26
Damage from a Change in the Climate Prof. Charles D. Kolstad Stanford University 25 July 2013 1

Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Damage  from  a  Change  in  the  Climate  

Prof.  Charles  D.  Kolstad  Stanford  University  

25  July  2013  

1  

Page 2: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

What  is  damage  from  climate  change?  

•  An  Individual:  the  harm  (posiHve  or  negaHve)  caused  by  a  change  in  climate  relaHve  to  a  baseline  (not  absolute),  expressed  in  terms  of  a  numeraire  good.    Equivalently,  WTP  to  avoid  change  in  climate.  

•  Society:  aggregate  damage,  summed  over  individuals,  space  and  Hme,  also  from  a  relaHve  change  in  climate,  expressed  in  terms  of  a  numeraire  good  

2  

Page 3: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Damage  equivalent  to  WTP  to  avoid  change  

3  

Climate  (eg  T)  

MWTP  

     T0                    T1  

WTP  

Page 4: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Metrics  and  AdaptaHon    

•  Climate  change  has  many  relevant  dimensions  – Mean  weather  –  Variance  of  weather  –  PrecipitaHon  vs  temperature  –  Sea  level  –  Catastrophic  change  –  Transient  –  Hme  path  of  the  change  

•  Implicit  is  that  individuals,  firms,  governments  take  steps  to  autonomously  adapt  to  the  change  in  the  climate  

•  Along  a  path  of  change,  firms  may  adjust  and  incur  losses  –  example  of  sea  level  rise.  

4  

Page 5: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Some  Categories  of  damage  

•  MARKET  IMPACTS  –  Energy  consumpHon  – Agriculture  – Water  resources  –  Infrastructure  –  Tourism  –  Real  Estate  

•  NONMARKET  IMPACTS  – Health  – AestheHcs  –  Ecosystem  services  

5  See  OECD,  2009  

Page 6: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Why  do  we  need  aggregate  measures  of  damage?  

•  Determining  the  right  amount  of  miHgaHon  •  Understanding  to  consequences  of  not  taking  acHon  

•  Use  in  IAMs,  which  in  turn  may  be  used  in  policy  analysis.  

•  Use  in  compuHng  Social  Cost  of  Carbon  for  use  in  decentralizing  carbon  regulaHons  

6  

Page 7: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

7  Tol,  2009  

Page 8: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Aggregate  Measures  of  Damages  from  CC  2.3  

-­‐1.4  -­‐1.9  

0.0  0.1  

-­‐1.5  

0.9  

-­‐2.2  

-­‐10.7  

-­‐0.5  

-­‐11.0  

-­‐10.0  

-­‐9.0  

-­‐8.0  

-­‐7.0  

-­‐6.0  

-­‐5.0  

-­‐4.0  

-­‐3.0  

-­‐2.0  

-­‐1.0  

0.0  

1.0  

2.0  

3.0  

0.0   1.0   2.0   3.0   4.0   5.0   6.0   7.0  

%GDP  

degree  cen8grade  

Tol  2002a  

Fankhauser  1995  

Tol  1995  

Mendelsohn  et  al.  2000a  (experimental)  

Mendelsohn  et  al.  2000a  (cross-­‐secHonal)  

Nordhaus  and  Boyer  2000  

Hope  2006  

Nordhaus  2007  (2.5  degrees)  

Nordhaus  2007  (6.0  degrees)  

Bosello  et  al.  2012  Source:  adapted  from  Tol  (2013)  

8  

Page 9: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Issues  of  AggregaHon  I  •  Across  Impacts  – Need  a  common  metric  so  that  a  variety  of  different  damages  may  be  aggregated.  

•  Across  space  – AggregaHon  of  losses  in  Bangladesh  with  losses  in  US  can  lead  to  US  swamping  Bangladesh.    Need  prioritarian  perspecHve  in  aggregaHon  (weights)  

– Differences  in  VSL  provide  possible  insights  into  welfare  weights  

9  

Page 10: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

10  Source:  OECD  

Page 11: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Issues  of  AggregaHon  II  

•  Across  Hme  – Market  discount  rate  is  the  rate  at  which  individuals  tade  off  Hme  

–  Social  discount  rate  is  what  governments  use  in  making  intertemporal  decisions  •  r=  d  +  e  g,  where  d  is  rate  of  pure  Hme  preference  for  well-­‐being,  g  is  growth  rate  of  consumpHon  and  e  represents  aversion  to  inequality    

•  d  is  typically  close  to  zero;  e  is  typically  1-­‐2;  g  is  typically  2-­‐3  (but  can  be  negaHve).    Implies  r  1-­‐2  Hme  the  growth  rate.  

•  d  is  purely  normaHve  but  some  reality  checking  with  how  willing  one  is  to  transfer  resources  to  poorer  people  at  same  point  in  Hme.  

11  

Page 12: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Social  Discount  Rate  from  Ramsey  Rule  

12  From:  Gollier,  2012  

Page 13: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Real  Market  Rates  

13  Source:  Gollier,  2012  

Page 14: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Measuring  Damage  

•  Requires  knowledge  of  the  quanHtaHve  biological  and  physical  impacts  of  a  parHcular  climate  change  

•  Ojen  the  natural  science  is  insufficiently  precise  to  be  able  to  generate  a  meaningful  esHmate  of  damage  

14  

Page 15: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Chain  of  CausaHon  Human  Ac8vity  

Mi8ga8on  

Emissions  (E)  (CO2,  CH4,  N2O,  HFC,  PFC,  NOx,  SO2,  …)  

Atmospheric  Concentra8ons  (C)  

Radia8ve  Forcing  (RF)  

Climate  Change    Temperature  (ΔT),  precipitaHon,  (Δp),  winds,  soil  moisture,  

extreme  events,  sea  level  (ΔSL)  

Impacts    Agriculture  and  forestry,  ecosystems,  energy  producHon  and  

consumpHon,  social  effects  

Damages    Welfare  loss  (e.g.,  monetary  units)  

Human  Ac8vity  15  

Page 16: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Methods  for  Measuring  Damage  

•  Revealed  Preference  Methods  –  Economist  always  suspicious  of  what  people  SAY  they  will  do  vs  what  they  ACTUALLY  do  –  acHons  speak  louder  than  words.  

–  Best  to  try  to  look  at  historic  analogues  rather  than  speculate  about  future  

–  Look  at  behavior  in  markets  related  to  environmental  harm  –  Eg,  how  much  to  house  prices  change  when  air  quality  gets  beoer  

–  Typically  preferred  since  real  economic  decisions  are  involved  •  Stated  preferences  

–  ConHngent  valuaHon  –  respondents  are  asked  their  WTP  –  Referenda  –  respondents  vote  on  expenditure  issue  with  associated  environmental  benefits  •  Eg,  Kahn  and  Matsusaka—California  voHng  on  environmental  proposiHons   16  

Page 17: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Revealed  Preference  Methods  

•  Hedonic  – Look  at  how  price  of  market  good  varies  with  different  levels  of  associated  environmental  good  

•  Household  producHon  – Look  at  expenditures  made  to  adapt  to  different  levels  of  the  environmental  good  –  defensive  expenditures  

– Travel  cost  is  an  example  which  is  widely  used  for  recreaHon  deman  

17  

Page 18: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Example:  Market  Impacts  on  Agriculture  Classic  Analysis  of  Medelsohn,  Nordhaus  and  Shaw  (1994)  

18  

Source:  M

ende

lsoh

n  et  al,  1994  

AVERAGE  LAND  VALUES  

Page 19: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Extensions  of  simple  cross-­‐secHon  analysis  

•  Panel  –  pooled  Hme  series,  cross  secHon.    •  Look  at  ag  profits  rather  than  land  prices;  look  at  weather  anomalies  as  well  as  climate  (Kelly  et  al,  2005)  

•  Applied  to  other  countries  (Series  of  papers  by  Mendelsohn  et  al  and  others)  

•  Improve  metrics  of  climate  change  (Deschenes  and  Greenstone,  2009)  –  Difficult  to  find  summary  staHsHcs  for  weather/climate  

•  One  problem  is  that  factor  prices  are  held  constant  •  Result  is  a  good  understanding  of  the  effect  of  climate  change  on  agriculture  

19  

Page 20: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Nonmarket  Impact:  Amenity  Values  of  Climate  (Maddison  and  Bigano,  2005)  

20  Source:  Kolstad,  2010  

Page 21: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Details  

•  Data  from  95  Italian  Provinces  •  1991-­‐95  •  Measure  discreHonary  income  •  Jan,  July  temperature,  precip  and  cloudcover  •  Conclude:  – 1  degree  increase  in  July  temp  causes  household  loss  of  350  euros  per  year  

– 1  mm  increase  in  Jan  precip  generates  loss  of  15  euros  per  year  

21  

Page 22: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Methods  for  Ecosystem  Services  

•  Use  value  has  generated  low  esHmates  of  value  (eg,  pharmaceuHcal  value  of  tropical  biodiversity)  

•  Ecosystem  services  are  generally  detached  from  markets  –  revealed  preference  methods  inadequate  (so  far)  

•  CV  one  of  the  only  methods  –  fraught  with  problems.  

22  

Page 23: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Social  Cost  of  Carbon  

•  Important  and  controversial  

•  Useful  for  decentralizing  carbon  regulaHon  domesHcally  

•  Defined  as  the  net  present  value  of  marginal  damages  with  respect  to  carbon  emissions,  condiHonal  on  a  baseline  regulatory  policy  

•  NOT  the  marginal  net  benefits,  as  incorrectly  used  by  US  Government  (though  concepts  are  the  same  absent  any  sort  of  carbon  regulaHon)  

23  

Page 24: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Co-­‐benefits  

•  Example:  If  carbon  regulaHons  are  Hghtened,  sulfur  emissions  may  decline  

•  Incorrect  to  value  those  co-­‐benefits  based  on  the  averted  damage  from  sulfur  

•  Correct  to  look  at  the  marginal  net  welfare  implicaHons  of  decreasing  carbon  by  one  ton,  condiHonal  on  a  trajectory  of  emissions  

•  Note:  Envelope  theorem  tells  us  if  the  sulfur  is  opHmally  regulated,  there  will  be  no  welfare  consequences  on  one  less  ton  of  sulfur  

24  

Page 25: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Benefits  Transfer  

•  Benefits  transfer  involves  using  a  study  done  in  one  locaHon  to  infer  the  damage  from  polluHon  in  a  different  locaHon.  

•  Carefully  done,  it  can  be  useful  •  Will  be  very  important  to  do  significant  amounts  of  benefit  transfer  in  the  climate  area,  parHcularly  for  nonmarket  damages.    World  is  too  big.  

25  

Page 26: Damage&from&aChange&in&the& Climate&web.stanford.edu/group/emf-research/docs/sm/2013/7-25/Kolstad_M… · Damage&from&aChange&in&the& Climate& Prof.&Charles&D.&Kolstad& Stanford&University&

Conclusions  

•  Significant  need  for  a  research  program  to  catalog  the  monetary  impacts  of  a  change  in  the  climate  

•  Beoer  esHmates  of  aggregate  damage  from  climate  change  are  long  overdue  

•  Will  we  ever  be  able  to  get  something  comprehensive  (like  the  “Cost  of  Clean”)?  

26