Indonesia’s moratorium and emissions from deforestation

How much would Indonesia’s moratorium on forest concessions have reduced emissions

from deforestation from 2000-2010?

J Busch – R Lubowski – E Ashkenazi – K Austin – F Boltz – M Hansen – B Margono – M Steininger–F Stolle– A Baccini

Jonah Busch, Ph.D. (Conservation International) World Resources Institute Side Event, UNFCCC COP 18

Millennium Hotel, Doha, Qatar Thursday, November 29, 2012

1. How much of Indonesia’s 2000-2010 emissions from deforestation occurred within oil palm concessions (kebun) and timber concessions (HTI)?

2. How much did the designation of a concession between 2000-2010

increase the annual deforestation rate at a particular site, relative to if that site hadn’t been designated a concession?

3. How much would Indonesia’s moratorium on new concessions have reduced emissions if applied from 2000-2010? 4. What carbon price would have achieved an equivalent reduction?


19% in oil palm concessions; 26% in timber concessions 2. How much did the designation of a concession between 2000-2010


on average, oil palm concessions increased deforestation by 60%, and timber concessions increased deforestation by 110%, controlling for year- and site- specific effects 3. How much would Indonesia’s moratorium on new concessions have reduced emissions if applied from 2000-2010? by 578 MtCO2e/10 yrs (8.3%) assuming no leakage, extrapolating

rates from dated to undated concessions 4. What carbon price would have achieved an equivalent reduction? a carbon price of $2.05 (mandatory) or $9.40 (voluntary)





1,442,462

566,778

1,859,364

190,202

4,718,690

Deforestation (ha), 2000-2010 Total: 8.78 million ha

7,835,348

10,830,952

6,509,198

7,207,299 61,859,374

Forest (ha), 2000 Total: 94.24 million ha

Oil palm concession

Logging concession

Timber concession

Protected area

Unprotected, non-concession

341,097

27,609

135,348

10,793

768,365

Palm oil conversion (ha), 2000-2010 Total: 1.28 million ha

1,691,808,273

414,866,793

2,280,133,378

132,747,802

4,187,169,312

Emissions (tCO2e), 2000-2010 Total: 8.71 billion tCO2e





2.2x (3.4x) more deforestation INSIDE oil palm (timber) concessions than OUTSIDE concessions…

-

20,000,000

40,000,000

60,000,000

80,000,000

100,000,000

120,000,000

140,000,000

160,000,000

180,000,000

2000

20

01

2002

20

03

2004

20

05

2006

20

07

2008

20

09

2010

Are

a (h

a)

Year

Indonesia forest loss, 2000-2010

Non-forest (+13%) Other forest (-8%) Timber (-27%) Logging (-5%) Oil palm (-17%) Protected (-3%)

2.2x (3.4x) more deforestation INSIDE oil palm (timber) concessions than OUTSIDE concessions…

…but concession land is geographically different

-

20,000,000

40,000,000

60,000,000

80,000,000

100,000,000

120,000,000

140,000,000

160,000,000

180,000,000

2000

20

01

2002

20

03

2004

20

05

2006

20

07

2008

20

09

2010

Are

a (h

a)

Year

Indonesia forest loss, 2000-2010

Non-forest (+13%) Other forest (-8%) Timber (-27%) Logging (-5%) Oil palm (-17%) Protected (-3%)

Unprotected Non-concession

(n=136,963)

Oil palm Concession (n=22,285)

Timber Concession (n=16,076)

Slope (%) 7.2

6.3 5.7

Elevation (m) 340

286 272 Distance to nearest highway (km) 65

56 46

Distance to nearest provincial capital (km) 226

239 211

Potential agricultural revenue ($/ha/yr) $245 $212 $259 Above and below ground biomass (tC/ha) 148

149 160

Peat extent (%) 15%

22% 66%

Forest cover, 2000 (%) 52%

48% 59%

Forest cover, 2010 (%) 48% 41% 45%

Deforestation (%/10 yr) 7.8% 17% 24%

Palm oil cover, 2010 (%) 2.9% 9.0% 2.9%

2000-2010 Before

concession After

concession

Oil palm, undated -1.62%

Oil palm, dated -2.35% -1.1% -2.5%

Logging, undated -

Logging, dated -0.47% -0.38% -0.51%

Timber, undated -0.28%

Timber, dated -2.98% -2.6% -3.5%

Protected, undated +1.46%

Protected, dated -0.40% -1.4% -0.37%

Other forest -0.77%

Total forest -0.94%

Non-forest +1.25%

2.3x (1.3x) more deforestation AFTER oil palm (timber) concessions than BEFORE concessions…

Average annual deforestation rate (%/yr)

2000-2010 Before

concession After

concession

Oil palm, undated -1.62%

Oil palm, dated -2.35% -1.1% -2.5%

Logging, undated -

Logging, dated -0.47% -0.38% -0.51%

Timber, undated -0.28%

Timber, dated -2.98% -2.6% -3.5%

Protected, undated +1.46%

Protected, dated -0.40% -1.4% -0.37%

Other forest -0.77%

Total forest -0.94%

Non-forest +1.25%

2.3x (1.3x) more deforestation AFTER oil palm (timber) concessions than BEFORE concessions…

…but deforestation was increasing nationwide

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

1.2%

1.4%

Def

ores

tatio

n ra

te (%

/yr)

Year

Average annual deforestation rate (%/yr)

To attribute deforestation to the designation of oil palm concessions, we need fixed effects regression to… -control for year-specific effects -control for site-specific effects matching methods? -not necessary due to panel data

Designation Average effect of concession designation on deforestation (ha/yr), controlling for year fixed-effects and site fixed-effects

Timber concession +110% (109-111%)

Oil palm concession +60% (57-64%)

Logging concession +61% (59-63%)

Protected area - 4.4% (1.2-7.5%)

Poisson regression; number of 3km x 3km site-years = 1,268,690





Oil Palm Concessions Total: 10.6 mha

post-2000

pre-2000

undated

Timber Concessions Total: 8.6 mha

Logging Concessions Total: 13.2 mha

Protected Areas Total: 14.8 mha

Aggregating impact to national level Scope of moratorium Reduction in

deforestation Reduction in

emissions from deforestation

New oil palm concessions in high-carbon forests (>150tC/ha) and peat lands

117,000 ha/10yrs (-1.3%)

153 MT/10yrs (-2.2%)

New oil palm + timber concessions in high-carbon forests (>150tC/ha) and peat lands

414,000 ha/10yrs (-4.7%)

578 MT/10yrs (-8.3%)

New oil palm + timber concessions in all forest

550,000 ha/10yrs (-6.3%)

628 MT/10yrs (-9.0%)

New oil palm + timber + logging concessions in all forest

628,000 ha/10yrs (-7.2%)

676 MT/10yrs (-9.6%)

New + existing oil palm + timber + logging concessions in all forest

1,486,000 ha/10yrs (-16.9%)

1367 MT/10yrs (-19.5%)

Uninformed extrapolation of concession dates to undated concessions; no leakage Caveats: No exemptions; no temporal shifting




3. How much would Indonesia’s moratorium on concessions have

reduced emissions if applied from 2000-2010? 4. What carbon price would have achieved an equivalent reduction?

Comparing incentive structures for

REDD+ using OSIRIS-Indonesia (Busch et al, PNAS, 2012)

Click-of-a-button decision support tool to

estimate and map the impacts of alternative REDD+ policy decisions on: -deforestation (ha/yr) -emission reductions (tCO2e/yr) -national and local revenue ($/yr) Benefits: -free -MS Excel interface -transparent -open-source -peer-reviewed, published, scientific -online: http://www.conservation.org/osiris

1. OBSERVED DEFORESTATION, 2000-2005 (Hansen, 2008)

Deforestation: 687,000 ha/yr Emissions: 860 million tCO2e/yr

2. LIKELY DEFORESTATION WITHOUT RED (unofficial “reference scenario”)

Deforestation: 693,000 ha/yr Emissions: 803 million tCO2e/yr

KALIMANTAN

JAVA

SUMATRA SULAWESI PAPUA

KALIMANTAN

JAVA

SUMATRA SULAWESI

PAPUA

3. LIKELY DEFORESTATION WITH RED ($10/tCO2e) Deforestation: 557,000 ha/yr

Emissions: 581 million tCO2e/yr Revenue: $2.2 billion.yr

KALIMANTAN

JAVA

SUMATRA

SULAWESI

PAPUA

SUMATRA

KALIMANTAN

JAVA

SULAWESI

SULAWESI

PAPUA

Carbon price needed for equivalent reduction

Scope of moratorium Reduction in emissions from deforestation

Voluntary incentives (simple)

Voluntary incentives (improved)

Mandatory incentives (e.g. cap-and-trade)

New oil palm concessions in high-carbon forests (>150tC/ha) and peat lands

153 MT/10yrs (-2.2%)

$2.60/tCO2e $1.10/tCO2e $0.50/tCO2e

New oil palm + timber concessions in high-carbon forests (>150tC/ha) and peat lands

578 MT/10yrs (-8.3%)

$9.40/tCO2e $2.75/tCO2e $2.05/tCO2e

New oil palm + timber concessions in all forest

628 MT/10yrs (-9.0%)

$10.20/tCO2e $2.95/tCO2e $2.25/tCO2e

New oil palm + timber + logging concessions in all forest

676 MT/10yrs (-9.6%)

$11.00/tCO2e $3.15/tCO2e $2.45/tCO2e

New + existing oil palm + timber + logging concessions in all forest

1367 MT/10yrs (-19.5%)

$24.40/tCO2e $7.05/tCO2e $5.60/tCO2e

(OSIRIS v1.5; Busch et al, PNAS, 2012)


19% in oil palm concessions; 26% in timber concessions 2. How much did the designation of a concession between 2000-2010


on average, oil palm concessions increased deforestation by 60%, and timber concessions increased deforestation by 110%, controlling for year- and site- specific effects

3. How much would Indonesia’s moratorium on new concessions have reduced emissions if applied from 2000-2010? by 578 MtCO2e/10 yrs (8.3%) assuming no leakage, extrapolating

rates from dated to undated concessions 4. What carbon price would have achieved an equivalent reduction? a carbon price of $2.05 (mandatory) or $9.40 (voluntary)

A moratorium on new oil palm concessions and timber concessions in high-carbon forests (>150 tC/ha) and peat lands would have had a substantial impact on reducing emissions from deforestation:

578 MtCO2e/10yrs (8.3%) Modest carbon gains could have been achieved by

expanding the scope of the moratorium to include secondary forests and new logging concessions

For Indonesia to achieve its 26-41% emission reduction

target, it would have had to expand the scope of the moratorium to address conversion within existing concessions or non-concession areas, or have put price-based instruments in place

Conclusion:

Terima kasih! Thank you!

Thanks to:

World Resources Institute Woods Hole Research Center

David and Lucile Packard Foundation

Comments and feedback welcome: http://www.conservation.org/osiris

[email protected]

REDD+: an overview

Payments ($) Developed countries provide finance

through funds or markets

Emission reductions (tCO2e/yr) •Developing countries reduce 15% of global GHG emissions from deforestation, etc •Co-benefits: biodiversity, clean water…

•UNFCCC sets basic rules •Timeline: agreement by 2015; implementation by 2020 •Forest countries decide how to achieve reductions •Price-based mechanisms or place-based policies

Palm oil in stove Palm oil in food Palm oil in car

Source: World Resources Institute

Data (~200,000 3km x 3km cells) Dependent variables: • Forest cover, 2000, 60m Landsat (Margono, Hansen et al, in prep) • Annual deforestation, 2000-2010, 60m Landsat (Margono, Hansen et al, in prep.) • Oil palm distribution, 2010, 250m Modis (Miettenen et al, 2011)

Explanatory variables: • Oil palm concession boundaries and dates (Ministry of Forestry, WRI, 2009/2010) • Timber concession boundaries and dates (Ministry of Forestry, WRI, 2009/2010) • Logging concession boundaries and dates (Ministry of Forestry, WRI, 2009/2010) • Protected area boundaries (WRI) and dates (various) Emission factors: • Forest biomass, 463m (Baccini et al, 2012) • Soil carbon (FAO, 2008) • Peat distribution and emissions (Wetlands International; Hooijer, 2010)

(1) Basic PES-style voluntary incentives

Site-scale accounting; historical reference levels

(2) District-scale accounting; historical reference levels

(3) District-scale accounting; projected reference levels

(4) District-scale accounting; projected reference levels +20% revenue sharing

(5) Well-structured voluntary incentives District-scale accounting; projected reference levels +20% revenue sharing

+20% responsibility sharing (6) District-scale accounting projected reference levels minus 10% +20% revenue sharing +20% responsibility sharing (7) Mandatory incentives, e.g. Cap & Trade

District-scale accounting; projected reference levels minus 10% 0% revenue sharing 100% responsibility sharing

(8) District-scale accounting; projected reference levels minus 26% 0% revenue sharing 100% responsibility sharing

PES CAT Well-structured Voluntary Well-structured voluntary REDD+

nearly as effective as cap-and-trade

Shortfall Surplus

Geographically prioritizing pilot programs: Expected abatement under REDD+ at $10/tCO2e

KALIMANTAN

JAVA

SUMATRA

SULAWESI

PAPUA

Where is forest carbon, AND where can money change behavior?

Sensitivities • Functional form • Included variables • Policy decisions • Model parameters -Carbon price -Price elasticity of demand for frontier agriculture (intranational leakage) -Exogenous agricultural price increase (international leakage) -Peat emission factor -Carbon data set -Social preference for agricultural revenue -National reference level -District level start-up costs -Per-hectare transaction costs

Economic incentives are just one

important component of a national

REDD+ strategy

The road ahead • Analysis in other regions: Peru, Madagascar, Bolivia, Mexico

• IDRISI GIS interface (w/ Clark Labs)

• Agricultural concessions and policies

• Degradation and reforestation

• Safeguards for REDD+

• Market integrity mechanisms: risk buffers, offset trade ratios,

conservative accounting • Matching payments for biodiversity,

water and other ecosystem services

• Community conservation contracts and green economic development

Documents

Indonesia’s moratorium and emissions from deforestation