somewhat  Effective  (SE),  Effective  (E),  Very  Effective  (VE),  Don’t  Know):  Newspapers  –    Display  Ad              Newspapers  –  Classified  Ad            Newspapers  –  Flyer  Insert            Magazine  –  Display  Ad                         144Online  –  Display  Ad            Online  –  Pop-­‐Up  Ad            Online  –  Text  Based  Ad              Please  answer  the  following  questions  on  your  trends  in  print  media  consumption.  I  ask  that  you  give  a  sense  of  your  consumption  patterns  five  years  ago,  today,  and  how  you  think  you  will  consume  in  five  years  from  now  using  the  following  scale:  Never  –  Rarely  –  Somewhat  Often  –  Often  –  Very  Often  Newspaper  or  magazine  delivered  to  home?  Books  purchased  from  bookstore?  Books  loaned  from  library?  Read  news  on  PC  or  laptop  computer  (including  newspapers,  magazines,  blogs,  etc.)  ?  Read  news  on  smartphone,  cellphone,  or  other  handheld  device  (from  media  outlets  such  as  newspapers,  magazines,  blogs,  etc.)?  Books  purchased  for  electronic  reader?    Listed  below  are  statements  about  the  relationship  between  humans  and  the  environment.    For  each  one,  please  indicate  whether  you  Strongly  Agree  (SA),  Mildly  Agree  (MA),  are  Unsure  (U),  Mildly  Disagree  (MD)  or  Strongly  Disagree  (SD).     1.  I  are  approaching  the  limit  of  the  number  of  people  the  earth  can  support        2.  Humans  have  the  right  to  modify  the  natural  environment  to  suit  their  needs        3.  When  humans  interfere  with  nature  it  often  produces  disastrous  consequences      4.  Human  ingenuity  will  insure  that  I  do  NOT  make  the  earth  unlivable          5.  Humans  are  severely  abusing  the  environment            6.  The  earth  has  plenty  of  natural  resources  if  I  just  learn  how  to  develop  them      7.  Plants  and  animals  have  as  much  right  as  humans  to  exist            8.  The  balance  of  nature  is  strong  enough  to  cope  with  the  impacts  of  modern  industrial  nations    9.  Despite  our  special  abilities  humans  are  still  subject  to  the  laws  of  nature        10.  The  so-­‐called  “ecological  crisis”  facing  humankind  has  been  greatly  exaggerated      11.  The  earth  is  like  a  spaceship  with  very  limited  room  and  resources          12.  Humans  were  meant  to  rule  over  the  rest  of  nature            13.  The  balance  of  nature  is  very  delicate  and  easily  upset            14.  Humans  will  eventually  learn  enough  about  how  nature  works  to  be  able  to  control  it    15.  If  things  continue  on  their  present  course,  I  will  soon  experience  a  major  ecological  catastrophe              Rate  your  level  of  environmental  concern  associated  with  the  following  scale:  None   Unconcerned   Somewhat  Concerned   Concerned   Very  Concerned  Deforestation  and  illegal  logging…              Global  warming  and  climate  change…            Mining  of  minerals  and  metals…            Depletion  of  non-­‐renewable  fossil  fuel  resources…            Nuclear  power  and  nuclear  waste…            Water  use  and  conservation…            Handling  of  electronic  waste…            Growth  and  increasing  consumption  in  developing  countries…              Growth  and  increasing  consumption  in  developing  countries…                   145What  types  of  environmental  issues  concern  you  most?  •  Local  (near  your  home)  •  Regional  (in  your  state  or  country)  •  Global  (affecting  the  entire  planet)    Many  environmental  issues  involve  difficult  trade-­‐offs  with  the  economy.  Which  of  the  following  statements  best  describes  your  view?  (a)  The  highest  priority  should  be  given  to  protecting  the  environment,  even  if  it  hurts  the  economy.  (b)  Both  the  environment  and  the  economy  are  important,  but  the  environment  should  come  first.  (c)  Both  the  environment  and  the  economy  are  important,  but  the  economy  should  come  first.  (d)  The  highest  priority  should  be  given  to  economic  considerations  such  as  jobs  even  if  it  hurts  the  environment.    Do  you  believe  that  I  have  a  responsibility  to  look  out  for  the  interests  of  future  generations,  even  if  it  means  making  ourselves  worse  off?  (a)  Yes  (b)  No