Industrial and terrestrial carbon leakage under climate policy fragmentation

Abstract

One of the main concerns in international climate negotiations is policy fragmentation, which could increase the carbon emissions of non-participating countries. Until very recently the carbon leakage literature has focused mainly on ‘industrial’ carbon leakage. However, there is another potential channel that has received little attention so far: the carbon leakage triggered by land-use change (‘terrestrial’ carbon leakage). In this article we use an integrated assessment model to explore these two forms of leakage in a situation where CO₂ emissions in all sectors, including those from land-use change, are taxed equally. Our results show that under different fragmentation scenarios terrestrial carbon leakage may be the dominant type of leakage up to 2050. When participating regions tax land-use emissions, forest area expands partly by shifting food and bioenergy production to non-participating regions. This reduces forest area in non-participating regions and increases their land-use emissions.

Policy relevance

Preventing industrial carbon leakage has been an important aspect of climate policy design. One clear policy implication of our study is that anti-leakage policy measures should also be considered for land-use change sources.

Keywords:

• AFOLU
• carbon leakage
• climate change
• deforestation
• energy

Acknowledgements

We thank P. Patel for his support with the GCAM model. The usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Calvin (2009) and Kriegler et al. (2015) offer a thorough discussion and quantification of the different channels by which carbon price differentials lead to changes in carbon emissions outside the regions taking domestic mitigation action and Antimiani, Costantini, Martini, Salvatici, and Tommasino (2013) assess alternative solutions to this type of carbon leakage.

2. Detailed information can be found at http://www.globalchange.umd.edu/models/gcam/ or at https://wiki.umd.edu/gcam

3. A full description of the agriculture and land-use module in GCAM can be found in Kyle et al. (2011) and Wise and Calvin (2011).

4. GCAM assumes exogenous crop productivity improvements along the century, although the implications of different climate change scenarios have been explored recently (Kyle et al., 2014). These values range from 0.25% per year to 1.4% per year depending on the region and year. After 2050, the default agricultural productivity grows at 0.25% per year. Previous work (Wise et al., 2009) has shown that the choice in productivity growth estimates has a significant impact on cropland area and land-use change emissions.

5. The 10 crop categories are corn, rice, wheat, other grains, sugar, root tuber, palm fruit, fibre crops, oil crops and other crops.

6. The six animal product categories are beef, dairy, pork, poultry, sheep, goat and others.

7. For a detailed representation of the implications of emission reduction in emission per capita in each scenario see Figure A1 in the appendix. Notice that as emissions from non-participant regions are included in the computation of global emissions, global emissions per capita are always above 0.5 tons of CO₂.

8. Non-CO₂ emissions are considered in the model and are linked to underlying human activities, but we do not impose a tax on them. If a tax on non-CO₂ were to be implemented, it would probably lead to higher levels of terrestrial carbon leakage given the relevance of methane emissions in the agricultural and livestock sectors.

9. In this study we are always referring to the notion of ‘strong’ carbon leakage, as CO₂ emissions embodied in imports are not accounted for in our model (Glen and Hertwich, 2008).

10. All the carbon that is stored in a forest converted to cropland is released at once, but it takes decades for afforestation to build up all the carbon storage potential in the new forests.

11. This effect can be observed in Figure A4 in the appendix, where regional disaggregation of carbon leakage is presented for each of the scenarios.

12. There is a shift in the model to those types of forests that have more potential to store carbon.

13. GCAM model (and in general most of the models of RCP database) assume a reduction of CO₂ from land-use change in the short term.

14. The index of fossil fuel includes the global price variation of crude oil, natural gas and coal. Each element is weighted according to the proportion of energy (EJ) in the primary energy mix.

15. The energy demand functions in the GCAM model include a set of parameters that capture the elasticity of the energy demand to the changes in either the prices of fuels (fuel price elasticity) or the prices of energy services (service price elasticity).

16. Figure A4 in the appendix shows that a substantial part of the global electricity mix in 2050 is still covered by fossil fuel with CCS technology (up to 12%). The greatest expansion due to mitigation policies is in nuclear power (up to 31%). The figure also shows that biomass does not have a prominent role in the energy system and the use of biomass with CCS is marginal in all the scenarios along the century.

17. See Figure A5 in the appendix.

18. Note that there are two independent CO₂ markets for developing regions and another for developed regions. This means that in those cases there will be two different CO₂ prices. CO₂ prices converge at the same time the emissions per capita converge.

19. Non-participant’s regions should benefit from fossil fuel price reduction and also from the increase in food/biomass production, but these positive effects are not captured in GDP as GDP is defined exogenously in the model.

Additional information

Funding

This work was supported by the European Union (FP7 program) under the CECILIA2050 project [grant number 308680]. We also thank financial support from Science and Innovation Ministry of Spain (ECO2015-68023) and Basque Government (IT-799-13). Iñigo Capellán-Perez thanks the REPSOL Foundation for support through the Low Carbon Programme.