Carbon footprints and the role of international trade in climate policy Do developed countries ‘out-source’ carbon emissions to developing countries? Glen Peters Center for International Climate and Environmental Research – Oslo (CICERO), glen.peters@cicero.uio.no
Carbon footprints and the role of international trade in climate policy Do developed countries ‘out-source’ carbon emissions to developing countries? Glen Peters Center for International Climate and Environmental Research – Oslo (CICERO), glen.peters@cicero.uio.no
Value added and the role of international trade in policy Do developed countries ‘out-source’ labour to developing countries? Glen Peters Center for International Climate and Environmental Research – Oslo (CICERO), glen.peters@cicero.uio.no
Outline • Background and framing • Defining “carbon leakage” • Research Questions • Methodology • GTAP-MRIO • Model comparisons • Applications • Chronological perspective
BACKGROUND AND FRAMING
Global Carbon Emissions from Fossil-Fuels, Cement, and Flaring Growth rate UNFCCC perhaps not 2010 meeting its objectives? 5.9% yr Growth rate 2000-2010 3.1% per year 2010 : Emissions: 9.1 PgC Growth rate 2009 Growth rate: 5.9% -1.3% per year Growth rate 1990 levels: +49% 1990-1999 1% per year KP ratified 2000-2010 Growth rate: 3.1% - Kyoto UNFCCC
The Kyoto Protocol View Legend: Developing countries Production (bottom, solid) critical in the future Consumption (top, dotted) Developed Countries … or not? (Annex B) The Kyoto Protocol has been a success! Developing Countries (non-Annex B)
Success or disaster? • Method of accounting critical to the outcome • Production/Territorial • Consumption for global pollutants? • Increase in net import five times greater than emission reductions (1990-2008) • Is this “carbon leakage”?
• Strong carbon leakage • Increase in emissions in non-Annex B countries as a result of mitigation policies in Annex B • Weak carbon leakage ( carbon migration ): • Increased production in non-Annex B to meet consumption in Annex B
Strong Carbon Leakage • Static CGE analysis • Put a carbon tax in Annex B countries • Estimate emission reductions in Annex B compared to emission increase in non- Annex B • Highly dependent on assumptions • Generally small (~5-20%, possibly >100%) • Difficult to measure (other factors) • Perhaps a few %
Weak Carbon Leakage • Dynamic analysis (usually historic) • Increase in emissions in non-Annex B due to consumption in Annex B • Not directly measurable • Based on data (usually MRIO) • Robust results, offsets emission reductions • Policy drivers not considered • Forward look studies difficult – Who predicted the growth of China?
Outsourcing vs Offshoring • Outsourcing • Get someone else to do the production • Offshoring (c.f., Pollution Havens) • Relocate activities • Strong carbon leakage (CGE changes) • Changes driven by change in tax regime • Weak carbon leakage (any change) • Changes driven by existing conditions
Broad research questions • What are the global repercussions of economic activity? • Exported and imported emissions • Static analysis, many studies now • Are emission reductions offset by increases in imported emissions? • Dynamic, backward looking analysis • Policy drivers are a different question
METHODOLOGY
GTAP-MRIO • Based on the GTAP database – Constructed for CGE analysis of agriculture and trade policies – 1997 (78 regions), 2001 (87), 2004 (112,129), 2007 (129) – 57 sectors WIOD - Groningen, 26/04/12
• “True” MRIO, no survey information • GTAP does not construct MRIO, but the “components” of an MRIO – Other MRIO’s construct an unbalanced MRIO, and then balance • Extremely fast to construct and solve – 1sec to construct MRIO from inputs – 6sec to normalise and invert MRIO WIOD - Groningen, 26/04/12
Three modes of operation • Emissions embodied in bilateral trade (EEBT) • Domestic emissions to produce exports • Simple, fast, transparent • GTAP-MRIO exogenous int. transport • GTAP-MRIO endogenous int. transport • True “carbon footprint” • Considers global supply chains
Three modes of operation • Same issues as in value added • Emissions embodied in bilateral trade (EEBT) • Full MRIO model (global supply chains)
UNCERTAINTY
“Results are too uncertain” • Obvious assumption • …but, not supported by data • Sources of uncertainty • Input emissions data • Definitions of “carbon footprint” • MRIOT • Smaller than most would think
Norwegian emissions 60 CDIAC Carbon dioxide (MtCO 2 ) Statistics Norway 50 40 30 20 Emissions data is 10 highly uncertain 0 1960 1970 1980 1990 2000 2010
Dutch emissions 200 190 EDGAR Energy emissions (MtCO 2 ) CDIAC 180 IEA_RA IEA_SA 170 NAMEA* 160 UNFCCC 150 Range in estimates 140 130 is about 12% 120 1970 1980 1990 2000 2010
Difference: GTAP vs EDGAR Allocation of emissions data to sectors highly uncertain
Input Emissions
Input Emissions
Definitions: EEBT vs MRIO
Model variation
Model variation
Netherlands 350 Territorial (MtCO 2 ) Peters et al (2011) Wiebe et al (2012) 300 Boitier (WIOD) 250 200 150 1990 1995 2000 2005 2010 350 Peters et al (2011) Consumption Wiebe et al (2012) 300 (MtCO 2 ) Boitier (WIOD) 250 200 150 1990 1995 2000 2005 2010
Netherlands 75 Net import (MtCO 2 ) Peters et al (2011) Wiebe et al (2012) 50 Boitier (WIOD) 25 0 1990 1995 2000 2005 2010 1.30 Peters et al (2011) 1.25 Wiebe et al (2012) Boitier (WIOD) 1.20 Ratio 1.15 1.10 1.05 1.00 1990 1995 2000 2005 2010
Model Comparisons • What causes largest differences? • 20%, Variations in input emissions • 5%, Definition of the Carbon Footprint • >1%, Different MRIOT (needs work) • We may give users the wrong impression about uncertainty • Must ensure we show results consistently
APPLICATIONS
Production – Consumption = -0.5GtCO 2 Update: Production – Consumption = 1.3GtCO 2
Key Findings (2001): • 5.3 GtCO 2 (22%) embodied in trade • Annex B Consumption 0.8 GtCO 2 higher than Production (6%) • OECD Consumption 1.3 GtCO 2 higher than Production (9%) • OECD, 1.3 GtCO 2
Key Findings (2004): • 6.2 GtCO 2 (23%) embodied in trade • Annex B Consumption 1.6 GtCO 2 higher than Production (12%) • OECD Consumption 2.1 GtCO 2 higher than Production (16%)
Changes over time 3500 OECD Net Emission Transfer into OECD 3000 Peters et al (2011) WIOD (Boitier) 2500 countries (MtCO 2 ) Wiebe et al (2012) 2000 1500 1000 500 0 1994 1996 1998 2000 2002 2004 2006
Shares over time ~1995 ~2000 ~2005 OECD 11 16 13 Peters et al (2011) 8 12 16 WIOD (Boitier) 8 12 17 Wiebe et al (2012) 14 17 23 • Compare to territorial emissions, studies more consistent (OECD outlier) • Longer, more consistent study needed
• World split into – 113 countries and regions – 57 sectors – 1990-2008, now 2010 • Focus on – Net emission transfer (exports minus imports) – Annex B (developed) vs. non-Annex B (developing) WIOD - Groningen, 26/04/12
• “Proxy” method – Construct 2 sector IOT using GDP data • GDP by expenditure gives time series • Detailed IOTs in 1997, 2001, 2004, (now 2007) give “proxy” data on imports to industry and emission intensities • Time series of trade data (GTAP) – Motivation is for quick analysis to analyse trends WIOD - Groningen, 26/04/12
Globally : Emissions from the production of traded goods and services 1990: 4.3 Gt CO 2 (20% of global emissions) 2008: 7.8 Gt CO 2 (26%) WIOD - Groningen, 26/04/12
Trade offset Kyoto reductions Net transfer in 2008 was five times achieved Kyoto Reductions …and our method is conservative
Aggregated results
WIOD - Groningen, 26/04/12
Light manufacturing dominates Non-energy intensive manufacturing is most relevant in carbon leakage debates
Extraction to Production Key Findings (2004): Production • 10.2 GtCO 2 (37%) to embodied in extraction • 6.4 GtCO 2 (23%) Consumption embodied in production Extraction to Consumption WIOD - Groningen, 26/04/12
Next steps? • What is causing the changes? • Climate policy? No, unlikely • Reduced trade barriers and costs? • Wage differentials? • General development? • China? • What to do about it? • Is it necessary to know the cause?
Policy implementation? • Consumption-based approaches are complementary to production-based • Step 1: Measure consumption-based emissions and track progress • Step 2: Design policy instruments to address issues • Step 3: Track progress
Recommend
More recommend
