McKinsey & Company has advised many forest nations on developing REDD+ plans, but its methodology has fundamental flaws. McKinsey's cost curve systematically downplays environmental impacts of deforestation and fails to address its true drivers. As a result, McKinsey-inspired plans could increase deforestation by promoting industrial activities like logging and plantations. Donor countries should scrutinize McKinsey's influence and ensure REDD+ plans protect forests and respect indigenous rights.
2. ii
Contents
!"#$%&'(#)*%++,-.
/0#)*&1-.)12)3!445) 6
!7&#-)8$9'7*#.)) :
;0,&)'*)8$9'7*#.),7<)=01)<1#*)'&)=1->)21-?) @
3!445A)8$9'7*#.B*)'72C%#7$#)'7)>#.)$1%7&-'#*)) D
Democratic Republic of the Congo (DRC) 6
Guyana 6
Papua New Guinea (PNG) 7
Indonesia 7
E721F-,G0'$A)8$9'7*#.)H)3!445)
McKinsey’s bad influence on national plans to reduce
forest emissions from deforestation and degradation I
/0#)8$9'7*#.)8JK)$%-(#A),7)1G&'$,C)'CC%*'17?)) LL
E721F-,G0'$A)8$9'7*#.B*))
+,-F'7,C),M,&#+#7&)$1*&)$%-(#)) L6
;0,&B*)=-17F)='&0)8$9'7*#.B*)+#&01<?) LN
LO)P1-#*&)$,-M17A)'7,-&'$%C,&#)1-)'7#G&?) LN
a. Carbon stocks and flows in plantations 15
b. Unrealistic precision 15
6O)4,&,)<#2'$'#7$'#*A)'7,<#Q%,&#)1-),M*#7&?) LD
Plantations: root and branch confusion 16
:O)R,*#C'7#)$,C$%C,&'17*A)+,7'G%C,&'7F),**%+G&'17*) LS
@O)P,(1%-'7F)'7<%*&-',C)'7&#-#*&*A)*>#=#<)G#-*G#$&'(#*) LI
Can McKinsey add up? 19
NO)817'&1-'7F),7<)$,G,$'&.A)%7=,--,7&#<)1G&'+'*+) LT
/0#)'+G,$&)12)8$9'7*#.B*),<('$#) 6U
LO)P,'C'7F)&1),<<-#**)&0#)-#,C)<-'(#-*)12)<#21-#*&,&'17)) 6U
6O)V,.'7F)&0#)G1CC%&#-) 6U
:O)E+G,$&'7F)17)7,&%-,C)21-#*&*)) 6L
P1-#*&*A)+1-#)&0,7)W%*&)$,-M17) 66
LO)/0-#,'7F)7,&%-,C)0,M'&,&*),7<)='C<C'2#)) 6:)
6O)X1$',C)'+G,$&*) 6N
8$9'7*#.),<('$#A)%72'&)21-)G%-G1*#)) 6D
3#$1++#7<,&'17*)21-)8$9'7*#.Y)-,'721-#*&)
7,&'17*Y),7<)<171-)$1%7&-'#*),7<)'7*&'&%&'17*) 6I
!7<71&#*) :U
5. v
Executive
summary
Rainforest degradation is a major
contributor to climate change, being
responsible for up to one fifth of global
emissions. If climate change is to
be tackled, the urgent prevention of
deforestation is essential.
Reducing Emissions from Deforestation and Degradation or REDD McKinsey’s advice does not, in any example studied by
schemes are intended to provide tropical forest nations with Greenpeace, lead to a cessation of deforestation or forest
financial incentives not to destroy or degrade their forests. The degradation. Often it defends destruction by industrial interests
concept of REDD has subsequently been expanded to also include on the erroneous grounds that it contributes to economic growth.
financial support for restoration, reforestation and afforestation, In DRC, for example, McKinsey legitimises a significant increase
making for an expanded mechanism widely known as REDD+. in industrial logging, with an increase of at least an additional 10
million hectares given as logging concessions.
Despite the world’s general failure to agree a deal on climate
change at the 15th Conference of Parties to the United Nations While McKinsey’s cost curve has been extremely influential in
Framework Convention on Climate Change at Copenhagen in government policy decisions, it has a number of fundamental
2009, donor countries nevertheless did pledge about $3.5 billion flaws. These include data deficiencies and dubious baseline
to kickstart REDD+. One year later more progress was made calculations, as well as basic mathematical errors and distortions
with an agreement to establish REDD+ at COP16 in Cancun (‘the within McKinsey’s carbon accounting method. Furthermore,
Cancun Agreement’). Meanwhile, around the world, rainforest McKinsey’s intellectual property rights on some of the data
nations have been endeavoring to become ready for REDD+ by underpinning its cost curve prevent proper scrutiny of its rationale.
engaging in national planning around how the scheme would be
domestically implemented. McKinsey’s approach provides an incentive to over-estimate
projected future levels of deforestation, allowing forest nations
McKinsey & Company is a giant, well-connected global consultancy to claim REDD+ funding for preventing destruction which was
firm which has been working to position itself as the market leader unlikely ever to have happened.
in REDD+ advice. According to McKinsey:
McKinsey co-authored studies barely acknowledge governance
‘Our clients … look to us for honest, objective, thoughtful, and issues within rainforest nations, such as the sheer scale of
experienced advice.’1 monitoring, reporting and verification, capacity-building and
governance challenges . This casts further doubt on the value of
The McKinsey ‘Climate Desk’ has been very successful in becoming McKinsey advice.
known as a leading provider of these services. It has attracted
commissions to advise many forest nations on how to draw up national McKinsey-inspired plans not only consistently fail to address
plans for applying for REDD+ funding. It appears most probable that the major drivers of deforestation, such as mining and
the rainforest nations featured in this report are generally following logging, they actually reward the industries and interests
and implementing the advice provided by McKinsey. that cause it. For example, in the DRC study, the palm oil industry
stands to gain as much as 1bn for the ‘relocation’ of concessions
However, when rainforest countries employ McKinsey to apply its that have not even been awarded.
trademarked cost curve to their REDD+ prospects, few if any of the
resulting plans meet basic standards of accuracy, rigour, utility or ethical McKinsey promotes a methodology that effectively encourages
acceptability. If implemented in their current form, these plans could its client governments to pursue an industry-orientated
actually result in an increase of deforestation and carbon emissions. development path at whatever cost to wildlife. In Indonesia for
example, the forecast of continuing expansion of pulp and oil palm
This Greenpeace report presents case studies on McKinsey’s influence plantations is a major threat to biodiversity. McKinsey accepts that
on REDD+ plans for four forest nations – Papua New Guinea (PNG), the deforestation will shift to other, still largely forested, islands such as
Democratic Republic of Congo (DRC), Indonesia and Guyana. Our key parts of Kalimantan and especially Papua. Kalimantan is home to the
findings include that: endangered Bornean orangutan.
6. vi
If followed, McKinsey’s advice will lead to an
expansion of monoculture plantations into farmland
and ecologically important non-forest lands.
McKinsey misleadingly classifies these lands as ‘marginal’
to justify their conversion to plantations. This could have
devastating impacts on local ecosystems and wildlife.
McKinsey – and its cost curve – systematically play
down the environmental impact of industrial logging
and deforestation for plantations. At the same time,
it routinely exaggerates the destructive impact of
smallholders and farmers. This leads to plans that advocate
large-scale acquisition of local people’s lands or settling of
subsistence farmers without sufficient attention to their land
rights, prior informed consent and compensation.
McKinsey’s advice has produced plans which have
been criticised by funding institutions and are unfit for
purpose. When rainforest countries employ McKinsey to
apply its cost curve to their REDD+ prospects they are at
risk of wasting money on advice that may be in violation of
safeguards in the Cancun agreement on REDD+ and other
decisions of the UNFCC, UN CBD and other international and
regional institutions.
As a matter of urgency:
McKinsey must publish all the data, assumptions and analysis
underlying its cost curve, and not hide behind intellectual
property rights to avoid proper scrutiny. It should revise its
methodology to include:
associated with abatement options,
Cancun agreement: protecting forest ecosystems
from conversion and degradation, and recognising
and implementing indigenous peoples and local
communities’ rights.
Rainforest nations should not commission further work
from McKinsey until the above conditions have been met.
Those which have worked with McKinsey should revise the
resulting plans to address concerns outlined in this report,
and make public all advice received so far from the company.
Donor countries and institutions should not provide further
funding for McKinsey until the above conditions have been
met. They should only agree to fund the provision of REDD+
advice where all parties agree to a fully open and transparent
tendering process and there is full public disclosure of advice
and full participation of local communities
They should focus their attention and incentives on REDD
through natural forest protection, rather than any ‘+’
activities. REDD proper provides the greatest mitigation
and adaptation benefits. Policies should prioritise ending
deforestation where it currently occurs and preventing it
from increasing in areas at risk .
8. 2
The story
of REDD+
Tropical forests are home to a staggering array of plant, animal and Two years of intense negotiations between 2007 and 2009 saw
human communities. While covering only about 10% of the total rainforest nations and potential donors fight to see their particular
terrestrial surface, they are home to considerably more than 60% of preferred blueprint for REDD adopted. As a result, there was a
all terrestrial and freshwater biodiversity.2 1.6 billion people depend significant expansion of the ground to be covered by REDD, as it
upon forests for their survival.3 They also play a vital role in stabilising moved from being a mechanism for forest protection to a tool for
global weather patterns and their degradation is a major contributor also promoting restoration, reforestation and afforestation, under
to climate change, being responsible for up to one fifth of global the new acronym of REDD+.
emissions.4 If climate change is to be tackled, therefore, the urgent
prevention of further deforestation is essential. In parallel, the World Bank continued to play a central role in
the development of REDD policy making and implementation
For this reason, and in the absence of an effective global treaty to through the Forest Carbon Partnership Facility (FCPF) and Forest
fight deforestation, many people began to see it as imperative that Investment Programme (FIP). The FCPF aims to prepare forest
reduction of emissions from deforestation and degradation of forests countries for participation in international carbon markets by
was included within the aims of the United Nations Framework supporting REDD readiness activities. It became operational in June
Convention on Climate Change (UNFCCC). In essence, REDD 2008. Thirteen countries (Argentina, Costa Rica, the Democratic
(Reducing Emissions from Deforestation and Degradation schemes) Republic of Congo, Ghana, Guyana, Indonesia, Kenya, Lao PDR,
contemplates that funding be made available for the developing Mexico, Nepal, Panama, the Republic of Congo and Tanzania) have
world to reduce emissions from deforestation and degradation so far submitted Readiness Preparation Proposals (R-PPs) for the
of forests. Later REDD+ emerged as a variant, with the plus sign FCPF, setting out potential REDD+ policies and activities, which
signifying the potential for funding flows for active enhancement of have been reviewed by ad hoc Technical Advisory Panels and
carbon stocks through programmes of reforestation or afforestation. the Participants Committee. The World Bank is conducting due
The publication of a joint proposal by Papua New Guinea and Costa diligence on these proposals with a view to entering into readiness
Rica at COP11 of the UNFCCC in 2005 proved to be the turning grant agreements of up to $3.6 million to assist these countries in
point for international interest in establishing a REDD scheme and a conducting the preparatory work they have proposed.7 The Forest
two year review into the practicalities was initiated. REDD was then Investment Program (FIP) is a targeted programme of the Strategic
formally incorporated within the climate Road Map agreed at COP13 Climate Fund (SCF), which is one of two funds within the framework
of the UNFCCC in Bali in 2007.6 of the Climate Investment Funds (CIF). The FIP supports developing
countries’ efforts to reduce deforestation and forest degradation
Complex political and methodological challenges meant that the (REDD). The following have been selected to be pilot countries for
issue had still not been resolved as countries began to negotiate the FIP: Brazil, Burkina Faso, Democratic Republic of Congo, Ghana,
the shape of a new global climate deal in the run up to the ill-fated Indonesia, Laos, Mexico and Peru.8
Copenhagen Conference of 2009.
As negotiations moved forward, pledges of funding from donor
Nonetheless, momentum for REDD was growing: reducing emissions countries grew in size, with $3.5 billion9 agreed at the UNFCCC COP15
from deforestation and degradation of forests was commonly at Copenhagen in 2009. In many cases it was unclear how much of this
described as the low hanging fruit of the UNFCCC negotiations. REDD money was additional and whether it would be disbursed bilaterally or
would aim to provide tropical forest nations with a financial incentive through multilateral institutions and processes. While less than the $25
not to destroy or degrade their forests. Rainforest countries saw billion which some estimated was required up to 201510 (or a minimum
REDD as a means of accessing some global financial value for their commitment of $10 billion over the next three years that some NGOs
standing forests, whilst rich countries saw it as a relatively cheap were calling for11), this was still a considerable amount of money,
and politically acceptable measure to address climate change. As the particularly for rainforest countries struggling with poverty alleviation
financial crisis hit and then deepened in 2008/9, it became a mantra and development needs. It was certainly enough money to attract the
that REDD offered supposed win/wins all round. attention of at least one international consultancy firm.
11. 5
McKinsey has attained the position of market the world that will result in further business going
leader in REDD+ advice and forcefully advocates in the direction of The Firm.
its approach in the countries where it works.
According to journalist Clayton Hirst, McKinsey is: In the Democratic Republic of Congo (DRC)
McKinsey itself proposed as part of its output
‘…the ultimate old boys’ network. Its tentacles reach the inclusion of a list of key REDD+ measures
into the boardrooms of Britain’s biggest companies to be adopted.25 This element did not feature in
and snake through Westminster’s corridors of the terms of reference for the contract, but the
power…. The McKinsey mob just keeps growing. The resultant 14 programmes now form a key part of the
firm, of course, doesn’t use such crude terminology country’s Readiness Preparation Plan for REDD+.26
for its former partners; the “alumni network” is its In Indonesia, McKinsey advocated afforestation
preferred phrase. One source close to McKinsey right from its first presentation,27 and saw the
says: “The alumni are seen as ambassadors to the policy adopted by the National Council for Climate
McKinsey brand. The network isn’t openly exploited, Change.28 In the confidential proposal, ‘Institutional
but the firm maintains a database of members and capability building for low carbon growth’, prepared
holds an annual reception for the alumni.’22 by McKinsey for the Indonesian government there is
‘a heavy emphasis on coaching of local government
As one might expect, there is evidence of these officials, DNPI [Indonesian National Climate Change
business practices being applied in McKinsey’s Council] secondees and institutional partners’.29
work on REDD+. One individual involved in an
official capacity with a rainforest nation’s REDD+ According to one observer, McKinsey remained ‘highly
process has told Greenpeace on condition of influential’ within Papua New Guinea’s (PNG) Office
anonymity that McKinsey uses contacts in of Climate Change and Development (OCCD) as
one country as sales reps to help it get work in recently as November 2010, with questions raised
another, while boasting to potential developing at Technical Working Group meetings being ‘mostly
country clients of its capacity to connect them answered … by McKinsey representatives’.30 There is
with donors and taking full credit for funding even evidence that McKinsey may be responsible for
deals concluded (Guyana) or international mentoring staff in the recently established OCCD.31
influence attained (PNG) by its existing clients.23 The same observer notes that Sebastian Schienle, a
Once its foot is in the door, the company works McKinsey representative stationed permanently in
to maximise its influence. PNG, was even part of the PNG delegation in Cancun.
McKinsey states: ‘We take an overall, Yet at the same time, McKinsey plays down
independent, and fact-based view of a client’s responsibility for most of the documents on which
performance. We rely on facts because they it works. Despite extensive evidence for its having
provide clarity and align people. Facts are the played a major part in the studies considered in
global management language. We work with this briefing, it is invariably credited merely with
facts to provide credible recommendations.’24 providing data, analysis or technical support. That
The claim to strict objectivity is not reflected this is McKinsey’s decision is suggested by the
in McKinsey’s advice on REDD+ which is heavily comment in its DRC project proposal that it will not
reliant on a set of distinctly subjective policy publish a report under its own name, so as to ensure
preferences. Put simply, with McKinsey national ownership of the results.32 The exception
advice you don’t get dispassionate analysis of is the ‘Pathways to a low carbon economy’ report
transparent data so much as the advocacy of a for Brazil, notably less controversial than the other
particular – and literally patented – policy view of studies discussed here.33
12. 6
REDD+: McKinsey’s
influence in key countries 34
Democratic One of the major problems with logging rates. The business as in December 2008. McKinsey
Republic of the the DRC study is that it clearly
attempts to both obscure the role
usual scenario for industrial
logging forecasts an increase
was credited with ‘independent
fact based assessment’ for this
Congo (DRC) of industrial logging in destroying in logging yield from 3-5m3/ document, but circumstantial
rainforests and to ensure a future hectares to 15m3/hectares evidence suggests it was largely
McKinsey was commissioned for the logging industry at the by 2030,42 then suggests that McKinsey’s work.48 McKinsey
to produce a study of DRC’s expense of small-scale farming. restricting the increase in received £313,000 from the UK
REDD+ potential in late 2009. Far from reducing and eventually yield to 10m3/hectares is an Department for International
It produced its report after just eliminating deforestation, it emissions reduction; 43 Development for REDD+ work
five weeks, and although it is proposes a significant increase in done on behalf of the Guyanese
credited only with technical concessions. A billion euros in subsidies44 Government supposedly between
collaboration on the study, there to the intensive farming June 2008 and March 2009.49
are grounds to believe that the The McKinsey co-authored industry (mostly palm oil for
published document is mainly DRC study underplays the role export) to divert plantation In addition to the FCPF, a number
McKinsey’s work.35 Early in 2010, of logging in deforestation, and establishment outside of of donors have been, or are to be,
the DRC released its Readiness simultaneously overestimates existing dense rainforest;45 approached for assistance with
Preparation Proposal for REDD, the likely expansion of the logging Guyana’s REDD+ preparation
which provisionally adopted all of sector in the future, allowing and implementation, but apart
McKinsey’s proposals. companies to misleadingly claim farmers without from small contributions from
that they have reduced their consideration for community Conservation International and
The DRC is one of the nine efforts when compared to what lifestyle and traditions the German Development Bank,
initial UN-REDD Programme would have happened without and without reference to the only funding agreed has
pilot countries, and has been REDD+ intervention.39 The overall indigenous peoples.46 been from Norway, which has
given direct funding to help effect is to support business as committed support of up to $250
launch its REDD+ process. In usual for logging companies – On 29th January 2011, the million by 2015. The funding
addition to REDD+ readiness whilst efforts to reduce emissions DRC’s Environment, Nature is supposed to be conditional
funding received from the are directed at subsistence Conservation and Tourism on ‘Guyana’s success in limiting
World Bank’s Forest Carbon farmers because their activity Minister, José Endundo, greenhouse gas emissions
Partnership Facility (FCPF) does not contribute to GDP announced that he would from deforestation and forest
and the UN, in 2010 support growth – regardless of its social legalise logging titles in 15 degradation’50 but the basis for
of up to $20 million for pilot and cultural value. million hectares of rainforest defining this process has
projects was being considered and proposed that the been controversial.51
by the Congo Basin Forest Fund, !"#$%&"$'()'' government lift the country’s
funded by Norway and the UK.36 *+"'!,-()$".'' moratorium on new logging Guyana’s approach involves the
No information is available on concessions,47 which would country being paid to retain its
funding promised for the actual
,/0#%*+/&"1'234' open up an additional 10 standing forests on the basis
implementation of the strategy. $*%1.'(),5%1"6 million hectares of forest to of their ‘economic value to the
exploitation. McKinsey’s nation’ were they cleared almost
McKinsey received $300,000 A significant increase in advice has legitimised entirely for timber, agriculture and
for its work, paid for by a industrial logging, with government policy. development at a hypothetical rate,
Multi-Donor Trust Fund (now an increase of at least an which is actually far above that ever
closed) overseen by the World additional 10 million hectares Guyana seen in the country. This rate (4.3%
Bank and funded by the UK, given as logging concessions;40 deforestation per year) would be
France, Belgium, Germany, Guyana is a UN-REDD partner around 20 times the government’s
Luxembourg and the EU.37 An argument that effectively country but receives no estimated current deforestation
McKinsey appears to have been states that companies should funding for its national REDD+ rate of 0.1-0.3%.52 The approach
appointed by direct agreement be paid (at a rate of $2 to programme. It published its is explained as appropriate to
rather than through an open 2.5 per tonne of CO2e)41 for approach to pursuing external Guyana’s status as a high forest
tendering process.38 doubling or trebling existing funding to avoid deforestation cover, low deforestation country.53
13. 7
!"#$%&"$'()'7%.#)#8$'' the actual implementation of of deforestation and ‘facilitate The effect is to make it seem 30
95#)'(),5%1"6 PNG’s REDD+ programme. progress towards performance- times cheaper to displace a small
A financial plan for interim based payments for emissions farmer than to challenge the
Almost no measures to funding requirements is being reduction’.70 Other funding for incursion of new plantations into
address the existing drivers developed.62 that period included $64.4 million natural forests.
of deforestation in Guyana. In from Australia and $30 million
fact logging would be allowed We have no information on from Germany for ‘Measures on Meanwhile, the logging industry
to increase by 20 times its who has paid for McKinsey’s Reporting and Verification’, work is declared off-limits. Discussing
current rate;54 work in PNG. towards a reference emission what it calls ‘sustainable
level and other preparation forest management’, the cost
Use of REDD+ funding :+"'1/,%;")*$'/)' work.71 According to a 2009 curve report claims that: ‘The
to facilitate ‘higher value <+(,+'!,-()$".' source, the UK, Japan and Norway alternative – stopping logging
agricultural development’, have also promised funding for altogether – would have
including biofuel production in
</&="1'#1>/,#*"' capacity building of Measurement the same effect on emission
‘unique and fragile’ Savannah 9/5(,("$'(),5%1()?6 Reporting and Verification reductions [as sustainable
ecosystems and wildlife rich (MRV)72, REDD+ markets and forest management], but has
wetlands;55 Continuation of large-scale fund distribution.73 a much higher opportunity
commercial logging under a cost and would not allow
Use of REDD+ funding to so-called reduced impact The Norwegian money was Indonesia to further develop
construct the Amaila Falls regime,63 yet a moratorium on paid direct to McKinsey and was its forest products industry.’81
Hydro-Electricity Project. new logging concessions not subject to a competitive McKinsey does not explain
A recent study suggests is explicitly rejected;64 tendering process as McKinsey the assumptions behind this
considerable impact on was already active in Indonesia statement, but its implication –
forests from clearance No measures to address in September 2009 when the that logging must continue and
through to building the plant mining, despite its role as a funding was agreed.74 that this will not compromise
and its access road: 750,000 major driver of deforestation; emission reductions – is central
tonnes(t) of biomass are to be Greenpeace has also seen to the proposals in the plan.
cleared from the dam site56 Major agricultural two McKinsey reports for
and the access route will intensification affecting the Indonesian government: !"#$%&"$'9&/9/$"1'()''
include 110km of a minimum subsistence farmers;65 Detailed project overview (Phase @)1/)"$(#'(),5%1"6
8m wide road cut through 3): Implementation support
primary forest.57 750,000t Afforestation and plantation for Central Kalimantan, from An additional 10 million
biomass is equivalent to 1.3Mt on pasture and other non- February 2010 and Detailed hectares of afforestation and
CO2 emissions.58 forest land, likely to impact on project overview (Phase 3): reforestation via plantations
areas with very high value to Institutional capacity building for with a lack of clarity as to
Papua New wildlife.66 low carbon growth. McKinsey whether industrial plantations
Guinea (PNG) Indonesia
charges approximately $3.6
million75 and $6.1 million76
are to replace natural forests;82
In 2010 Papua New Guinea respectively for capacity building. The payment of large sums
published three documents Indonesia is one of the initial of money, effectively
relating to its national climate UN-REDD Programme The cost curve for Indonesia compensation,83 to divert
change REDD+. Though McKinsey pilot countries and receives contains flawed assumptions, establishment of pulp and palm
is credited only with data direct support for its national which significantly bias the oil plantations from forested
and analysis for the first two programme.67 Indonesia has final outcome to protect the land when improvements to
documents, and not at all for the also been selected as a World interests of industrial forestry productivity could mean only
third, there is strong evidence Bank Forest Investment and agri-business. The cost of a marginal increase of new
for McKinsey being largely Programme (FIP) pilot country. reducing emission from limiting land area would be needed to
responsible for all three.59 At the Copenhagen conference plantation expansion into natural meet government targets for
in December 2009, President forests is set as high as possible, production expansion;84
PNG is one of the nine UN-REDD Yudhoyono pledged Indonesia to by assuming that there are no
Programme pilot countries reduce overall emissions by 26% alternative locations possible
and receives direct support by 2020 using domestic funding – nearly $30/tonne CO2e or increased greenhouse gas
for its national programme, only, while aiming to increase $20,000/ha.77,78 In contrast, emissions by proposing that
which ‘aims at initiating the that figure to 41% with help of the forecast costs of reducing the definition of ‘forest’ will
quick start phase of readiness international funding.68 emissions from smallholder be more than 30% canopy
support for REDD+’.60 In agriculture are minimised to cover. According to the
addition to funding of $6.4 As of May 2010 (the most recent include only the monetised joint Indonesia National
million from UN agencies, PNG figures available), Indonesia had value for production79 – a figure Development Planning
has received or been promised received or been pledged FCPF of $1/tonne CO2e80– which Agency–UN-REDD draft
funding from Australia (up to $3 and UN-REDD funding totalling clearly recognises neither the National REDD+ Strategy85,
million), Japan (¥700 million) $9.2 million (the UN-REDD transaction costs nor, more 10% canopy cover (the FAO
and the EU (unspecified).61 contribution of $5.6 million being importantly, the wider social, threshold for definition of
No information is available on funded by Norway69), and $80 environmental and cultural forests) would be classified as
any donor commitments for million from FIP to address drivers impacts of such an intervention. ‘high carbon’.
17. 11
The McKinsey
MAC curve: an
optical illusion
McKinsey has risen to prominence within the assumptions relied upon in its calculations.86 Due to
climate change and REDD+ spheres through its the company’s stringent application of intellectual
global greenhouse gas abatement cost curve, property rights on its data, the outside world has
which the company conceived in 2007 and no way of knowing how McKinsey arrives at the
updated in 2009. The so-called ‘McKinsey curve’ different cost estimates attributed to various
has been extremely influential in setting the terms abatement measures.87 The potential victims of
of the debate for international carbon reduction a REDD+ policy which displaces local farming will
regimes and other marginal abatement cost (MAC) thus never have access to the reasoning behind
curves inspired by the McKinsey model have since why this policy was deemed cheap in the first place,
become hugely influential in carbon abatement let alone considered acceptable.
policy. They are a simple way of ordering and
presenting different options for reducing emissions The use of projected emissions raises another set of
and typically look like a succession of rising steps, thorny issues. McKinsey cost curves typically work
each one a different potential measure, its height on assumptions about what a country’s emissions
representing its cost, and its width representing will be in 2020 or 2030, so it is necessary to
the amount of carbon abatement it could deliver. calculate, based on current trends, what emissions
are likely to be at that date before the abatement
The cost curve approach to carbon reduction potential and the associated compensation for
has many immediate attractions, not least that it REDD+ action can be calculated. This introduces
allows policymakers to focus on the least expensive a clear incentive to inflate projections in order to
measures first and to get an idea of the total cost be paid more for not actually producing emissions.
of a given level of emission reduction. But as with The dangers of this approach are clearly illustrated
many simple presentations of a complex reality, in the case studies in this report, in particular the
MAC curves can disguise significant dangers; in projections for logging yield in the DRC and PNG,
particular, where there are flaws in underlying which result directly from McKinsey advice.
assumptions about comparative costs.
Cost curves for REDD+ are not able, and do
This is especially true when it comes to costing not seek, to integrate the web of social and
the measures in REDD+. For example, if the true environmental values associated with tropical
costs of displacing local subsistence farming are forests beyond their carbon sequestration and
underestimated – as this report argues they are storage potential. MAC curves treat tropical
– by ignoring transaction costs and wider social forests like a carbon abatement technology, rather
and environmental impacts, whilst the costs of than recognising them as some of the world’s most
addressing industrial logging are overestimated (for complex living systems, supporting a staggering
example by exaggerating the economic value of variety of biodiversity, as well as being of great
logging to the economy), and these assumptions are economic and cultural importance to humans.
built-into the cost curve, then every policy decision
flowing from the use of the curve will tend to favour It is this basic lack of understanding – along with
logging interests over those of small-scale farmers. some rather fundamental mistakes in biological
The result will not just be socially destructive, but carbon accounting – which too often seduce
may prove impossible to implement, economically policymakers away from measures to protect natural
irrational, and ineffective in reducing emissions. forests in favour of plantations and industrial scale
logging, for example. Until these flaws are
McKinsey claims to ‘rely on facts because they addressed, the use of the MAC curve in forest
provide clarity and align people’, but it is entirely policy making will remain at best misleading,
unwilling to transparently disclose the data and and at worst dangerous.
18. 12 UN B E L I E VA B L E
CO V E R T
Global consultancy firm, McKinsey advises rainforest
McKinsey’s, cost curve has nation governments on reducing
been extremely influential in emissions from deforestation. Yet
setting the terms of the debate it keeps most of its assumptions
for international carbon commercially confidential. Since
reduction regimes. these flawed cost curves are at
the heart of its advice, HOW CAN
C S THE REDD+ PLANS McKINSEY
AL
C U L AT I O N INSPIRES BE TRUSTED?
McKinsey’s FA L S E E CO N O MY
MARGINAL
The height of each bar shows how much
uch
ABATEMENT
COST
CO2e abatement measures cost. Butut
these costs are misleading because only
t
the missed opportunity costs get
CURVE included, and McKINSEY EXCLUDES
CERTAIN SIGNIFICANT COSTS
FOR REDD+ such as transaction,
ES
R
a
agri re for
an rn
bsisten
bsiste ers’
gal.
implementation, monitoring and legal.
Abatement cost ive i ood are ened,
bar doesn’ in
ba fi
¤ per tCO2e ost, no e st
of tt
60
2nd generation biofuels
40
Reduced pastureland Organic
conversion soil restoration
20
Reduce slash and Geothermal
burn agriculture conversion
Grassland
management
0 5 10 15
Building efficiency new build
-20 Waste recycling Degraded
land reforestation
-40
Cars full hybrid
Insulation retrofit (residential)
n I N T E R AC T I O N S
-60
Tillage and residue management
due
McKinsey puts each CO2e abatement
VAC
Retrofit residential HVAC measure in a bar to show its clients which
-80
are the mo cost effective. But the bars
most
Appliances residential are not fle
flexible enough to allow for even
th simplest
the simple of interactions. In reality, if
-100
Residential electronics one
one measure
on measu is increased or lowered, then
another m
an h
another measure can change in response.
Sin e it
energ
Lighting – switch incandescent eff ie
effi ess of an on
to LED (residential) red s, bars
s
M in
Since McKinsey’s assumptions are not available for public scrutiny, th cost
this
curve has been redrawn for illustrative purposes without using origin data.
strative original
19. INC R E DIBL E SNAPSHOT 13
McKinsey’s secrecy means that McKinsey’s cost curves only
t scientific community and
the focus on one year, usually
policymakers can’t see or 2030. But even where it’s
challenge the assumptions
c possible to predict costs, the
behind how McKinsey arrives curve doesn’t show the trends
a different cost estimates or
at over a period of time.
emission savings. McKINSEY’S
e McKINSEY’S CURVE IGNORES
WORK IS NOT OPEN TO
W DEVELOPMENTS BEFORE
PUBLIC SCRUTINY. AND AFTER 2030 THAT
MIGHT BE IMPORTANT.
LO S T E M I S S I O N S
McKinsey cost curves predicts CO2e
saving potential for each abatementnt
measure in 2030, shown by the width of
h
each bar. But McKinsey doesn’t showow
the emissions that accumulate over ar Gas plant CCS retrofit
period of time or their contribution to
Coal CCS retrofit
global warming, which could be much ch
more significant than presented. ba s,
Iron and steel CCS new build
an
afforestatio save
CO2e, it ignore Coal CCS new build
it save it
as it
Power plant
rainforest or even degraded rest
biomass
co-firing
Cars plug-in hybrid
Degraded
forest reforestation Solar PV
Solar CSP
Nuclear
20 25 30 35 40
40
Pastureland afforestation Abatement potential
ential
Low penetration wind
High penetration wind
GtCO2e per year
r
Reduced intensive
agriculture conversion
MISSING BENEFITS
The x axis shows the potential CO2e
savings via the width of ea measure,
each
FALSE SENSE OF CERTAINTY ’s no of e but doesn’t factor in any additional
st in benefits or costs. Thes missing
These
’s st rv does no beyo
benefits and costs, beyond carbon
McKinsey presents forest-related future re
ref rang of r
rtaint
abatement costs as certainties. But sincece emissions, ought to be iinfluencing
margins of error ca be greater than cost
can ost REDD+ plans, and also have implications
hav
betw ot
differentials between measures, it’s not a
for other policy areas.
realistic to predict which will be cheaper.
er.
fo
Costs may vary for REDD+ abatement t
measures due to location, land use
change, policy and market forces. Ra
Rainforest o er
ive fo
ere;
t e
intensiv agri t nversion
be va
e doesn’ t differen
do
doesn’t
o rt nit st e
ed differen
of in in differen
21. 15
What’s wrong
with McKinsey’s
method?
AB'' /&"$*',#&D/)6''
C It is not made clear anywhere whether the
()#&*(,%5#*"'/&'()"9*E' existing carbon stock of land targeted for
plantations has been taken into account.
Measuring carbon accurately, with clear and For example, the DRC study recommends
verifiable methodology, is critical to the success afforestation on ‘shrubby savannahs or forest-
of REDD+ programmes. However, forest carbon savannah mosaic’89 – but without data on
accounting systems are complex, controversial and the carbon which is already stored in these
still a matter for debate. This makes it particularly ecosystems, it is impossible to calculate whether
important that REDD+ plans show exactly how they putting plantations on them will actually reduce
have calculated any emissions savings, so that they emissions – or by how much. Yet the same
can be assessed independently and verified. report gives emissions from logging as net
figures – that is, assuming regrowth of trees
Yet McKinsey keeps most of the workings of its cost which will in turn reduce the overall impact on
curve commercially confidential, this means that its emissions. The result of these two approaches
calculations of forest carbon savings are hidden and taken together, is likely to exaggerate the
therefore can’t be verified. Each potential action, emission reduction potential of plantations,
such as preventing logging, or planting trees, is and minimise the negative impacts of logging
given a cost per tonne of carbon saved and assessed – resulting in an inevitable bias in the kinds of
for its total abatement potential – but there is solutions proposed in the plan.
almost no indication of how these results were
reached. Many of the assumptions and calculations The same DRC study shows similar distortions.
underpinning the results of the cost curve are Its agroforestry case study shows carbon
concealed as if in the workings of a black box. sequestration by a plantation equivalent to around
150 tonnes of carbon stored per hectare.90 This
There is evidence of major problems with the figure is around the same amount of carbon
cost curve carbon accounting methodology: sequestered by untouched primary forests in the
region 91 despite the fact that the plantations are
a. Carbon stocks and flows described as being harvested for fuelwood and
in plantations construction. This is grossly unrealistic.
McKinsey co-authored studies focus almost b. Unrealistic precision
exclusively on carbon flows (emissions and
absorption), usually given at two static points in The McKinsey cost curve generates predictions
time – today’s current emissions and net flows with unrealistic precision. For example, a
in 2030.88 They do not describe how this carbon fact sheet on the Indonesia cost curve gives
stock – ie carbon stored in forests and soils – might emissions reduction estimates in 2030 to two
change over time. This makes interpreting the decimal places. This level of precision obviously
figures given for ‘reforestation and afforestation’ gives an exaggerated picture of the reliability of
(ie plantations) particularly difficult. the estimates.92
The numerous errors and biases suggest that McKinsey lack an understanding
of the fundamentals of carbon accounting. The confusion of net and gross
emissions, the neglect of effects on carbon stocks and, most importantly,
the persistent failure to display a robust and transparent carbon accounting
methodology seriously undermine the documents’ credibility.
22. 16
PLANTATIONS: FB'' #*#'1"G(,("),("$6''
2
ROOT AND BRANCH ()#1"H%#*"'/&'#D$")*E'
CONFUSION It is not only McKinsey’s secrecy that is
troubling: in some instances the so-called
data that McKinsey has used to produce
Indonesian Government
recommendations may simply not exist.
documents confuse different
plantation types and consider In the DRC study, for example, a table is given
commercial plantations as showing confidence in individual emissions
factors.99 The table reveals that illegal logging
‘carbon sequestration’ and
and fuelwood factors have been reached
‘sink enhancement’. despite there being ‘no exact data available’. It
is unclear what assumptions have been made
or analysis done in the absence of this data.
The conclusions on industrial logging are also
highly suspect, due to the lack of governance,
control and law enforcement, and the level of
corruption in the DRC logging sector.
The DRC, PNG and Indonesia studies rely
heavily on plantations, usually referred to in In the development scenario set out for
the reports based on McKinsey’s advice as Guyana, meanwhile, evidence-based planning
‘afforestation and reforestation’. Although is largely abandoned in favour of speculation.
these plans do not explicitly advocate The Low Carbon Development Strategy
replacing natural forest with plantations, using suggests extensive agricultural and forestry
plantations in the emissions abatement figures development, including on large areas of land
acts to mask ongoing deforestation. which are almost certainly unsuitable for such
activity. A Guyanese forest expert, Janette
Each of the cost curve reports bases its Bulkan, has commented on the ‘extreme
predictions on what it calls ‘conservation’93 infertility of most of [Guyana’s] forest-
plantations or ‘afforestation aiming to sequester covered hinterland soils’100 which makes
carbon’,94 that is, plantations not intended for them ‘much less likely to be convertible to
harvest. Such plantations have no economic use financially-profitable, ecologically-sustainable
other than to attract REDD+ credits. These plans agriculture than in neighbouring Brazil’.
would pay developing countries to hand over
large areas of what may be biodiverse and useful With a similar disregard for basic data
land to ineffective plantations, while continuing to or evidence to support its assumptions
cut down natural forest. and proposals, PNG’s Interim Action Plan
proposes that measures to increase yields
There is also the possibility that the McKinsey- and market access in subsistence and
inspired plans could lead to REDD+ funding smallholder agriculture would save 9-15
supporting pulpwood and oil palm plantations, megatonnes of CO2e per year by 2030,101
which do not sequester significant amounts but admits that ‘the abatement effect of
of carbon.95 Greenpeace has previously noted these measures is unproven’.
that Indonesian government documents
confuse different plantation types and consider Elsewhere, data from countries in different
commercial plantations as ‘carbon sequestration’ continents is used to attempt to construct
and ‘sink enhancement’.96 This possibility is arguments in support of McKinsey’s favoured
admitted in the Indonesia and PNG cost curve REDD+ interventions. The PNG report, the
reports. For example, it is suggested for PNG that Climate Compatible Development Strategy,
if reforestation included forestry plantations, cites evidence from African countries in support
this would require ‘further research/analysis of proposals for ‘agricultural extension’, ignoring
… to calculate the abatement potential’.97 The the different ecological and cultural conditions
possibility of REDD+ funding going to commercial affecting PNG farmers.102 The document
plantations is not ruled out. claims that ‘Technical appendices containing
this data and analysis are available on request
Although not one of the principal case studies from the Department of Environment and
considered in this report, McKinsey’s advice Conservation’103 but Greenpeace requests for
to the government of Brazil is illustrative these appendices have been unsuccessful.
here. McKinsey’s report suggests that both
‘commercial forestry operations’ such as ‘pulp These examples suggest these reports are
production’ and ‘reforestation using native not based on hard evidence. They present
species...not for commercial use’98 could form possibilities as if they were firm policy plans,
part of REDD+ plantation programmes. backed by inadequate, if not absent, data.
24. 18
KB'' #>/%&()?'()1%$*&(#5'()*"&"$*$6'
C reduction potential, and add to the time it takes
$="<"1'9"&$9",*(>"$ to implement a REDD+ strategy.’ 118
McKinsey co-authored studies repeatedly use tricks This is particularly important for programmes
of data presentation to protect or promote industrial relating to smallholder or subsistence agriculture and
logging and large-scale agricultural interests at the fuelwood collection, where implementation costs
expense of subsistence farming. The methodology are likely to be very high. The failure to include the
of the cost curve contains implicit assumptions on costs and difficulties of communication with large
the relative value of different activities, particularly numbers of people in remote areas, added to the
logging, industrial agriculture and subsistence or failure to account for the value of non-monetised
smallholder agriculture. The overall effect is that the land uses, means that the financial and social cost of
potential emissions savings from targeting small- programmes tends to be underestimated and their
scale agriculture are repeatedly overestimated, and potential effectiveness overestimated.
their costs underestimated, in comparison to tackling
the commercial drivers of deforestation. In contrast, forecast emissions reductions from
reduced plantation expansion are costed at the
The country reports for Indonesia, the DRC and theoretical opportunity cost based on the value of lost
PNG, for example, base their calculations of the cost production. In this instance the maximum opportunity
of emissions reduction from avoiding deforestation cost – nearly $30/tonne CO2e or $20,000/ha119 – is
and degradation on a theoretical ‘opportunity based on an assumption that plantations will not be
cost to the nation’ which excludes ‘transaction, established at all if they are not on forested land,120
communication and information costs,’ 116 that is, despite admitting that much expansion could be
the cost of implementing an emissions reduction relocated to non-forest land much more cheaply.121
programme. This tends to misrepresent the costs In effect, the cost of reducing plantation expansion
and desirability of different emissions reduction is inflated while the cost of reducing smallholder
options. In DRC, McKinsey is actually very explicit expansion is minimised until it is virtually meaningless.
in equating the abatement cost to ‘the reduction
of profit margin incurred by the company’.117 As The partial use of the concept of opportunity cost
the World Bank review of the DRC’s R-PP argues, also skews the cost curve’s priorities. While the
‘[Transaction and implementation costs] can REDD levers (smallholder agriculture and plantation
significantly increase costs, reduce the emissions development) are based on opportunity cost – defined