6 - WB (2018) - Nesting and BS Experiences
6 - WB (2018) - Nesting and BS Experiences
6 - WB (2018) - Nesting and BS Experiences
April 2018
Approaches to REDD+ Nesting
Lessons Learned from Country Experiences
April 2018 Donna Lee
Pablo Llopis
Rob Waterworth
Geoff Roberts
Tim Pearson
Table of Contents
Executive Summary 3
Abbreviations 4
4. Lessons Learned 31
Cookstoves
Climate-smart
agriculture
Protected area
in Australia and New Zealand. For these systems, There are many different forms and definitions of
projects form part of a national policy response (for nesting, each developed for specific policy purposes.
example, a domestic carbon trading or tax system) and While focusing on REDD+, many of the issues
are therefore more likely to be consistent with national described in this paper also apply to the broader land
accounting and contribute to targets. This approach is sector—an important consideration in light of the Paris
more likely to represent the future of REDD+ nesting— Agreement and countries’ efforts to meet their NDCs.
for example, Colombia is pursuing a system that uses
revenues from a carbon tax to provide incentives for 1.2. Why Nest?
local-scale projects.
Nesting requires considerable policy and technical
By taking this broad view of nesting, the paper support from both the private and public sectors. As
assumes that it can include the following: such, it is important to understand why a country may
consider nesting and the potential costs and benefits.
● Benefit-sharing approaches, where a large-scale The many reasons to consider nesting include but not
program (national or subnational) generates are limited to the following:
emission reductions and/or receives carbon (or
REDD+ results-based) finance, and shares it with Providing early and future benefits. Local-scale
smaller-scale units to incentivize local actions activities (including carbon projects) can stimulate
private investment, provide operational on-the-
● Systems that engage in carbon accounting at ground capacity often lacking in countries, and offer
multiple scales—for example, countries that have lessons (and results) that, ideally, can be replicated.
GHG commitments at higher scales (such as If the conditions are right, such activities/projects
Kyoto Protocol or Paris Agreement NDCs) or may can support and become critical building blocks for
be generating carbon units or receiving results- jurisdictional programs. Some have been able to
based finance at higher scales, but also allow access finance from the voluntary carbon market.
smaller-scale units to generate benefits tied to
GHG performance Creating a pathway for governments to implement
policies to reduce emissions, particularly in countries
● The incorporation of local-scale activities/projects where mitigation is expected to occur on private or
into jurisdictional REDD+ schemes as well as community-controlled land. For example, a government
subnational programs into national REDD+ may enact a policy that provides incentives to
schemes landholders or managers who operate at a smaller
Figure 2.1: Options for the allocation of funds from the sale of ERs
Allocation of funds
SUMMARY
Simple: This option is the simplest from a Stronger incentives: Certain types
technical perspective. There are no concerns, of stakeholders may respond well to
for example, with double counting or aligning performance metrics (e.g., the private
multiple MRV systems. sector).
BENEFITS Flexible: A country can decide how to spend Catalyzes private investment: Because a
the funds in a variety of ways (including to potential return on investment can engage
achieve non-carbon benefits). This may also private finance, this option may be useful for
include using funding to leverage private governments with insufficient resources or
investments. that do not have strong fiscal levers.
Allocation of funds
subnational scale may consider whether to: (a) issue, allocating a share of issued ERs (at the jurisdictional
sell, and monetize ERs, and then share that finance level) to local actors, or (b) allowing local actors to
with local actors; or (b) allocate ERs (figure 2.2). If a issue (for performance achieved at the local scale) a
country decides to fully monetize ERs from performance prescribed number of ERs themselves. In most cases,
achieved at the higher scale, there will likely not be ER a government will likely want to ensure that the total
generation at smaller scales (to avoid double counting). number of ERs issued (at local scales) is at, or below,
This is an instance where the jurisdiction is in full control the total number of ERs verified at the highest (for
of ERs generated and uses the subsequent finance example, national) scale.
received as part of a benefit-sharing system.
Boxes 2.4 and 2.5 illustrate two examples where a
If a country decides to share a portion of the ERs country has chosen to allocate finance or ERs as ex
with local actors, it may do so through two means: (a) post rewards.
Projects are capped with regard to how much they may receive from the Carbon Fund payments. This is
due to the risk that projects may perform extremely well compared to the overall jurisdictional (province-
wide) performance, and thus claim a high percentage of payments from the Carbon Fund. For this reason,
the remaining ERs not purchased by the Carbon Fund will go into a pool of in-kind ERs that can be
provided to individual projects (which may then monetize them) for performance achieved beyond that
rewarded by the Carbon Fund (for example, due to the imposed cap).
*The DRC case study is based on an advanced draft benefit-sharing plan developed by the country in consultation with ER
Program stakeholders. The negotiations between DRC and the FCPF Carbon Fund regarding an ERPA are ongoing; it is not
clear at this stage if the provisions (which are analyzed in this paper) will be accepted by donors for contract signature.
SUMMARY
Allocation of funds
% of $ to % of $ to % of $ to % of $ to % of ERs to % of ERs
higher level smaller higher level smaller higher level to smaller
jurisdiction scale units jurisdiction scale units jurisdiction scale units
SUMMARY
Sharing finance or ERs between the national (or subnational) government and local actors is a continuum from 0
percent to 100 percent; deciding the best proportional split will depend on needs and an assessment of the costs
and benefits of national versus local allocation as well as other available financial sources.
The share of finance/ERs distributed to the The share of finance/ERs distributed to local
DESCRIPTION national or subnational government for its actors (e.g., for their contributions to overall
contribution to performance. performance of the jurisdiction).
BENEFITS • There is no negotiation process with local • The upward feedback process is simpler.
entities. • Provides direct reward for performance,
• Increases national commitment to encouraging private sector engagement
achievement of REDD+. in REDD+ in the country and achieving
the efficient and cost-effective emission
reductions.
• The process of negotiation for the attribution
• Decision power is centralized and lobby
of funding to each local unit is much more
activities are easier, which can detract from
complex and can jeopardize the overall
CHALLENGES optimal jurisdictional performance.
operation of the jurisdictional system.
• The incentive for local entities to perform at
• Monitoring of financial flows is complex and
their optimum is lower.
may be subject to corruption, if it exists.
SUMMARY
The case where projects generate carbon units separate from the higher-scale, jurisdictional
DESCRIPTION
program
• MRV mismatch: Allowing projects to generate their own ERs may require development of MRV
rules and systems to minimize mismatch at different scales and for some entity to take on the
liabilities for mismatches.
CHALLENGES • Double counting: Where projects are allowed to sell carbon units internationally, systems are
needed to avoid counting the same unit twice within the same context (e.g., Paris Agreement).
• Special care needs to be taken to separate emission reduction certificates created by the
smaller units, that is, registry procedures need to be reliable.
Flow of ERs
“Nested”
Flow of finance
activities
feed
back or
ERs used for achievement of Remaining ERs contribute,
the (unconditional) NDC to national
emission
reductions
% of $ to % of $ to % of $ to % of $ to % of ERs to % of
Projects issue Domestic higher level smaller smaller
higher level higher level ERs to
ERs and engage system jurisdiction scale scale
jurisdiction jurisdiction smaller
in international that rewards units units scale
trading projects
units
STAND ALONE
PROJECTS
Mai Ndombe province generates and issues ERs for jurisdictional performance
Carbon Fund ERPA (purchase of X% of ERs achieved) In-Kind ERs (remaining Y% of ERs achieved)
Possible
distribution of ERs
Allocation of funds to projects
Figure 3.1: Nested systems vary regarding the level of required MRV alignment
NESTED SYSTEMS
Allocating Allocating Allocating Allocating ERs Stand-alone Stand-alone
finance, finance, finance, based on GHG projects with projects with
ex-ante grants ex-post rewards ex-post rewards performance own MRV in own MRV and
using proxy based on GHG domestic only trading ERs
measures performance systems
EXAMPLES
Amazon Fund Mai Ndombe ER Program Australian ERF
(original)
Amazon (new)
GRIF
National Project
• Uses (medium resolution) Landsat to measure • May use higher resolution imagery to measure
forest cover change forest cover change
• Carbon stock estimates from national forest • Often collects and uses own site-specific carbon
inventory or default values stock estimates
• Stratification based on national forest classes; • More strata than national GHG inventory due to
may combine strata to reduce uncertainties use of data with finer spatial scales
• Often only includes above- and below-ground • May choose or be obligated to measure
biomass; excludes non-CO2 gases deadwood, litter, soil, and non-CO2 gases
measure land cover change using Landsat imagery accuracy. In most cases, this led to the use of site-
and derive carbon stock and carbon stock change specific ground data and project-specific assumptions
estimates from national forest inventories, International of additionality and leakage.
Panel on Climate Change default emission/removal
factors (for gap filling), and, in some developed Given these differences, it is not surprising that
countries, existing growth models. In many developing national systems will produce different estimates of
countries, initial estimates from national systems forest-related emissions (and/or removals) compared
have only been developed over the past two to three to projects. At the national scale, a country may
years and will require ongoing improvement. When also aggregate data differently than smaller scales
forest carbon projects were being developed, many for various reasons. For example, some countries
developing countries could not provide national data for the purpose of developing a national forest
or systems to consider or work with. Furthermore, many reference emission level (FREL) have stratified their
of the guidelines applied, such as the Verified Carbon data into fewer forest classes—balancing availability
Standard (VCS), had very different requirements, such of data, measurement costs, and uncertainty, and
as shorter time series of data but a higher degree of also simplifying the implementation of an accuracy
KEY POINTS
There are differences in data used by national governments, for example, for reporting GHG inventories
to the UNFCCC, compared to that used by forest carbon projects. Such differences are typically more
significant with legacy projects.
Data to measure results may flow top-down or bottom-up (or both). In Brazil, data are flowing from the
national system to the states; in Australia, a more sophisticated system integrates data flows in both
directions.
It is much easier to nest subnational units than projects. Alignment of data reduces the level of mismatches
that occur when measuring performance at multiple scales. Subnational jurisdictions typically use national
data and therefore are more easily aligned to national estimates. There may be instances where subnational
jurisdictions use higher-resolution spatial data or region-specific carbon stocks—in such cases, the national
system can integrate data into its systems more easily than project data, particularly if such units use
administrative boundaries that aggregate to cover 100 percent of the national territory.
assessment. Other differences may exist through use. Higher levels of accuracy may require resources
projects requiring data to be at more regionally or unavailable to a government, and more granular
locally specific scales. For example, carbon stock data can be a trade-off with uncertainty estimates—
estimates (typically from a national forest inventory, which at higher scales is sometimes larger than the
or NFI, plot system across the entire country) may expected change (or performance). The level of
differ when averaged up at the national scale versus accuracy of measurements required at multiple scales
provided at a project or regional scale. may also depend on the risk tolerance of the actors
involved (and who bears that risk). Precision may
If a country’s nested approach allows stand-alone be less of an issue so long as the system does not
projects, a higher level of accuracy and precision, generate biased results.
and a more granular stratification (of areas and
estimates of carbon stock for forest/land classes), 3.2. Baseline Setting or Allocation
may be needed at the national level. In some
instances, it is simply not possible for a project to Currently, the way in which carbon projects develop
use national-level data—for example, there may baselines and account for emission reductions is
be insufficient accuracy of the spatial data that is often fundamentally different to the methods used
required for the standard that the project wishes to for jurisdictional approaches. Large-scale reference
Where legacy projects exist, nesting can become both politically and technically challenging—particularly if
projects developed baselines prior to the higher-level jurisdiction. Table B3.3.1 illustrates three very different
baseline methods used within Zambia (at project and national scales).
If a project is located in an area of high-expected deforestation (for example, a “hot spot” or frontier), it is
fair to expect higher than average deforestation. However, it is also the case that several projects—applying
project-based methodologies and located within a single region—may quickly exceed a historical average
baseline of the region. In such cases, agreements will need to be made that often require projects to take
a “haircut” on their baseline (and thus their expected performance), sometimes putting at risk the business
models in place.
levels often use historical averages to calculate (or It is easier to nest afforestation/reforestation (A/R)
project forward) a reference level. Results-based and improved forest management (IFM) projects than
financing for large-scale programs (largely funded it is to nest avoided deforestation projects.
by donor governments) also has preferred the use of This is largely the case because for A/R, in most cases,
historical averages (with upward adjustment for high the baseline is zero since the starting point is an area of
forest cover, low deforestation, or HFLD, countries). land that is non-forest, which is similar to the baseline
By comparison, projects use a range of methods and assumed at the jurisdictional level (for example, the
models to project forward an assumed business-as- Kyoto Protocol’s accounting rules for A/R are “gross-
usual scenario. net,” which basically sets the baseline to zero). For
14
12
10
8
MtCO2
0
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
In Mai Ndombe’s ER Program, projects will be able to receive a portion of payments from the Carbon Fund
based on performance. In return, such projects will be required to: (a) use agreed baselines, (b) only
generate credits against the agreed baseline (that is, they may not develop a separate baseline under, for
example, the VCS, and separately issue Verified Carbon Units), and (c) accept the capped amount that any
one private sector or large community project may receive from the Carbon Fund. The PMU is expected to
develop guidance and information on how future project baselines may be developed.
*The DRC case study is based on an advanced draft benefit-sharing plan developed by the country in consultation with ER
Program stakeholders. The negotiations between DRC and the FCPF Carbon Fund regarding an ERPA are ongoing; it is not
clear at this stage if the provisions (which are analyzed in this paper) will be accepted by donors for contract signature.
When the resolution from CONAREDD+ was issued, 13 REDD+ projects developed in the Amazon biome
had already been validated under the VCS, and few of them had used PRODES data to identify their
baseline and build their reference levels. The CONAREDD+ Resolution means that projects within states—
specifically those states that will further allocate funds through a state-level nested system that provides
incentives to projects—will need to invest in remodeling the baseline and, in many cases, will have to
assume a lower generation of credits that may jeopardize their financial feasibility.
When the resolution from CONAREDD+ was issued, 13 REDD+ projects developed in the Amazon biome had already been validated under the VCS, and few of them had used PRODES data to identify their baseline and build their reference levels. The CONAREDD+ Resolution means that projects within states—specifically those states that will further allocate funds through a state-level nested system that provides incentives to projects—will need to invest in remodeling the baseline and, in many cases, will have to assume a lower generation of credits that may jeopardize their financial feasibility.
KEY POINTS
For avoided deforestation or avoided forest degradation, because there are different dynamics and
pressures on forests occurring across a landscape, jurisdictional baselines (for example, average
deforestation rates) in most cases cannot be simply downscaled to project-level actions.
The stock-flow method used by Brazil is one solution to providing incentives for both standing stock (that is,
conservation of existing forests) and reductions in deforestation—but it is more easily applied to nesting of
subnational areas into a national system.
It is much easier to nest A/R and IFM projects than avoided deforestation because baseline or reference
level construction methods are more similar at small and large scales.
Smaller-scale baselines (and results measurements) do not necessarily need to perfectly match higher
(jurisdictional) baselines as long as, in aggregate, they add up and do not exhibit bias—that is, that smaller-
scale performance, on average, is aligned with the overall higher-scale performance.
Developing nested baselines can change significantly the incentives (and validity of business models) for
smaller-scale activities.
Emission
Reduction Unit
(ERU) transferred
to Country B
Country A: AAU
Assigned surplus
amount
(KP target)
Country B: AAU
Actual deficit
emissions
Country A: Country B:
Actual Assigned
emissions amount
(KP target)
Mismatches between national GHG inventories and calculation of units by projects are generally unknown;
however, the relatively low volume of traded units compared to AAUs minimizes the risks of misalignment.
envelope of jurisdictional performance. However, not tied to the generation of credits (assets) or not
where jurisdictions and projects or subunits with the used in compliance systems. Where carbon finance
jurisdictional area are accounting simultaneously, there operates more akin to payments for (ecosystem)
needs to be a mechanism to avoid double counting. services, and no rights or titles are transacted, double
In addition, because often it is not possible to perfectly usage of the emission reduction is less problematic
nest projects within a jurisdiction, there will be a need to (although some donor governments are concerned
“true up” the accounting and manage liabilities for the with double payment for such reductions). Selling into
expected quantitative correction that may occur. This voluntary carbon markets—for example, to entities
requires clarity on which party or parties are responsible whose purchase of the offset does not appear under
both for managing the mismatches and for the liabilities any accounting within the UNFCCC—also is less
for any “overshoot” that occurs. Alternately, systems problematic. These may be considered as situations
(for example, buffers) may be developed to share the where double claiming may not be desirable, but it
burden up front of such overshoot. does not affect the integrity of any single system.
It is worth noting for this topic that these issues Boxes 3.8 and 3.9 describe examples of avoiding
are less critical for performance-based payments double counting.
To maintain consistency with Australia’s international obligations, the Emissions Reduction Fund project
proponents are issued Australian Carbon Credit Units, which are managed through a centralized registry,
the Australian National Registry of Emissions Units. These credit units are directly exchangeable for Kyoto
units, and then can be traded, surrendered, or canceled through the registry. Thus, if a project is issued a
credit that can’t be reconciled through the national inventory system, the Australian government has this as a
liability. The methods, therefore, aim only to recognize abatement that can be identified through the national
inventory system.
With regards to REDD+, countries with conditional 3.4. Using Proxies Rather Than GHG
NDCs may consider selling ERs as part of the
conditional portion of emission reductions. Or, if a Metrics
country expects to overachieve at the national level— If a jurisdiction is allocating finance or ERs based on ex
that is, exceed its unconditional NDC—it has the post performance, it will need to decide what metrics
flexibility to allow even more ERs to be sold (figure to use, and these may be best determined based
3.2). If a purchasing country uses these reductions on the actions or outcomes it wishes to incentivize.
to achieve its NDC, those reductions will need to be Carbon performance may be one option, or metric,
debited from the selling country’s national accounts, as used to allocate finance. Performance could also
in the Kyoto Protocol, to avoid double counting. be measured, for example, by hectares of forest
Australia has a spatially explicit NGGI system, which uses a Tier 3 model (FullCAM) to estimate the emission
and removals from forests. This modeling framework was designed to overcome differing governance and
data collection arrangements for the forests; where the national government is responsible for reporting
GHGs to the UNFCCC, it is the state governments who have constitutional responsibility for managing public
lands, including forests. As a result of this approach, for nesting, the spatially explicit nature of Australia’s
GHG inventory supports the adoption of payments based on project-level performance in reducing GHG
emissions, as all project areas are covered by the national system (that is, the same tools can be used for
projects as are used by the national system), and the national system can in turn be improved to better
match the project-level estimates.
The Zambia model (box 3.12) is not currently used to government of Acre as the jurisdictional REDD+
share carbon revenues; however, it is provided as an program proponent.
example of a potential model that could be adapted. It
shows how a system may be built on an already existing The government of Guatemala has been involved
system to provide benefits to local communities. directly or indirectly in the development of different
REDD+ projects and considered the certification of a
subnational jurisdictional REDD+ program under the
3.5. Why Has Nesting Failed in Some VCS JNR. The subnational baseline and the future
Cases? deforestation model developed for the jurisdictional
REDD+ program were geographically explicit and
In some cases, although there may be utility in
had different strata that represented the various forest
providing projects with incentives, circumstances types occurring in the subnational region. This created
have created barriers to developing a nested system. a significant source of economies of scale for the
In the Brazilian state of Acre, Law 2308 (SISA Law, implementation of private initiatives in the northern
October 22, 2010) created a legal framework for lowlands of Guatemala. The projects seeking validation
operationalizing a jurisdictional REDD+ program. in the subnational region could use the data and
This legal framework considered a registry in which information from the subnational baseline to construct
projects developed within the geographical scope of their own baselines and nest into the subnational
the jurisdictional program could be registered if they jurisdiction. When the Guatemalan government
were previously recognized and integrated into Acre’s decided to seek finance from the World Bank’s FCPF,
IES Carbon Program. When the legal framework of the the process of validation of the jurisdictional scheme
jurisdictional REDD+ program was promulgated, there under the VCS JNR was left on standby. This is
were already REDD+ private initiatives in Acre that had because the reference level developed under the VCS
been validated and verified (that is, had issued carbon JNR only considered the deforestation component; per
credits) under the VCS. The divergence between the approach assumed by the country under the FCPF,
technical aspects employed in the private initiatives forest degradation and carbon stock enhancement
and in the jurisdictional REDD+ program, however, components also would need to be included. As a
were significant and created disincentives for nesting— result, the projects were not nested and, instead,
for both the developers of private projects and the continued operating as stand-alone activities.
• What are the policy (setting of reporting/accounting rules, transparency requirements, whether and how
to integrate legacy versus new projects, etc.) and operational (sharing of data, process for transferring
funds/units, operation of underpinning systems, etc.) challenges of nesting and how can they be
overcome?
• What technical measurement and monitoring systems could best support nested approaches?
• What are the risks inherent in a nested system (for example, MRV mismatches, double counting, etc.)
and how should responsibility for them be shared? Who is willing to take on such risks? How can
registries help to manage some of these risks?
• What are the conditions under which a nested approach or system makes sense (for example,
centralized versus decentralized government systems, balance of private versus public lands, where
decision-making authorities lie)? What legal frameworks enable nesting (or not)?
• Where should the costs be borne for both the implementation actions that reduce emissions (or
enhance removals) and the MRV of GHG performance? Where does the intellectual property sit for
measurements of GHGs?
• What are the technical and administrative requirements expressed by demand-side actors (to date,
donor governments or voluntary standards) for the jurisdictional REDD+ initiative?