China’s Influence on Deforestation in Brazilian

China’s  Influence  on  Deforestation  in  Brazilian   Amazonia:  A  Growing  Force  in  the  State  of  Mato  Grosso   China  influences  deforestation  in  Brazilian  Amazonia  in  a  variety  of  ways,  including   the  direct  influence  of  Chinese  enterprises.  We  examine  these  issues  and  present   data  on  the  growth  of  China’s  role  in  Brazil’s  soy  and  beef  sectors,  which  are  major   drivers  of  deforestation  in  the  country’s  Amazon  region.  We  concentrate  on  the  state   of  Mato  Grosso,  where  soy  and  beef  production  are  dominant  forces  and  where   China  is  the  principal  destination  for  exports.  China  also  purchases  other   commodities  from  Brazilian  Amazonia,  such  as  iron  ore  and  timber.  Chinese   financing  is  increasingly  influencing  and  accelerating  infrastructure  development   projects  such  as  a  planned  railway  that  would  connect  Mato  Grosso  to  ports  on  the   Amazon  River  in  order  to  facilitate  soy  exports.   We  find  that  increases  in  Brazil’s  exports  to  China  are  significantly,  positively   associated  with  increasing  deforestation  rates.  Nevertheless,  deforestation  has  been   declining  in  recent  years,  thanks  to  improved  regulation,  including  a  powerful  new   measure  that  disqualifies  any  operation  with  environmental  irregularities  from   benefiting  from  public  loans.  However,  the  future  of  this  policy  is  not  secure.  The   influx  of  money  from  the  booming  Chinese  export  market  is  reshaping  the  Brazilian   political  landscape.  The  “ruralist”  voting  block  representing  large  landowners  has   used  its  newfound  influence  to  push  for  relaxing  environmental  regulation.  If  Brazil   is  to  consolidate  its  gains  against  deforestation,  especially  during  an  agricultural   export  boom,  it  will  need  to  hold  fast  to  its  regulatory  progress  and  resist  the  call  to   sacrifice  long-­‐term  conservation  goals  for  short-­‐term  export  revenue.   1. Introduction As  a  country  with  vast  natural  resources,  including  agricultural  land,  timber,   hydroelectric  capacity  and  mineral  deposits,  Brazil  is  a  logical  source  of  imports  to   supply  China’s  burgeoning  demand.  Brazil  is  also  a  logical  destination  for  Chinese   investment,  particularly  in  the  extraction  or  production  of  commodities  and  the   transport  infrastructure  needed  to  facilitate  export.  Brazil’s  political  stability  and   openness  to  foreign  investment,  combined  with  the  financial  power  provided  by   China’s  strong  economy,  translate  into  the  impressive  growth  and  scale  of  China’s   presence  in  Brazil.  China  has  become  Brazil’s  largest  trading  partner  and  the   greatest  source  of  export  surplus  from  agricultural  goods  (US$85  billion  in  2011),   contributing  to  Brazil’s  economic  growth  and  to  reducing  the  country’s  vulnerability   to  external  economic  crises  (da  Nóbrega,  2012).  It  should  therefore  not  come  as  a   surprise  that  China  exerts  multiple  influences  on  events  in  Brazil,  often  to  the   detriment  of  Amazon  forest.   Brazil’s  state  of  Mato  Grosso  (Figures  1  and  2)  is  one  of  the  main  targets  for  Chinese   investment  and  is  the  source  of  most  of  one  of  the  major  Brazilian  exports:   soybeans.  Mato  Grosso  is  twice  the  size  of  the  US  state  of  California  and  is  one  of  the   most  important  agricultural  areas  in  South  America  and  in  the  world.  In  the  2012-­‐ 2013  agricultural  year  Mato  Grosso’s  estimated  production  was  23.5  million  tons  of   soybeans  (29%  of  Brazil’s  production),  with  7.0  million  hectares  planted  and  a   productivity  of  2959  Kg/ha  (Brazil,  IBGE,  2013).  Improvements  in  transport   infrastructure  are  expected  to  allow  a  much  larger  area  to  be  planted  with  soy,   2   especially  in  the  northern  part  of  the  state.  Chinese  land  purchases  in  Mato  Grosso   have  been  mostly  for  soy,  but  they  also  include  some  areas  for  cotton  production.   Figure  1:  Cities,  highways,  railways  and  protected  areas  mentioned  in  the  text   Cities:  1  =  Itacoatiara,  2  =  Porto  Velho,  3  =  Carauarí,  4  =  Altamira,  5=  Barcarena,  6  =  Cuiabá,  7  =   Santarém,  8  =  Juína,  9  =  Porto  dos  Gaúchos,  10  =  Vilhena,  11  =  Sapezal,  12  =  Lucas  do  Rio  Verde,   13  =  Belém,  14  =  Manaus,  15    Cruzeiro  do  Sul,  16  =  Anápolis,  17  =  Sorriso,  18  =  Querência,  19  =   Santa  Cruz,  20  =  Piura.  Protected  areas:  21  =  Amazonia  National  Park,  22  =  Juruena  National   Park,  23  =  Xingu  Indigenous  Park.  Highways:  24  =  BR-­‐319,  25  =  BR-­‐163,  26  =  MT-­‐322,  27  =  MT-­‐ 319. Railways:  28  =  Transcontinental  Railway,  29  =  North-­‐South  Railway,  39  =  Ferronorte Railway.  Other:  31  =  Carajás  Mine.  Brazilian  states:  AC  =  Acre,  AM  =  Amazonas,  AP  =  Amapá,  GO =  Goiás,  MA  =  Maranhão,  MG  =  Minas  Gerais,  MT  =  Mato  Grosso,  PA  =  Pará,  PR  =Paraná,  RJ  =  Rio de  Janeiro,  RO  =  Rondônia,  RR  =  Roraima,  RS  =  Rio  Grande  do  Sul,  SC  =  Santa  Catarina,  SP  =  São Paulo,  TO  =  Tocantins. 3   Figure  2:  Dams,  rivers  and  waterways  mentioned  in  the  text   Dams:  1  =  Belo  Monte  Dam,  2  =  Samuel  Dam.  3  =  Jirau  Dam,  4  =    Santo  Antônio  Dam,  5  =   Guajará  Mirim  (Cachoeira  Riberão)  Dam,  6  =  São  Luiz  do  Tapajós  Dam,  7  =  Jatobá  Dam,  8  =   São  Simão  Alto  Dam,  9  =  Salto  Augusto  Baixo  Dam,  10  =  Sinop  Dam,  11  =  São  Manoel  Dam,   12  =  Chacorão  Dam.  Rivers:  13  =  Madeira  River,  14  =    Xingu  River,  15  =  Amazon  River,  16  =   Teles  Pires  River,  17  =  Juruena  River.  18  =  Arinos  River,  19  =  Tapajós  River.  Waterways:  20   =  Guaporé  Waterway,  21  =  Madeira  Waterway,  22  =  Tapajós  Waterway.  Brazilian  states:   AC  =  Acre,  AM  =  Amazonas,  AP  =  Amapá,  GO  =  Goiás,  MA  =  Maranhão,  MG  =  Minas  Gerais,   MT  =  Mato  Grosso,  PA  =  Pará,  PR  =Paraná,  RJ  =  Rio  de  Janeiro,  RO  =  Rondônia,  RR  =   Roraima,  RS  =  Rio  Grande  do  Sul,  SC  =  Santa  Catarina,  SP  =  São  Paulo,  TO  =  Tocantins.   4   2. Brazil’s  Exports  to  China  and  the  World Brazil’s  exports  to  the  world,  including  China,  increased  dramatically  over  the  2003-­‐ 2008  period.  The  annual  mean  growth  rate  of  Brazilian  exports  in  the  1990-­‐2002   period  was  5.6%,  but  the  annual  rate  jumped  to  22%  for  the  2003-­‐2008  period   (Bittencourt  et  al.,  2012,  p.  102).  The  global  economic  downturn  then  led  to  a  fall  in   total  Brazilian  exports  in  2009,  but  economic  recovery  was  followed  by  a  new   export  record  in  2010.  Exports  bound  for  China  grew  much  faster  than  the  trend  for   Brazilian  exports  in  general:  over  the  2000-­‐2008  period  exports  to  China  grew  at  an   average  annual  rate  of  40.4%.  They  even  grew  by  23.1%  in  2009  when  Brazil’s  total   exports  fell  by  22.2%  (Bittencourt  et  al.,  2012,  p.  103).  The  percentage  of  Brazil’s   exports  represented  by  China  grew  from  only  2%  (or  2.8%  including  Hong  Kong  and   Macao)  in  2000  to  19%  (or  20.4%  including  Hong  Kong  and  Macao)  in  2012,  making   China  Brazil’s  number  one  individual  market,  as  shown  in  Figure  3  (UN  Comtrade,   2013).  This  percentage  doubled  in  the  2008-­‐2012  period.  The  increased  exports  to   China  were  mirrored  by  decreases  in  exports  to  the  United  States  and  the  European   Union.     Figure  3:  Brazil’s  Exports  to  China  as  a  Share  of  All  Brazilian  Exports,   2000-­‐2012.   Source:  UN  Comtrade  (2014).   5   The  nature  of  what  was  being  exported  also  changed,  with  primary  commodities   increasing  from  38.9%  in  2000  to  44.9%  in  2008  and  51.0%  in  2009,  when  China   became  dominant  (Bittencourt  et  al.,  2012,  p.  106).  The  increasing  role  of  China  is  an   important  factor  in  this  shift.  In  2000,  Brazil  accounted  for  27%  of  Latin  America’s   primary  and  resource-­‐based  products  (PRBP)  exports  to  China,  but  that  share  grew   to  48%  in  2012,  after  a  peak  of  51%  in  2011,  as  shown  in  Figure  4  (UN  Comtrade,   2013).  For  soybeans  and  other  seeds  this  percentage  grew  from  39%  in  2000  to   79%  in  2012,  after  a  peak  of  83%  in  2009  (SITC  Revision  3,  222).  For  iron  ore  and   iron  concentrate  the  percentage  each  year  was  always  between  87  and  91%  (SITC   Revision  3,  281).  For  crude  petroleum,  after  some  variation  in  the  2000-­‐2003   period,  the  percentage  stabilized  between  59  and  65%  after  2007  (SITC  Revision  3,   333).  All  of  these  products  have  significant  impact  on  the  environment.   Unfortunately,  cross-­‐national  studies  indicate  that  export  of  primary  commodities  is   associated  with  the  least  gains  in  indicators  of  human  wellbeing  in  the  exporting   countries  (Carmignani  &  Avom,  2010).   6   Figure  4.  Share  of  Brazil  in  2000-­‐2012  Latin  American  exports  to  China,   by  type   Notes:  Commodities  are  defined  using  SITC  Rev.3.  Soybeans  and  other  seeds  corresponde  to  222,   iron  ore  and  concentrates  to  281,  and  crude  petroleum  to  333.  Source:  UN  Comtrade  (2013).     3. China  and  deforestation  in  the  state  of  Mato  Grosso Some  of  the  main  drivers  of  Amazon  deforestation  include  roads,  agribusiness   development  (cattle  pasture,  soybean  production,  logging  and  agro-­‐industrial   expansion),  fire  and  mining  (Brown,  2004;  Fearnside,  2005a,  2008a).  Hargrave  &   Kis-­‐Katos  (2011)  analyzed  the  economic  causes  of  deforestation  in  the  Brazilian   Amazon  with  a  regression  method.  They  investigated  the  deforested  area  using  the   theoretical  model  of  Angelsen  (1999),  who  argued  that  deforestation  can  be   explained  by  the  expected  profits  from  land  use,  but  that  liberalization  and   macroeconomic  issues  may  also  be  relevant.  Angelsen  and  Kaimowitz  (1999)   showed  that  agricultural  and  forestry  exports  may  lead  to  more  deforestation.  In   general,  the  literature  points  to  such  deforestation  causes  as  the  area  of  soybeans   harvested,  the  area  of  cattle  pastures,  the  prices  of  these  commodities,  roads,   7   population  density,  cattle  herd  size,  geographical  variables  related  to  climate  and   soil,  rural  credit  policies  and  economic  growth  (Morton  et  al.,  2006;  Barona  et  al.,   2010;  Martins  &  Pereira,  2012).  Hargrave  and  Kis-­‐Katos  (2011)  estimated   deforestation  as  a  function  of  these  possible  causes  for  each  municipality  (county)  in   the  Legal  Amazon  (this  5-­‐million  km2  administrative  region  includes  the  states  of   Acre,  Amapá,  Amazonas,  Maranhão,  Mato  Grosso,  Pará,  Rondônia,  Roraima  and   Tocantins).  They  show  soybean  prices  and  environmental  fines  to  be  factors   influencing  deforestation.  Their  findings,  however,  did  not  explicitly  take  into   account  the  exports  to  China  as  the  main  consumer  of  Amazonian  soybeans.  An   econometric  study  of  deforestation  in  Mato  Grosso  using  municipality-­‐level  data   from  2001  to  2010  shows  the  strong  role  of  soy,  beef  and  wood  exports  (Moreira,   2013).  Moreira  (2013)  did  not  separate  exports  to  China  from  those  bound  for  other   destinations,  but  as  the  largest  purchaser  of  these  commodities,  particularly  for   soybeans,  China’s  role  is  clear.  Even  before  the  recent  surge  of  exports  to  China,   Nepstad  et  al.  (2006)  argued  that  China's  demand  for  Brazilian  soybeans  has   stimulated  production  and  also  deforestation  in  Mato  Grosso,  the  biggest  Brazilian   producer.     We  concentrate  our  attention  on  Mato  Grosso,  where  the  effect  of  soybeans  is   paramount,  rather  than  timber,  beef  and  minerals  (which  are  also  exported  to   China).  Soybeans  are  a  major  force  in  clearing  of  the  cerrado  (central-­‐Brazilian   savanna)  and  in  Amazonian  deforestation  both  in  areas  that  are  climatically  and   topographically  appropriate  for  planting  soy  and  in  areas  that  are  not  good  for  soy   but  where  deforesters  gain  access  using  soy-­‐related  transportation  infrastructure   (Fearnside,  2001a,  2007).  The  expansion  of  soy  into  pasture  areas  in  Mato  Grosso   also  leads  to  displaced  deforestation  for  cattle  in  Pará,  the  state  bordering  Mato   Grosso  to  the  north  (Arima  et  al.,  2011).  In  addition  to  its  impact  on  deforestation,   the  migration  of  ranchers  to  Pará  and  the  expansion  of  pasture  there  can  exacerbate   land-­‐tenure  conflicts,  usually  at  the  expense  of  small  farmers  and  traditional  peoples   (e.g.,  Fearnside,  2001b).   8   The  rapid  rise  in  exports  of  products  such  soy  and  beef  to  China  have  consequences   for  Amazonian  deforestation  that,  while  they  may  appear  obvious,  are  nevertheless   complex  to  quantify  and  interpret.  This  direct  impact  of  commodity  exports  is  only   the  tip  of  the  iceberg  of  Chinese  influence  on  Amazonia.  Money  earned  from  this   trade  is  strengthening  Brazilian  agribusiness  interests,  with  profound  effects  on   domestic  politics  that  are  reflected  in  legislative  and  administrative  changes   weakening  environmental  protection.  Impacts  can  also  be  expected  from  Chinese   financing  under  negotiation  for  infrastructure  such  as  a  railway  linking  the  state  of   Mato  Grosso  to  a  port  on  the  Amazon  River.  Mato  Grosso  is  a  major  focus  of   expansion  of  soy,  cotton  and  intensified  cattle  production.  Chinese  purchases  of  land   for  agriculture  and  timber  imply  an  increasing  direct  role  in  commodity  production.   Other  impacts  come  from  exports  from  mining  and  from  the  processing  of  minerals,   especially  the  demands  for  charcoal  for  pig-­‐iron  smelters  and  for  electricity  from   hydroelectric  dams  for  aluminum  smelters.     4. Brazil’s  exports  to  China 4.1  Interpreting  export  data   Data  on  exports  by  product,  origin  and  destination  (for  quantities  as  well  as  value  in   US  dollars)  are  provided  by  the  Aliceweb  system  (Online  Information  System  for   Analysis  of  Foreign  Trade)  of  the  Brazilian  Ministry  of  Development,  Industry  and   Foreign  Trade  (MDIC).  Chinese  demand  for  Brazilian  exports  reached  approximately   US$30  billion  in  2010,  having  increased  tremendously  since  2000,  with  annual   geometric  rates  from  2000  to  2010  of  about  44%  for  non-­‐agricultural  products  and   34%  for  agricultural  products  in  values  FOB3.  Exports  from  the  Legal  Amazon  to   China  have  also  increased  at  an  impressive  rate  of  52%  yearly  from  2000  to  2010,   3  “Free  on  board,”  or  the  value  at  the  port  of  shipment  net  of  all  domestic  transportation  and  loading   costs   9   increasing  from  US$104  million  in  2000  to  US$6631  million  in  2010  (research  data   based  on  Brazil,  MDIC,  2012).   Exports  to  China  from  the  Brazilian  Legal  Amazon  rose  from  an  average  of  13%  of   total  Brazilian  exports  to  China  in  the  2000-­‐2004  period  to  an  average  of  23%  in  the   2005-­‐2010  period.  The  change  occurred  abruptly  from  14.6%  in  2004  to  23.3%  in   2005  and  then  kept  stable.  Almost  half  of  the  change  in  the  2004-­‐2005  period  can  be   attributed  to  FOB  price  increases,  but  for  the  2009-­‐2010  period  (when  FOB  prices   were  lower)  the  physical  quantities  explain  most  of  the  increase  in  value  (research   data  based  on  Brazil,  MDIC,  2012).   An  interesting  point  to  observe  is  that  the  share  of  export  value  represented  by  a   soybean  composite  (soy  grain,  soy  oil  and  other  vegetable  oil)  decreased  from  a  high   of  66.3%  in  2003  to  34.4%  in  2010,  with  an  average  of  47.7%  over  the  last  five   years.  Iron  and  manganese  now  represent  a  large  share  of  Brazilian  exports  to   China,  reaching  63.1%  in  2010  (49.8%  as  a  5-­‐year  average).  In  2010  about  83%  of   the  Legal  Amazon’s  exports  were  from  the  states  of  Pará  (52%  -­‐  iron)  and  Mato   Grosso  (31%  -­‐  soybeans).  The  share  of  iron  plus  soybeans  in  these  two  states   decreased  from  79%  in  2006  to  72%  in  2008,  but  then  rebounded  to  high  levels:   89%  in  2009  and  83%  in  2010  (research  data  based  on  Brazil,  MDIC,  2012).   Figure  5  graphs  the  exports  of  soybeans  from  Mato  Grosso  and  of  iron  from  Pará.   The  impressive  increase  is  apparent.  The  value  of  iron  from  Pará  and  of  soybeans   from  Mato  Grosso  totaled  US$5.5  billion  in  2010  (research  data  based  on  Brazil,   MDIC,  2012).   10   Figure  5.  Value  of  exports,  soybeans  from  Mato  Grosso  and  iron  from  Pará   Note:  Values  are  shown  in  USD,  on  an  FOB  basis.  Source:  Brazil,  MDIC  (2013).   4.2  Interpreting  deforestation  data   Data  on  the  area  planted  in  soybeans  and  the  size  of  the  cattle  herd  were  obtained   from  IBGE  (Brazilian  Institute  of  Geography  and  Statistics).  Data  on  area  deforested   in  the  Brazilian  Amazon  are  available  from  the  Prodes  project  of  the  National   Institute  for  Space  Research  (Brazil,  INPE,  2013).  The  database  reports  deforested   area  in  hectares  at  the  municipal  level  for  2000-­‐2010.  Note  that  these  data  only   report  the  clearing  of  forest,  not  the  clearing  of  cerrado  (central  Brazilian  savanna)   that  represents  much  of  the  soybean  area  in  the  state  of  Mato  Grosso.  We  note  that   in  the  past  there  have  been  significant  problems  with  Brazilian  deforestation  data   (e.g.,  Fearnside,  1997a),  but  that  transparency  and  reporting  have  been  greatly   improved  for  the  years  covered  in  our  quantitative  analysis.   Looking  at  deforestation  in  the  Legal  Amazon  and  comparing  it  to  exports  to  China,   to  soybean  planted  area  and  to  cattle  herd  size,  Figure  6  exhibits  an  interesting   behavior.  In  Figure  6,  based  on  Brazil,  INPE  (2013)  deforestation  data  and  on  Brazil,   IBGE  (2012)  agricultural  data,  the  cumulative  area  may  give  a  misleading   interpretation  because  this  area  is  not  the  increase  in  deforestation.  Cumulative   11   deforested  area  clearly  has  a  positive  relationship  with  area  of  soybeans,  size  of  the   cattle  herd  and  value  of  exports,  but  for  each  year  there  is  a  different  value  for   exports  and  soybean  area,  while  deforested  area  is  a  cumulative  value.  One  point  to   be  observed  is  that  recent  expansion  of  soybeans  in  Mato  Grosso  is  taking  place  in   preexisting  pastures  that  represent  areas  deforested  at  some  time  in  the  past.   Advance  of  soybeans  into  pasture  areas  in  Mato  Grosso  has  long  been  believed  to   displace  ranching  activity  into  forest  areas  such  as  those  in  Pará,  contributing  to   deforestation  there  (Fearnside,  2001a;  Fargione  et  al.,  2008).  Recently  this  effect  has   been  demonstrated  statistically  (Arima  et  al.,  2011).  Brazilian  diplomats  currently   deny  this  effect,  and  in  March  2014  they  were  successful  in  getting  mention  of  it   deleted  from  the  summary  for  policy  makers  of  the  Fifth  Assessment  Report  of  the   Intergovernmental  Panel  on  Climate  Change  (Garcia,  2014).   Figure  6.  Cumulative  area  of  deforestation  (km²),  Soybean  planted  area  (ha),   Cattle  herd  (head)  in  the  Brazilian  Legal  Amazon,  2000-­‐2010.     Sources:  Brazil,  INPE  (2013);  Brazil,  MDIC  (2013).  Data  normalized  to  year  2000  =  100.   In  Figure  7,  increase  in  deforested  area  is  depicted  for  each  state  in  the  Legal   Amazon  using  Brazil,  INPE  (2013)  data.  It  can  be  observed  that  deforestation  rate   has  declined  since  2001,  staying  below  10,000  km²  in  2009-­‐2010.  The  state  of  Pará   had  the  largest  increase  in  deforested  area  (almost  3400  km²),  or  52.6%  of  the  total   12   increase  in  2010.  Mato  Grosso  was  an  important  state  up  to  2008  with  a  rapid   expansion  in  agriculture  and  cattle  raising,  but  there  was  a  significant  reduction  in   2009-­‐2010:  Mato  Grosso  accounted  for  only  12%  of  the  total  increase  in  2010.     Figure  7.  Annual  increase  in  deforested  area  (km²)  in  each  state  in  the   Brazilian  Legal  Amazon,  2001-­‐2010.     Source:  Brazil,  INPE  (2013).   Data  from  Brazil,  INPE  (2013)  and  Brazil,  MDIC  (2012)  exhibit  a  strict  positive   relationship  between  cumulative  deforestation  and  the  Legal  Amazon’s  exports,   with  a  high  correlation  (0.82).  A  different  pattern  is  seen  if  increase  in  deforestation   is  plotted  against  the  value  of  exports.  This  shows  an  inverted  relationship  with  a   negative  correlation  of  -­‐0.72  with  total  exports  (Figure  8).     13   Figure  8.  Increase  in  deforested  area  (km²)  in  the  Brazilian  Legal  Amazon   against  the  value  of  exports  to  China,  2001-­‐2010.     Sources:  Brazil,  INPE  (2013);  Brazil,  MDIC  (2013).   Looking  at  the  data  in  this  way,  exports  to  China  are  not  a  driving  force  of   deforestation  directly,  and  more  detailed  observation  is  needed.  Soy  is  a  major   product  in  total  exports  to  China,  but  what  is  its  relationship  to  the  increase  in   deforestation?  Figure  6  indicates  a  pattern  associating  cumulative  deforestation   with  soybean  area  and  cattle  herd  size.  Figure  8  shows  the  increase  in  deforestation   against  these  two  variables  using  Brazil,  INPE  (2013)  and  Brazil,  IBGE  (2012)  data.   14   Figure  9.  Increase  in  deforested  area  (km²)  in  the  Brazilian  Legal  Amazon   against  the  difference  in  soybean  planted  area  and  in  cattle  herd  size,  2001-­‐ 2010.     Sources:  Brazil,  IBGE  (2012);  Brazil,  MDIC  (2013).   According  to  Figure  9,  the  increase  in  deforested  area  is  accompanied  by  increases   in  either  the  cattle  herd  or  the  soybean  area.  However,  the  regression  line  had  non-­‐ significant  parameters  and  the  correlations  with  the  increase  in  deforestation  were   only  0.3  for  soybeans  and  0.5  for  cattle.  On  the  other  hand,  there  is  a  strong  positive   correlation  (0.85)  between  the  differences  in  soybean  area  and  in  cattle  herd  size.   The  same  pattern  was  found  by  Marta  and  Figueiredo  (2008).  A  regression  was   performed  for  the  entire  Legal  Amazon,  aggregating  the  data  from  the  nine  states  in   the  region.  The  deforestation  process  is  quite  complex,  leading  us  to  use  a  systems   approach4.  This  approach  allowed  us  to  relate  explanatory  variables  to  both  a   deforestation  equation  and  an  exports  equation.  Here  we  will  discuss  some  results   of  this  system.   4  The  details  for  the  system  of  equations  as  well  as  its  results  are  provided  in  the  Technical  Appendix.   15   Deforestation  rates  throughout  Brazilian  Amazonia  have  declined  substantially   since  2004,  with  a  statistically  significant  negative  trend.  The  decline  through  2008   is  explained  by  falling  international  prices  of  soy  and  beef  in  the  2004-­‐2006  period,   together  with  a  worsening  exchange  rate  of  the  Brazilian  real  against  other   currencies  from  the  point  of  view  of  exporters,  but  deforestation  continued  to   decline  after  2008  despite  a  recovery  in  commodity  prices  (Assunção  et  al.,  2012).     Figure  10  shows  the  relation  of  exports  to  soybeans  and  cattle  according  to  Brazil,   MDIC  (2012)  and  Brazil,  IBGE  (2012)  data.  The  two  dispersion  plots  in  Figure  8   present  essentially  the  same  information:  a  positive  correlation  between  the  exports   to  China  and  the  sizes  of  the  herd  and  the  crop.  The  correlation  is  0.69  for  soybean   area  and  0.67  for  cattle  herd.     Figure  10.  Total  exports  from  the  Brazilian  Legal  Amazon  to  China  against  the   soybean  planted  area  and  cattle  herd,  2001-­‐2010.     Sources:  Brazil,  IBGE  (2012);  Brazil,  MDIC  (2013).     Note:  Total  exports  from  Legal  Amazon  to  China  in  FOB  US$  billion.  Soybean  planted  area  in  10³   hectares.  Cattle  herd  in  millions  of  head.   The  idea  is  that  soybeans  are  planted  in  already-­‐existing  cattle  pastures  (Marta  &   Figueiredo,  2008).  The  soybeans  are  then  exported  and  there  is  a  kind  of  learning   process  where  exports  lead  to  an  effect  on  the  next  period’s  exports.  The  preceding   period’s  exports  and  cattle  herd  size,  as  well  as  current  soybean  planted  area,   16   soybean  price  and  beef  price  explained  the  change  in  deforestation  in  the  current   period.     On  the  other  side,  the  soybean  plantations  are  occupying  previous  pastures.  The   persistence  of  cattle  herds  occupying  land  is  a  restriction  on  soybean  cultivation  and   exports.  The  land-­‐use  shift  from  old  degraded  pasture  into  soybean  cropping  can,  in   some  way,  have  a  positive  effect  on  exports.  The  problem  is  that  expansion  in   soybean  area  leads  to  increase  in  deforestation.  Increasing  cattle  herd  size  also  has  a   statistical  impact  on  the  increase  in  deforestation,  similar  to  other  studies  that  have   found  a  strong  effect  of  cattle  herd  size  on  deforestation  (e.g.,  Alencar  et  al.,  2004;   Kaimowitz  et  al.,  2004;  Arima  et  al.,  2005).  Nevertheless,  increasing  the  herd  size   appeared  to  reduce  exports  to  China  and  this  reduction  would  lead  to  a  decrease  in   deforestation.     Two  factors  can  help  explain  the  relationship  between  cattle  and  deforestation  in   Mato  Grosso.  First,  is  the  recent  spread  of  techniques  for  improving  pasture   productivity  (or  pasture  stocking),  particularly  in  Mato  Grosso,  as  well  as  the   growth  of  feedlots.  Feedlot  holding  capacity  in  Mato  Grosso  grew  from  668,000  head   in  2009  to  883,000  in  2014;  interviews  with  ranchers  conducted  by  the  Mato  Grosso   state  government  indicate  a  dramatic  surge  of  intention  to  hold  cattle  in  feedlots  in   the  northern  part  of  the  state,  where  plans  for  this  practice  doubled  from  2013  to   2014  (IMEA,  2014).  These  feedlot  meats  are  mainly  for  high-­‐quality  beef  exports.   Although  our  analysis  is  at  the  level  of  the  Legal  Amazon  for  all  variables,  the  effect   of  intensification  would  probably  be  substantially  less  important  in  other   Amazonian  states  such  as  Pará.     A  second  factor  is  that  China  only  recently  permitted  the  importation  of  Brazilian   beef.  Until  that  point,  the  increase  in  herds  represented  a  restriction  on  products   that  could  be  exported  to  China,  namely  soy,  leading  to  more  deforestation  for   pasture.  For  example,  reducing  soybean  exports  would  lead  to  pasture  and  herd   increases.  Prices  of  soybeans  and  beef  also  impact  deforestation  rates,  with  soybean   17   price  as  a  direct  effect  and  beef  price  having  an  indirect  negative  effect  on   deforestation  rates.  Now  that  China  has  authorized  imports  of  Brazilian  beef,  it  is   expected  to  exert  pressure  on  the  Brazilian  beef  market,  competing  with  the   Brazilian  soybean  exports.   Even  though  states  like  Mato  Grosso  and  Pará  exhibited  increases  in  cattle  herd  size   after  2010,  the  herd  in  2010  was  about  the  same  size  as  in  2005  for  Pará,  and  Mato   Grosso.  The  herd  remained  constant  over  the  2005-­‐2008  period  and  had   remarkable  increases  of  more  than  5%  annually  in  2009  and  2010.  Mato  Grosso  had   small  increases  in  deforestation  in  2009  and  2010,  although  it  was  one  of  the  states   with  the  largest  increases  in  deforestation  between  2000  and  2008,  clearing  land  for   herd  expansion  in  the  following  years.     The  relations  of  soybeans  and  cattle  to  deforestation  over  the  2000-­‐2010  period   were  complicated  by  other  factors  that  influenced  the  deforestation  process  to   differing  degrees  in  each  year.  The  efforts  of  environmental  authorities  to  control   illegal  deforestation  through  inspections  and  fines  have  varied  substantially   (Nepstad  et  al.,  2014).  Increased  efforts  to  control  deforestation  appear  to  have  had   a  significant  effect  from  2008  onwards,  whereas  before  2008  deforestation  rates   track  soy  and  beef  prices  closely  (Barreto  et  al.,  2011;  Assunção  et  al.,  2012).  In   addition,  the  periods  immediately  preceding  elections  are  normally  characterized  by   deforestation  increases  both  as  a  result  of  political  pressure  to  relax  environmental   enforcement  (especially  at  the  state  level)  and  as  a  result  of  anticipation  by   deforesters  that  election  results  will  bring  relaxed  enforcement  and/or  amnesties   forgiving  past  violations  (see  Fearnside,  2003).   A  key  change  that  occurred  in  2008  was  a  decision  by  Brazil’s  Central  Bank  that  no   public  bank  loans  could  be  given  to  operations  with  environmental  irregularities   reported  by  agencies  such  as  IBAMA  (BACEN  Resolution  3.545/2008).  Unlike  fines   given  by  IBAMA  and  other  agencies,  which  can  be  circumvented  with  a  seemingly   endless  sequence  of  appeals,  the  restriction  on  bank  loans  has  real  teeth  and  an   18   immediate  effect.  The  credit  restriction  greatly  increases  the  impact  of  the   environmental  inspection  programs,  even  if  the  programs  themselves  do  not  change   substantially  in  scale  and  even  in  the  face  of  the  inability  of  agencies  like  IBAMA  to   collect  most  of  the  fines.  This  recent  change  in  bank  policy  was  not  included  in  our   model,  and  its  effect  remains  as  a  suggestion  for  further  study.  Increased  exports   were  possible  at  a  time  of  decreasing  deforestation  due  to  rising  per-­‐hectare  soy   yields,  clearing  in  non-­‐forest  vegetation  types  (i.e.,  cerrado),  and  soy  expansion  in   former  cattle  pastures  (the  indirect  effects  of  which  would  be  displaced  beyond  the   borders  of  Mato  Grosso  to  increase  pasture  in  Pará).   4.3  Other  commodities  exported  from  Brazilian  Amazonia   4.3.1  Timber   China  has  cut  almost  all  of  its  natural  forests  and,  despite  large-­‐scale  plantations  of   fast-­‐growing  trees,  the  country  has  a  tremendous  demand  for  timber  such  as  that   from  Brazil’s  Amazon  forest.  Unlike  European  and  North-­‐American  markets,  China   is  willing  to  buy  wood  from  almost  any  species  of  tropical  tree.  An  example  of  this   occurred  when  timber  was  sold  prior  to  the  1988  flooding  the  Samuel  Dam  in  the   state  of  Rondônia  (Fearnside.  2005b).  Export  of  raw  logs  from  Brazil  has  been   prohibited  since  1965,  but  an  exception  was  opened  to  allow  logs  from  Samuel  to  be   exported  (Nogueira,  1988).  From  1987  through  1989,  a  continuous  chain  of  barges   arrived  in  the  Amazon  River  port  of  Itacoatiara  with  logs  for  loading  on  ships,  and   one  ship  loaded  with  logs  departed  for  China  every  two  weeks  during  this  period.   The  exception  opened  for  the  relatively  small  Samuel  reservoir  area  had  allowed   logs  to  be  illegally  exported  from  vast  areas  in  western  Amazonia.   Chinese  companies  purchased  several  bankrupt  sawmills  in  Manaus  in  1996,   thereby  gaining  the  forestland  holdings  of  the  sawmill  companies.  Together  with   Malaysia,  land  purchases  totaled  4.5  million  hectares  in  the  state  of  Amazonas   (Amazonas  em  Tempo,  1996).  Most  of  the  forestland  bought  by  Chinese  companies   19   was  in  the  municipality  of  Carauarí.  A  major  increase  in  logging  activity  was   expected  at  the  time,  but  this  did  not  occur  (presumably  due  to  the  substantial   bureaucratic  barriers  to  obtaining  approval  of  forestry  management  plans).  Other   countries  have  been  satisfying  most  of  the  world’s  demand  for  tropical  timber,   including  the  demands  of  China.  However,  Brazil  has  by  far  the  largest  stock  of   remaining  tropical  forest,  and  the  pressure  of  this  demand  is  bound  to  focus  on   Brazil  once  available  stocks  elsewhere  are  exhausted  (Fearnside,  1989a).   4.3.2  Alumina,  aluminum  and  iron   Chinese  companies  have  interests  in  alumina  (Al2O3:  the  precursor  of  primary   aluminum)  in  Barcarena,  Pará,  where  Alumina  Brasil-­‐China  (ABC)  and  Aluminum   Corporation  of  China  Limited  (Chalco)  have  a  joint  venture  with  the  Brazilian   mining  company  Vale  (Vale,  2009).  The  power  demand  for  this  electricity-­‐intensive   industry  contributes  to  Brazil’s  push  for  a  massive  increase  in  building  hydroelectric   dams  in  Amazonia  over  the  next  decade.  Brazil’s  2011-­‐2020  ten-­‐year  energy-­‐ expansion  plan  (Brazil,  MME,  2011)  calls  for  30  large  dams  to  be  built  in  the  Legal   Amazon  by  2020,  a  rate  of  one  dam  every  four  months.  The  Chinese-­‐Brazilian   alumina  plant  will  be  an  important  beneficiary  of  the  Belo  Monte  Dam,  now  under   construction  on  the  Xingu  River,  near  Altamira,  Pará.  Belo  Monte  has  environmental   and  social  impacts  that  extend  far  beyond  the  areas  that  will  be  directly  flooded,  and   the  dam  is  likely  to  justify  much  larger  upstream  reservoirs  to  regulate  the  river’s   flow  (Fearnside,  2006).  The  dam  has  functioned  as  a  “spearhead”  in  creating   precedents  that  weaken  Brazil’s  environmental  licensing  system  and  prepare  the   way  for  the  many  dams  proposed  under  the  energy-­‐expansion  plan  (Fearnside,   2012).  In  February  2014  a  consortium  led  by  Chinese  State  Grid  won  the  bidding  for   the  R$5  billion  (US$2  billion)  contract  to  build  the  transmission  line  for  Belo  Monte.   The  expansion  of  Amazonian  dams  also  receives  a  boost  from  China’s  equipment   sales,  as  in  the  case  of  the  turbines  from  Dong  Fang  Electric  Corporation   International  and  Dong  Fang  Electric  Machinery  for  the  Jirau  Dam  now  under   20   construction  on  the  Madeira  River.  The  influence  of  both  Brazil  and  China  in   expanding  carbon  credit  for  hydroelectric  projects  under  the  Kyoto  Protocol’s  Clean   Development  Mechanism  has  further  increased  the  profitability  of  dams  (Fearnside,   2013a,b).     Iron  from  Brazil  is  now  largely  exported  to  China  (Soares,  2012)  (and  previously   shown  in  the  introduction  of  this  chapter).  The  Chinese  market  has  eclipsed  the   European  purchasers  that  dominated  exports  from  the  Carajás  Mine,  in  Pará,  when   the  mine  was  opened  in  the  1980s.  Processing  of  part  of  the  ore  for  export  as  pig-­‐ iron  consumes  charcoal,  providing  a  longstanding  source  of  pressure  on  the  forests   of  eastern  Amazonia  and  a  challenge  to  environmental  and  labor  authorities   (Fearnside,  1989b).  The  environmental  and  social  impacts  of  charcoal  production   made  European  iron  imports  from  Amazonia  a  target  for  criticism  from  non-­‐ governmental  organizations  (e.g.,  Sutton,  1994),  but  this  is  no  longer  evident  now   that  exports  have  shifted  to  China.  Brazil’s  export  of  iron  alloys,  although   representing  far  smaller  quantities  than  iron  in  the  form  of  ore  or  pig-­‐iron,  is  the   fastest-­‐growing  category  of  exports  to  China,  tripling  from  4000  to  12,000  tons   between  2013  and  2014  (CEBC,  2014).  Producing  iron  alloys  consumes  a   tremendous  amount  of  electricity  and  creates  a  miniscule  amount  of  employment  in   Brazil:  1.1  jobs  per  GWh  of  electricity  consumed,  or  even  less  than  primary   aluminum,  which  creates  only  2.7  jobs  per  GWh  (Bermann  &  Martins,  2000,  p.  90).   5. China  and  political  shifts  in  Brazil The  political  influence  of  the  “ruralist”  voting  block  that  represents  large   landholders  in  Brazil’s  National  Congress  is  has  increased  markedly  due  to  the  large   amounts  of  money  entering  Brazil  from  the  export  of  soybeans,  China  being  the   number  one  source  of  these  earnings.  The  shift  of  Brazil’s  economy  towards   agricultural  commodity  exports  (strengthening  the  influence  of  large  landowners)   and  away  from  manufacturing  (weakening  the  influence  of  industrialists  and  labor   21   unions)  is  affecting  virtually  every  aspect  of  Brazil’s  politics  (e.g.,  ISA,  2014).  Effects   include  the  positions  of  the  current  presidential  administration  on  environmental   issues  (Santilli,  2014;  Smeraldi,  2014).  The  “ruralist”  block  is  currently  trying  to   revert  the  Brazilian  Central  Bank’s  resolution  blocking  loans  for  agriculture  and   ranching  from  the  Banco  do  Brasil  (BB),  Caixa  Econômica  Federal  (CEF)  and  Banco   da  Amazônia  (BASA)  for  properties  with  pending  fines  for  environmental  violations.   China’s  influence  on  this  transition  extends  beyond  the  boost  to  “ruralist”  influence   from  soy  income:  China’s  exports  of  cheap  manufactured  goods  to  Brazilian   manufacturers’  former  export  markets  has  cut  deeply  into  Brazil’s  exports  from  this   sector,  and  China’s  direct  export  of  manufactured  goods  to  Brazil  further  displaces   Brazilian  manufacturing  and  reduces  the  political  influence  of  this  sector  within   Brazil.  Brazil  has  maintained  an  approximate  balance  in  terms  of  monetary  value   between  exports  to  and  imports  from  China  (Figure  11).  This  balance  may  be   influenced  by  trade  negotiations  between  the  two  countries  in  which  the  Chinese   interest  in  maximizing  its  exports  could  help  explain  the  close  parallel  between  the   increasing  monetary  flows  in  the  two  directions.  Unlike  countries  that  have  little   domestic  manufacturing  to  lose,  the  effect  in  Brazil  is  significant.  The  increasing   exports  of  agricultural  commodities  and  imports  of  manufactured  goods  contribute   to  a  shift  in  political  influence  in  Brazil  from  the  manufacturing  to  the  agribusiness   sectors,  with  consequences  for  environmental  policies.   22   Figure  11.  Brazil’s  trade  balance  with  China   Source:  Authors’  calculations  based  on  UN  COMTRADE  data.   6. Financing  from  China Another  issue  is  the  effect  of  Chinese  finance  in  Latin  America.  China  is  a  new  and   growing  source  of  funding  with  less  environmental  restrictions,  lower  interest  rates   and  different  size  loans  (Gallagher  et  al.,  2012).     Official  data  regarding  the  foreign  direct  investment  (FDI)  in  Brazil  are  found  in   Brazil  BCB  (2014),  in  the  census  from  the  Brazilian  Central  Bank.  The  Chinese  stock   of  FDI  in  Brazil  (in  capital  shares  and  immediate  investor)  changed  from  US$582   million  in  2010  to  US$  1,093  million  in  2012.  Looking  as  final  investor,  who  occupies   the  top  of  the  control  chain,  this  FDI  stock  in  capital  shares  went  from  US$  7,874   million  to  US$  10,226  million  in  the  same  period  (Brazil´s  total  FDI  stock  is  US$   617,384  million  for  all  countries).  This  2012  value  is  divided  by  sector  as:  extractive   Industries  (82.3%);  manufacturing  (1.3%);  trade  and  repair  of  vehicles  (2.6%);  and   others  (13.9%).     23   Most  Chinese  loans  are  for  the  oil,  iron,  steel,  energy  and  telecom  sectors.  From   2007  to  2012,  the  China  Brazil  Business  Council  (CBBC)  “recorded  a  total  of  60   announced  Chinese  investment  projects  for  a  total  of  US$  68.5  billion”  (CBBC,  2013),   of  which  54  were  in  the  2010-­‐2012  period  and  47  were  partially  or  completely   state-­‐funded.  Regarding  the  investment  motivation,  57%  were  resource  seeking,  but   lately,  from  2011-­‐2012,  market  oriented  investments  dominated.  Chinese   investment  projects  are  distributed  amongst  14  Brazilian  sectors,  as  shown  in   Figure  12.   Figure  12.  Chinese  investment  projects  by  sector  (number  of  projects),  2007   to  June  2012.     Source:  CBBC  (2013,  p.  14).   The  Chinese  investment  is  highly  concentrated  in  the  three  states  of  the  Southeast   macro-­‐region:  São  Paulo  (27%  -­‐  Banking,  Telecom,  Automotive  and  Electronics),   Minas  Gerais  (18%  -­‐  Machinery  and  Equipment)  and  Rio  de  Janeiro  (17%  -­‐  electrical   energy,  oil  and  gas)  (Figure  13).  Mato  Grosso  and  Amazonas  each  accounts  for  5%.   24