Ecological Economics 84 (2012) 57–65

Contents lists available at SciVerse ScienceDirect

Ecological Economics

j ourna l homepage: www.e lsev ie r .com/ locate /eco lecon

Income-based environmental responsibility

Alexandra Marques a,⁎, João Rodrigues a, Manfred Lenzen b, Tiago Domingos a

a IN+, Center for Innovation, Technology and Policy Research, Energy and Environment Scientific Area, Mechanical Engineering Department, Instituto Superior Técnico,Technical University of Lisbon, Av. Rovisco Pais n°1, 1049–001 Lisbon, Portugalb Integrated Sustainability Analysis, School of Physics A28, University of Sydney, NSW 2006, Australia

⁎ Corresponding author. Tel.: +351 218419441.E-mail address: alexandra.marques@ist.utl.pt (A. Ma

0921-8009/$ – see front matter © 2012 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.ecolecon.2012.09.010

a b s t r a c t

a r t i c l e i n f o

Article history:Received 23 March 2012Received in revised form 19 September 2012Accepted 23 September 2012Available online 23 October 2012

Keywords:Carbon responsibilityDownstream responsibilityIncome responsibilityMulti-regional input–output analysis

A strongmitigation effort is underway to reduce the levels of anthropogenic greenhouse gas emissions. An alloca-tion of the mitigation effort among multiple agents requires understanding which agent is responsible for whatemissions, according to a defined measure of carbon responsibility. The metric adopted by current internationalclimate policy is production-based (or territorial) responsibility. However, other types of responsibility havebeen discussed in the literature, namely consumption-based (or upstream) responsibility and downstream re-sponsibility. In this paperwe study the latter, which is little explored in the literature. We clarify the term throughanovel nomenclature, income-based responsibility andpresent a case-study,with the quantificationof income-basedresponsibility for 112 world regions, and the comparison of the results with production and consumption-basedresponsibilities.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

In the economic process primary factors of production, such asnatural resources, labor or capital are applied in production processesto deliver consumption goods and services. Typically these processesgenerate adverse environmental pressures. There is a widespreadperception that the current levels of consumption and the consequentenvironmental pressures are unsustainable, leading to environmentalproblems, such as changes in climate patterns, scarcity of drinkingwater or degradation of arable land. In order to control and minimizethe negative externalities that result from the economic process it isnecessary to quantify environmental pressures and to allocate re-sponsibility for them to the economic agents involved.

Climate change is currently a priority area of environmental policy(EEA, 2010; OECD, 2008; UNEP, 2007), with a lot of attention focusedon anthropogenic greenhouse gas (GHG) emissions. Many scientistsbelieve that in order to reduce these risks a substantial reduction ofGHG emissions is needed (mitigation).

The global nature of climate change requires global action; for exam-ple, if one large emitter does not commit to its responsibility andmitiga-tion target, it is unlikely that the rest of the world can compensate for it.Moreover any emitter that stays out of any agreement will benefit fromthe action taken by others;mitigation efforts to copewith climate changecan be considered public goods which allow for free-riding phenomena(Stern, 2007), potentially impairing or delaying climate policy. The extentof global participation is critical because the higher the participation ratein any action taken, the lower the cost (OECD, 2003). One of the goals of

rques).

rights reserved.

the current climate debate is to design fair and equitable climate policiesthat are globally accepted.

In this work, we explore a principle of responsibility seldomaddressed in the literature, downstream responsibility. First, Section 2describes production-based responsibility, the metric adopted by currentclimate policy. Section 3 describes consumption-based responsibility,a frequently proposed alternative to production-based responsibility.Section 4 describes downstream responsibility and presents the rationalefor a novel nomenclature. Section 5 compares the three approaches. InSection 6 a case study is presented. Finally, Section 7 concludes.

2. Production-based Responsibility

Since 1992, in the Earth Summit, international negotiations are inplace seeking a global agreement for the attribution of GHG emissionsresponsibility and reduction. But only in 1997 significant results wereachieved with the Kyoto Protocol. Under the Protocol each countryshould report, through a national GHG inventory, the ‘emissions andremovals taking place within national (including administered) terri-tories and offshore areas over which the country has jurisdiction.’ Forthe Annex I countries who ratified it, a binding target of GHG emissions'reduction was also established: 5% (on average), during the 2008–2012period against 1990 levels (UNFCCC, 1998).

Under the Kyoto Protocol, a country holds responsibility for all theemissions that are directly generated by the production processes thattake place within its borders. This type of responsibility is often calledproducer-based responsibility, and accounts for the direct emissions of acountry (Fig. 1).

The geographic boundary established for GHG inventories does notconsider the transfer of emissions through international trade, enabling

DIRECT EMISSIONS

COUNTRY A

PRODUCTION-BASED RESPONSIBILITY

A

DIRECT EMISSIONS

COUNTRY A

UPSTREAMEMISSIONS EMBODIED

IN IMPORTS

UPSTREAMEMISSIONS EMBODIED

IN EXPORTS

CONSUMPTION-BASED RESPONSIBILITY

B

DIRECT EMISSIONS

COUNTRY A

DOWNSTREAMEMISSIONS ENABLED

BY IMPORTS

DOWNSTREAMEMISSIONS ENABLED

BY EXPORTS

INCOME-BASED RESPONSIBILITY

C

Fig. 1. Schematic representation of a country's (A) production-based responsibility,(B) consumption-based responsibility and (C) income-based responsibility. Arrowsrepresent the flow of emissions.

58 A. Marques et al. / Ecological Economics 84 (2012) 57–65

carbon leakage (Pedersen and de Haan, 2006; Peters et al., 2011b), andcreates difficulties for the assignment of emissions generated through in-ternational activities, like aviation andmaritime transportation (currentlynot included in national inventories) (IPCC, 1997).

The limitations of this approach became evident when big emitters,like the USA and China, refused to ratify the protocol or to commit tobinding targets. Underlying these decisions were issues of fairness andfear of impaired competitiveness among countries (Whalley and Walsh,2009). Developing economies that are highly dependent on exports,like China and India, claim that they should not bear the responsibilityfor production from which they do not benefit in terms of consumption(BBC, 2009). On the other hand, the USA and other developed economies,fear that their economies' competitiveness will be impaired if they haveto cope with any binding target that is not also applied to developingeconomies. Environmental regulations can draw away investors, pro-moting the relocation of industries to environmentally unregulatedeconomies (the pollution haven hypothesis) and enhancing any poten-tial carbon leakage (Stern, 2007). These positions have not changed, ascould be seen in the COP17 meeting, held in Durban in 2011. In thesenegotiations, USA, China and India acknowledged and committed tothe need of curbing carbon emissions, however they did not establishany target. On the other handCanadawithdrew from theKyoto Protocolwhile Japan and Russia refused to sign for a second commitment period(UNFCCC, 2012).

To address the issues of competitiveness and carbon leakage, thereare two often suggested mechanisms: they are, an integrated emis-sions trading system and carbon-motivated border tax adjustments

(Alexeeva-Talebi et al., 2008; Ismer and Neuhoff, 2004). In an inte-grated emissions trading system firms exporting to participatingcountries are included, and requested to obtain emissions certifi-cates for the CO2 content of their exports (Alexeeva-Talebi et al.,2008). Carbon-motivated border tax adjustments consist in tariffsimposed on imports from foreign producers that should reflect thecost of carbon as if the goods imported were to be produced in theparticipating country (Ismer and Neuhoff, 2004; Whalley, 2009).This mechanism also allows rebates for the domestic carbon tax onexported goods (Alexeeva-Talebi et al., 2008; Ismer and Neuhoff,2004). Thus, participating countries would be partially refunded fromtheir carbon abatement costs, whereas non-participating countriesexporting to participating countries would face a penalty (Ismer andNeuhoff, 2004) and thus have an incentive to join the scheme. The effec-tiveness of thesemeasures aswell as their compatibility with theWorldTrade Organization are issues still under debate (Alexeeva-Talebi et al.,2008; de Cendra, 2006; Ismer and Neuhoff, 2004; Whalley, 2009).

Another option to overcome production-based responsibility limita-tions is to consider a measure that includes indirectly generated emis-sions (Eder and Narodoslawsky, 1999; Peters and Hertwich, 2008b).Indirect emissions can be generated due to two distinct and oppositedriving forces: supply and demand (Eder and Narodoslawsky, 1999).The responsibility principles that result from this are describe in thenext Sections.

3. Consumption-based Responsibility

Consumption-based responsibility has been discussed by severalauthors, as an alternative to production-based responsibility that couldaddress its main limitations, namely carbon leakage (for example,Ahmad and Wyckoff, 2003; Andrew and Forgie, 2008; Bastianoni et al.,2004; Davis and Caldeira, 2010; Eder and Narodoslawsky, 1999;Munksgaard and Pedersen, 2001; Peters and Hertwich, 2008a; Peterset al., 2011b).

This metric measures the emissions generated to produce a country'sfinal demand for goods and services. These equal the emissions stemmingfromwithin the national territoryminus the domestic emissions requiredto generate exports plus the foreign emissions required to generate im-ports (Fig. 1). So in fact, consumer-based responsibility is a carbon tradebalance, as was pointed out by Rodrigues et al. (2010), Serrano andDietzenbacher (2010) and Kanemoto et al. (2012).

For a certain product, thismetric takes into account all the emissionsgenerated along its supply chain prior to the delivery to final demand;for that reason these are often called upstream embodied emissions.

The adoption of this type of responsibility is supported by China.Around 20% of China's emissions are generated in the production of ex-ports, therefore China has claimed that a fair agreement should takeinto consideration that those emissions take place to produce goodsthat are not consumed by the Chinese people: ‘We produce productsand these products are consumed by other countries, especially the de-veloped countries. This share of emissions should be taken by the con-sumers but not the producers’ (BBC, 2009).

Unlike production-based inventories, consumption-based responsi-bility accounts for the emissions generated through international trade,minimizing the effects of carbon leakage by holding countries responsi-ble for the emissions embodied in their trade balance (Andrew andForgie, 2008; Bruckner et al., 2009; Marques et al.; Peters, 2008).

A measure of responsibility based on consumption takes into con-sideration international trade. In the context of climate policy the appli-cation of this responsibility principle would work as a type of “borderadjustment,” but not interfering with WTO regulations, since it isnot formally a trade-measure (Peters, 2008). However, consumption-based GHG inventories can be used to measure the carbon content ofimported goods and services and aid in the implementation of bordertax adjustments that would reflect the costs of a certain product interms of carbon (Peters, 2008).

59A. Marques et al. / Ecological Economics 84 (2012) 57–65

4. Income-based Responsibility

Anothermetric of environmental pressure is downstream responsibil-ity (Gallego and Lenzen, 2005; Lenzen and Murray, 2010; Lenzen et al.,2007; Rodrigues and Domingos, 2008; Rodrigues et al., 2006, 2010).This metric measures the emissions enabled by primary suppliers andthat are required to generate a country's income through wages, profitsand rents (payments to primary factors of production). These equal theemissions generated within national boundaries minus the domesticemissions generated downstream of imported products plus the foreignemissions generated downstream of exported products (Fig. 1).

Froma technical point of view, the accounting of downstreamrespon-sibility has the same characteristics as consumption-based responsibility;it also accounts for the emissions generated through international tradeand thus can be used to minimize the effects of carbon leakage.

Downstream responsibility also works as a carbon trade balance(Rodrigues et al., 2010). For a certain product, this measure accountsfor all the emissions generated downstream in its supply chain until de-livery to final demand; for that reason (Lenzen and Murray, 2010)named these types of emissions downstream enabled emissions.

There are applications of the concept at the country level (Lenzen andMurray, 2010;Marques et al.; Rodrigues et al., 2010), and at the corporatelevel (Lenzen andMurray, 2010;WBCSD andWRI, 2010). For example, atthe country level Rodrigues et al. (2010) analyzed the downstream re-sponsibility for six world regions, and showed that developed economiesand fossil fuel exporters are the regions who enabled more emissions togenerate their income, which indicates that these regions are receivingan economic benefit from emissions that did not occur within their bor-ders. Lenzen and Murray (2010) studied the downstream responsibilityassociated with the provision of inputs to a certain number of Australianindustries; this study was the first quantification of downstream respon-sibility at the industry level. Moreover Lenzen and Murray (2010) alsoprovided a substantial clarification of the term, by providing a matchbetween the less known downstream-based vocabulary and the well-known consumption-based vocabulary. At a corporate level a recentstudy (Schücking et al., 2011) showed how the supply of capital bybanks, through investment decisions, enabled the emissions of GHG.The results showed that the self-proclaimed greener banks are actuallythose who enable more GHG emissions through investments in thecoal industry (Petherick, 2012; Schücking et al., 2011).

Downstream responsibility has never received the same discussionas its consumption-based cognate. The current market-based economyputs emphasis on the consumption process since its main driving forcesare consumers' needs (Lenzen et al., 2007). Therefore, accepting thatwhobenefits fromemissions are the consumers is a natural choice. Con-sumers can opt for productswhose production causes less environmen-tal impacts, pushing companies to adopt cleaner production processes.But the price of suchproducts is likely to differ from the price of the dirt-ier alternative (Laroche et al., 2001). Thus, the expression of environ-mental preferences through purchases cannot be dissociated from thebudget constraint allocated to consumption and, ultimately, to theconsumer's source of income.

What if more environmentally efficient products are the most ex-pensive ones? The truth is that most of the time the cost of a productinfluences the consumer choice because it will reduce its real in-come. People want to consume more and for that need to generatemore income. This is done through the supply of primary factors ofproduction.

In a downstream approach the emphasis is put on how this moneyis earned. The benefit from the generation of the emissions is passed tosuppliers in the form of income. Actual examples of applications of thistype of responsibility are easy to find. For example, if we accept thatcigarette companies should compensate their customers for anyhealth problem they might incur due to cigarette consumption, weare following a downstream responsibility principle. Or, for example,if we believe that banks and other financial institutions should follow

the Equator Principles (EP, 2006), then we are also following a down-stream responsibility principle.

However, the size of the literature and research on this topic is mod-est. We believe that this situation results from lack of clear intuition onthe concept due to a naturally oriented market-based thinking. Here weaim to clarify and illustrate the potential applications of the concept ofdownstream responsibility, continuing the work of Lenzen and Murray(2010).

Will the general public recognize the term ‘upstream responsibility’as meaning ‘consumption-based responsibility’? The term upstream,per se, indicates the direction from where carbon responsibility arrivesto an agent, but it says nothing about the economic process that causedthe emissions. Thus this term was replaced in the literature by a moreintuitive one: consumer-based responsibility. Downstream responsibil-ity also describes from where carbon responsibility arrives to an agent,but it also says nothing about the economic process that caused theemissions. In this case the process in question is the supply of primaryinputs to production processes that enable emissions to occur. There-fore the term downstream responsibility could be replaced by supply-based responsibility. However if we agree that responsibility should beplaced on those that take some benefit from carbon emissions, supply-based responsibility is not informative regarding that benefit, whereasconsumption-based responsibility is. The supply of primary inputs, to aproduction process involves the provision of labor (or any other primaryfactor of production) in return of a salary, or, in a broader view, an income.Therefore we propose the term income-based responsibility to addressdownstream responsibility.

Whereas the use of consumption-based responsibility as a metric tolevy a carbon tax would be a tax on consumption, the equivalent use ofincome-based responsibility would function as an income tax. For ex-ample, such a tax would reduce more strongly the income earned by ashareholder of a coal-fired power plant than the income earned by aninvestor in renewable energy. This assuming that producers are not fac-ing perfectly inelastic demand and that therefore the amount of the taxis not completely passed on to consumers. The same logic applies tocountries. The way the income is generated, in a country whose mainactivity is oil extraction, would enable more emissions, thus resultingin a higher responsibility (and thus a higher tax rate), than if thatcountry's main activity was fruit production.

A recent report by IDE-JETRO and WTO (2011) draws attention tothe drastic changes that occurred in the structure of internationaltrade in the last decades. Many products are no longer made in a singlecountry, but instead production chains have become fragmented, withdifferent countries specializing in specific stages of the supply chain,leading to a move from trade in goods to a trade in tasks (Hummels etal., 2001; IDE-JETRO andWTO, 2011). This shift increases the trade vol-ume of intermediate goods, which are re-exported several times duringthe processing stage, before reaching the country of assembly into afinal good, which can itself be exported.

The emergence of global production chains changed the paradigmof international trade, from a situation in which the last step in a sup-ply chain accounted for the most added value to a situation in whichit only represents a small fraction (IDE-JETRO andWTO, 2011). Underthis current paradigm new statistical metrics, complementary to thetraditional ones, need to be implemented in order to provide a clearview of international trade. IDE-JETRO and WTO (2011) propose theuse of international trade of value added. This measure enables thecorrect determination of the relative importance of each region thattakes part in a global supply chain. This report and the new frameworkit presents opens the door for wider applications of income-basedresponsibility.

5. Comparison of Responsibility Principles

The extent to which an agent is responsible for a certain environmen-tal pressure depends on the responsibility principle chosen. Different

60 A. Marques et al. / Ecological Economics 84 (2012) 57–65

principles will yield different results, mainly due to differences in theboundaries adopted (Eder and Narodoslawsky, 1999).

In production-based responsibility the agent is only responsible forthe direct pressures it generates. In consumption and income-based re-sponsibility the boundary is expanded to include all the pressures,direct and indirect, that occur respectively upstream of the activityunder consideration (consumption-based), or downstreamof the activ-ity under consideration (income-based) (Eder and Narodoslawsky,1999; Lenzen and Murray, 2010; Rodrigues et al., 2006; Zhou et al.,2010).

Production-based responsibility accounts for emissions generatedthrough production processes. The agent who directly emits shouldhold responsibility for its emissions. In the consumption-based re-sponsibility all the emissions that were generated upstream alongthe supply chain to deliver goods and services to final demand areaccounted. The agent holding the responsibility is the consumer,who is viewed as the agent who benefits from emissions; had con-sumers not demanded goods and services the emissions would nothave occurred. The income-based perspective accounts for all theemissions that where generated downstream along the supply chaindue to the supply of primary factors of production. The agent holdingthe responsibility is the supplier of primary factors of production,who benefits from emissions because it receives a payment fromthe inputs supplied. Both producer and income-based approachessee income as the benefit taken from emissions. However the bound-aries differ; for the former only the production process is analyzed,accounting for direct effects; whereas in the latter the whole supplychain is taken into consideration and both direct and indirect effectsare accounted for. Consumption and income-based accounting workas a carbon trade balance. For an agent consumption-based responsi-bility considers the emissions embodied in the goods it imported, andleaves out the emission embodied in the goods exported (Kanemotoet al., 2012; Rodrigues et al., 2010; Serrano and Dietzenbacher,2010). For income-based responsibility, the emissions enabled bythe goods it exported are considered whereas the emissions enabledby the goods imported are left out (Rodrigues et al., 2010). As a result,agents who embark on international exchanges typically exhibitmore different results depending on the responsibility principle ana-lyzed. In Table 1 a succinct comparison between these responsibilityprinciples is presented.

As mentioned, the adoption of a responsibility principle will yielddifferent outcomes for different agents, making them better or worseoff depending on the principle chosen. If the production-based respon-sibility principle is applied exporting countries will display higher re-sponsibilities than importing countries. This means, for example, thatrapidly developing economieswould hold responsibility formore emis-sions than developed economies (see, for example, Davis and Caldeira,2010; Peters and Hertwich, 2008a; Rodrigues et al., 2010). As men-tioned in Section 2 the fairness of this measure has been questioned(Whalley and Walsh, 2009; Zhou et al., 2010).

Shifting to a consumption-based approach would mean that finalconsumers, instead of producers, assume total responsibility for theemissions generated. Highly importing countries would hold moreresponsibility than highly exporting countries. Generally speaking,developed countries, with higher levels of consumption, would havehigher responsibility than developing countries (see, for example, Davis

Table 1Comparison between different types of approaches to GHG emissions responsibility.

Responsibility Production Consumption Income

Effects Direct Indirect IndirectEmissions Direct Embodied EnabledDirection - Upstream DownstreamScope Production Demand SupplyAgent Producer Final consumer Primary suppliers

and Caldeira, 2010; Peters and Hertwich, 2008a; Rodrigues et al., 2010).This metric is seen as fairer by developing economies than production-based responsibility, as it would force developed economies to changetheir lifestyle patterns in order to meet their GHG goals (Zhou et al.,2010). To date and up to our knowledge, there is no concrete policy rec-ommendation based on consumption-based responsibility. However, webelieve that if such apolicywould be implemented itwould raise issues offairness as production-based responsibility raised. For example, consum-ing countries provide an economic benefit to producing countries that isnot accounted for if a strictly consumption approach is followed.

Through an income-based approach, it would be primary sup-pliers, instead of producers or consumers who would assume total re-sponsibility for the emissions generated. Countries exporting goodswho will enable emissions in other regions will be worse off thancountries whose exports do not enable emissions. This metric wouldpush countries to analyze how they generate their income. It is con-tradictory that a country that domestically promotes clean technolo-gies generates the majority of its income through the exports of fossilfuels. It is reasonable to accept that it is fair that those benefitingmore, in terms of income, from emissions should bear the responsibil-ity. However, in terms of climate policy, any measure placing full re-sponsibility on one actor is unlikely to be accepted, since it willnever be perceived as fair for all the agents involved in negotiations.

One option to overcome this issue is the adoption of a shared respon-sibility approach. With this type of approach responsibility is distributed,according to a pre-defined criteria, amongst the different agents. Severalcombinations of producer, consumer and income-based responsibilityare possible. Rodrigues et al. (2010) provide an extensive review of thistopic. For example, Bastianoni et al. (2004) developed a shared responsi-bility indicator, the Carbon Emissions Added (CEA). For a country thiswould equal their producer responsibility, plus total upstream embodiedemission, normalized by the total of emissions worldwide. Ferng (2003)proposed a shared responsibility indicator, according to which a countrywould be responsible for a share of its production-based responsibilityplus a share of its consumption-based responsibility. Ferng (2003) statesthat for each country a fair share should be established but no alloca-tion rule was determined. Gallego and Lenzen (2005) and Lenzen et al.(2007) developed a shared responsibility framework that assignsresponsibility either to producers and consumers (production-basedand consumption-based responsibilities) or producers and workers-investors (production-based and income-based responsibility). Theshare of responsibility allocated to each agent should be proportionalto its value added (Lenzen et al., 2007). Rodrigues et al. (2006) axiom-atically derived a shared responsibility indicator that is the average be-tween an agent's consumption-based responsibility and income-basedresponsibility.

There are some empirical results on shared responsibility (for exam-ple, Andrew and Forgie, 2008; Cadarso et al., 2012; Ferng, 2003; Lenzen,2007), focusing on producers and consumers. We believe that the lackof results on shared responsibility including suppliers is due to thelack of understanding of this approach, as referred in Section 4.

The construction of a GHG inventory based on the production-basedresponsibility principle, or on the consumption or income-based princi-ples, requires different types of data. Production-based inventories usedata that are already available through statistical offices (IPCC, 1997),and therefore present less uncertainty (Peters, 2008). Consumption-based and income-based inventories needmore data, that is not usuallyavailable through statistical offices, namely bilateral trade data, andthus have more uncertainty associated (Peters, 2008; Peters et al.,2011b). As a consequence, a production-based responsibility measureis probably seen as more transparent and straightforward than a mea-sure based on consumption or income-based responsibility. To over-come this, the scientific community has been committed to providereliable data that can be used to compute the latter types of responsibil-ity. Examples of this are the recently launched global databases: EORAfrom the University of Sydney (Lenzen et al., 2012), EXIOPOL and

Table 2Producer, consumer, and income-based responsibilities, population and GDP of GTAPregions (year 2004) (Mt CO2, million, thousand million USD2004).

GTAPcode

Region's name Producer Consumer Income Pop. GDP

AUS Australia 315.27 305.47 376.61 19.94 637.79NZL New Zealand 28.34 34.30 29.45 3.99 96.44XOC Rest of Oceania 17.22 17.75 15.56 8.71 21.28CHN China 4071.13 3147.11 3450.95 1307.99 1674.13HKG Hong Kong 54.70 97.23 92.20 6.96 163.01JPN Japan 924.98 1214.09 1058.83 127.92 4658.74KOR South Korea 344.35 335.40 316.03 47.64 676.50TWN Taiwan 220.70 167.41 199.11 22.76 305.29XEA Rest of East Asia 75.80 57.27 65.68 25.36 25.59KHM Cambodia 2.81 3.71 2.19 13.80 4.88IDN Indonesia 295.57 261.55 307.17 220.08 254.70LAO Lao PDR 1.40 2.00 1.11 5.79 2.45MYS Malaysia 125.32 68.91 145.18 24.89 114.90PHL Philippines 67.38 72.68 52.22 81.62 84.48SGP Singapore 38.20 58.33 46.76 4.27 106.81THA Thainland 192.72 144.05 148.13 63.69 161.70VNM Vietname 72.93 67.97 55.31 83.12 43.03XSE Rest of Southeast

Asia7.44 9.08 18.02 51.30 5.59

BGD Bangladesh 28.78 41.11 25.00 139.21 55.91IND India 919.76 860.79 760.23 1087.12 641.26PAK Pakistan 111.19 126.67 88.15 154.79 94.73LKA Sri Lanka 10.86 15.63 9.25 20.57 20.08XSA Rest of South Asia 8.35 13.97 7.77 56.32 13.90CAN Canada 460.01 424.99 512.81 31.96 979.13USA United States of

America4879.14 5511.71 4650.48 295.41 11673.38

MEX Mexico 327.08 353.65 189.41 105.70 683.24XNA Rest of North

America3.15 4.95 1.49 0.13 5.89

ARG Argentina 118.20 88.41 124.44 38.37 150.40BOL Bolivia 8.96 8.52 9.39 9.01 8.78BRA Brazil 234.81 215.53 241.66 183.91 616.54CHL Chile 54.98 44.07 52.67 16.12 89.64COL Colombia 45.19 48.14 62.14 44.92 97.46ECU Ecuador 17.31 21.09 26.82 13.04 29.97PRY Paraguay 2.87 4.55 4.79 6.02 8.42PER Peru 25.09 30.06 27.98 27.56 68.63URY Uruguay 4.02 6.30 3.39 3.44 13.69VEN Venezuela 123.52 88.30 179.07 26.28 108.23XSM Rest of South

America1.86 2.37 1.23 1.39 3.52

CRI Costa Rica 4.14 6.39 5.11 4.25 19.47GTM Guatemala 8.47 13.49 6.47 12.29 27.45NIC Nicaragua 3.51 4.56 1.85 5.38 4.39PAN Panama 4.87 7.86 5.24 3.18 12.60XCA Rest of Central

America11.00 16.51 8.51 14.07 24.15

XCB Caribbean 142.85 139.65 89.92 38.45 193.12AUT Austria 52.27 82.86 68.14 8.17 292.31BEL Belgium 72.39 124.15 85.20 10.42 352.31CYP Cyprus 7.05 9.24 5.45 0.83 15.42CZE Czech Republic 99.41 81.15 81.68 10.23 108.03DNK Denmark 44.27 62.00 65.20 5.41 243.73EST Estonia 15.03 13.55 10.68 1.34 10.22FIN Finland 57.67 69.19 64.69 5.24 185.92FRA France 255.58 410.46 302.32 60.26 2046.47DEU Germany 599.25 804.45 733.92 82.65 2740.50GRC Greece 74.78 94.89 77.69 11.10 205.20HUN Hungary 42.71 52.20 37.33 10.12 99.65IRL Ireland 33.97 46.59 52.95 4.08 182.24ITA Italy 332.60 476.05 340.80 58.03 1677.82LVA Latvia 6.45 11.84 5.85 2.32 13.47LTU Lithuania 9.42 14.51 9.15 3.44 21.20LUX Luxembourg 9.73 11.25 9.83 0.45 31.86MLT Malta 2.73 3.43 1.56 0.40 5.32NLD Netherlands 165.81 172.01 155.14 16.23 578.98POL Poland 240.70 212.64 202.00 38.56 233.62PRT Portugal 50.19 64.34 45.92 10.44 167.72SVK Slovakia 24.67 25.91 20.51 5.40 41.55SVN Slovenia 12.62 13.93 12.97 1.97 32.52ESP Spain 266.76 324.69 258.40 42.65 1039.90SWE Sweden 37.41 69.97 64.65 9.01 346.41GBR United Kingdom 438.29 657.35 541.88 59.48 2123.60

Table 2 (continued)

GTAPcode

Region's name Producer Consumer Income Pop. GDP

CHE Switzerland 26.69 72.40 59.52 7.24 357.54NOR Norway 52.45 46.51 143.27 4.60 250.05XEF Rest of EFTA 4.62 5.64 4.68 0.32 15.71ALB Albania 4.24 5.77 3.82 3.11 8.99BGR Bulgaria 41.83 31.29 32.12 7.78 24.57BLR Belarus 50.59 43.54 48.50 9.81 21.96HRV Croatia 15.20 20.30 11.58 4.54 33.93ROU Romania 76.53 69.01 58.11 21.79 74.42RUS Russian

Federation1332.95 1016.77 1464.48 143.90 569.84

UKR Ukraine 217.62 126.61 159.47 46.99 60.98XEE Rest of Eastern

Europe5.89 8.15 3.05 4.22 2.60

XER Rest of Europe 70.96 68.06 56.35 14.29 44.98KAZ Kazakhstan 161.61 134.85 157.05 14.84 44.35KGZ Kyrgyzstan 5.18 5.71 4.66 5.20 2.21XSU Rest of former

Soviet Union132.88 94.27 128.90 37.40 20.20

ARM Armenia 3.38 4.29 2.47 3.03 3.34AZE Azerbaijan 24.18 26.86 21.59 8.35 8.73GEO Georgia 2.43 4.67 2.36 4.52 4.47IRN Iran 299.80 301.86 340.16 68.80 157.86TUR Turkey 163.34 192.71 153.30 72.22 295.83XWS Rest of West Asia 909.16 707.86 1235.51 118.40 691.10EGY Egypt 120.29 101.71 113.73 72.64 76.81MAR Morocco 31.86 38.01 27.92 31.02 50.25TUN Tunisia 18.38 18.61 14.76 10.00 27.99XNF Rest of Norh Africa 127.64 110.65 191.62 38.10 112.39NGA Nigeria 39.92 38.16 101.83 128.71 68.57SEN Senegal 4.15 5.91 2.24 11.39 7.20XWF Rest of West

Africa19.85 34.51 20.36 117.42 50.73

XCF Rest of CentralAfrica

7.80 11.54 27.42 35.36 38.01

XAC Rest of South C.Africa

9.09 14.40 37.72 71.34 23.89

ETH Ethiopia 3.70 6.74 2.34 75.60 7.28MDG Madagascar 1.36 2.03 1.84 18.11 4.35MWI Malawi 0.55 1.57 0.63 12.61 1.79MUS Mauritius 1.83 3.80 2.43 1.23 5.92MOZ Mozambique 1.60 3.40 2.04 19.42 6.09TZA Tanzania 3.06 6.51 2.99 37.63 11.47UGA Uganda 2.26 3.51 2.90 27.82 7.27ZMB Zambia 1.77 3.09 2.14 11.48 5.40ZWE Zimbabwe 8.78 6.89 7.81 12.94 4.08XEC Rest of Eastern

Africa21.04 34.27 25.33 99.73 50.19

BWA Botswana 3.76 6.38 4.53 1.77 8.72ZAF South Africa 329.12 213.38 312.64 47.21 213.93XSC Rest of SACU 3.44 6.33 4.86 4.84 9.06

21731 21731 21731 6405 40962

Producer, consumer, and income-based responsibilities, population and GDP of GTAPregions (year 2004) (Mt CO2, million, thousand million USD2004).

61A. Marques et al. / Ecological Economics 84 (2012) 57–65

WIOD databases from European consortiums (Timmer, 2011; Tukkeret al., 2009) and the Asian IDE-JETRO (IDE, 2006).

6. Case Study

In this Section we present the results of the quantification ofincome-based responsibility, for 112 world regions (either countries orgroups of countries), using theGlobal TradeAnalysis Project (GTAP) data-base. We compare the results with production and consumption-basedresponsibilities.

6.1. Data and Methodology

Weused the GTAP 7.1 database (Narayanan andWalmsley, 2008) tobuild a full Multi-Regional Input–Output (MRIO) model (Peters et al.,2011a) covering the whole world, for the year 2004. An input–output(IO) model provides a description of an economy, for a given year

62 A. Marques et al. / Ecological Economics 84 (2012) 57–65

based on monetary flows between industrial sectors (Miller and Blair,2009). The distinctive feature of aMRIOmodel is that it allows the anal-ysis of the interactions and interdependencies between different sectorsin different countries.

If e is the vector of direct GHG emissions of every sector in theworldeconomy, the vectors of upstreamemissions embodied in final demand,u, and of downstreamemissions embodied in primary inputs, d, are cal-culated as:

u′ ¼ e′x−1� �

I� Zx−1� �−1

y; ð1Þ

d ¼ v I−1 � x−1Z� �−1

x−1e� �

; ð2Þ

where lowercase are vectors, uppercase are matrices, vectors are in col-umn format, ' is transpose and ˆ is diagonal matrix. The source data forthe calculations are: Z, the matrix of inter-sectoral transactions; y, thefinal demand vector; v, the vector of added value; and x, the vector oftotal output.

The derivation of Eqs. (1) and (2) can be found in Rodrigues et al.(2010), under an attributional interpretation, and in Oosterhaven

(1996), under a consequential interpretation. The term I−Zx−1� �−1

is the Leontief inverse (Leontief, 1936) and the term I−x−1Z� �−1

is

the Ghosh inverse (Ghosh, 1958).In anMRIO a country is defined as a set of sectors. So the production/

consumption/income-based responsibility of a country is, respectively,the sum of eis, uis and dis over the appropriate set of sectors i. Both up-stream and downstream emissions are conserved quantities, whichmeans that the consumption-based and the income-based responsibil-ity of theworld are both equal to the production-based responsibility ofthe world,∑iu′

i ¼ ∑id′i ¼ ∑ie′i . This implies that the choice of the re-

sponsibility principle can be interpreted as an allocation problem inwhich a pie (the world total emissions) can be cut into different slices(the responsibility of each region).

6.2. Results

The responsibility of each region according to each principle isreported in Table 2. In order to facilitate the subsequent discussionwe se-lected 15 representative regions for which we present figures. These re-gions are individual countries and two aggregate EU regions: EUR-15(Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland,Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, UnitedKingdom) and EUR-10 (Cyprus, Czech Republic, Estonia, Hungary,

Fig. 2. Difference between consumer and income-based responsibilities and produ

Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia). The latteris the group of countries that joined the European Union in 2004.

Fig. 2 presents the difference between consumption and production-based responsibility and the difference between income and production-based responsibility, as a percentage of production-based responsibility,for the selected regions. This figure illustrates the effect of the choice ofthe responsibility metric in the apparent contribution of a country toglobal emissions. The income-based responsibility of Norway(NOR) is 2.7 times higher than its production-based responsibilitywhile its consumption-based responsibility is slightly smallerthan its production-based responsibility. Venezuela (VEN), Australia(AUS), Russia (RUS) and Brazil (BRA) display the same pattern. Theconsumption-based and income-based responsibilities of Switzerland(CHE) are 2.7 and 2.2 times higher than its production-based responsibil-ity, respectively. For Singapore (SGP), EU-15 and Japan (JPN) the rela-tion between consumption and income-based responsibilities andproduction-based responsibility is the same as for Switzerland.South Africa (ZAF), China (CHN), EU-10 and India (IND) have lowervalues of consumption and income-based responsibility thanproduction-based responsibility.

Fig. 3 displays the per capita producer, consumer and income-basedresponsibility of the selected regions. Per capita production-basedvalues tell us how many tons of CO2 are generated inside the country'sborder per inhabitant. Per capita consumption-based values indicatethe emissions that are required to satisfy the final demand of goodsand services of a citizen of that country. Finally, income-based valuesrepresent the emissions that are required to generate the average in-come of that country.

We see that citizens of wealthier economies are, on average, re-sponsible for more CO2 emissions than citizens from least developedeconomies. We also see that in wealthier regions per capita consumerresponsibility is typically higher than producer responsibility. This ob-servation indicates that, in these regions, the (upstream) carbon em-bodied in imports exceed the (upstream) carbon embodied in exports.This phenomenon is particularly striking in small open economies,such as Singapore (SGP) or Switzerland (CHE), who rely strongly on in-ternational trade. On the other hand, for bigger economies, like UnitedStates or China (CHN), the indirectmeasures of responsibility are closerto the direct measure due to the predominance of the domestic effects.This occurs because consumer and income-based responsibilities arecarbon trade balances (Kanemoto et al., 2012; Rodrigues et al., 2010;Serrano and Dietzenbacher, 2010), and in larger economies domestictrade outweighs international trade.

Fig. 4 shows the different responsibility metrics per unit of GDP.We find that developing economies have higher carbon intensitiesthan more developed economies. The relations between producer,

ction-based responsibility (in % of production-based responsibility, in 2004).

Fig. 3. Per capita producer, consumer and income-based responsibilities (in ton CO2, in 2004).

63A. Marques et al. / Ecological Economics 84 (2012) 57–65

consumer and income responsibilities are maintained. In the case ofrich countries the differences are smoothed (for example, NOR). Inthe case of not so wealthy countries differences are sharpened (forexample, VEN). Brazil (BRA) appears amongst more developed econ-omies, an exception to this pattern. This may be explained by the factthat this country relies heavily on hydro power and ethanol. In 2004,the fraction of domestic electricity generated from hydro power was82.8%, only surpassed by Norway (NOR) with 98.8% (IEA, 2006).These results show that developed economies are more carbon effi-cient, generating per each dollar GDP less emissions than developingeconomies.

With these results we see that whether the analysis is done for abso-lute, per capita or per GDP values, there is a group of countries that con-sistently displays higher income-based responsibility. These are Norway(NOR), Venezuela (VEN), Australia (AUS) and Russia (RUS). A similaritybetween all these countries is the fact that they are all fossil fuel ex-porters. Thus, under the income-based perspective, we see that countrieswhich earn income from the export of fossil fuels (or other goods that en-able emissions to occur downstream along the supply chain) are held re-sponsible for the emissions generated abroadwhen the fossil fuel is burnt.

Interestingly, using this indicator we find that Norway, a countrywhose electricity is obtained mostly from hydropower and whichhas an otherwise very ‘clean’ economy (Peters and Hertwich, 2006;

Fig. 4. Per 2004 dollar GDP producer, consumer and in

Yamakawa and Peters, 2011), has the highest per-capita income-based GHG responsibility in the world.

7. Discussion and Conclusion

The aim of this work was to clarify and illustrate the informationprovided by the concept of income-based responsibility, i.e., the emis-sions enabled downstream by primary factors of production.

We suggest the use of income-based responsibility as a novel no-menclature to replace downstream responsibility, as a more intuitiveterm which also has a clear symmetry with the more popular metricof consumption-based responsibility. Thus, while consumption-basedresponsibility provides information on how GHG emissions are re-quired to generate final demand, income-based responsibility providesinformation on how GHG emissions are enabled to generate income.

To illustrate this metric we present empirical values of income-based responsibility and compare them to direct emissions (the metricused in theKyoto protocol and normally referred to as producer respon-sibility) and the alternative metric of consumer-based responsibility.

We find that, for some countries, different metrics yield very differ-ent responsibility values, while for other countries the responsibility isvery similar, irrespective of the principle chosen. Unsurprisingly, the

come-based responsibilities (in ton CO2, in 2004).

64 A. Marques et al. / Ecological Economics 84 (2012) 57–65

choice of metric is more important for smaller and more open econo-mies than for bigger economies.

The most emblematic case is that of Norway. This is a rich nation thatis considered one of the ‘cleanest’ countries in the world, with low pro-ducer and consumer responsibilities. However, its income is generatedat the expense of enabling large CO2 emissions abroad through the exportof fossil fuels. We have shown that income-based responsibility is ableto quantify this type of situation, while the alternative production orconsumption-based indicators cannot.

In climate policy, the analysis of a country's performance from thepoint of view of all responsibility principles provides a more integrat-ed understanding on how carbon emissions can be mitigated. Under aproduction-based perspective, mitigation options are circumscribedby domestic borders. Under a consumer-based perspective, importingcountries can improve their environmental performance through theselection of clean suppliers from abroad. Under an income-based per-spective, the symmetric possibility is open to exporting countries. Forexample, under the income-based perspective Norway can increaseits environmental performance by deciding to sell fossil fuels onlyto countries with carbon efficient production chains.

Another application for income-based responsibility could be itscombination with consumption-based responsibility in a single met-ric of shared responsibility (Gallego and Lenzen, 2005; Lenzen et al.,2007; Rodrigues et al., 2006).

Acknowledgements

We would like to thank the financial support of FCT through grantPTDC/AMB/64762/2006 and funding Pest-OE/EEI/LA0009/2011, FCT andthe MIT Portugal Program through scholarship SFRH/BD/42491/2007(to AM).

References

Ahmad, N., Wyckoff, A., 2003. Carbon dioxide emissions embodied in internationaltrade of goods. Technical Report DSTI/DOC(2003)15. Organization for EconomicCo-operation and Development, Paris, France.

Alexeeva-Talebi, V., Löschel, A., Mennel, T., 2008. Climate policy and the problem ofcompetitiveness: border tax adjustments or integrated emissions trading? CentreFor European Economic Research (ZEW) Discussion Paper, 08–061.

Andrew, R., Forgie, V., 2008. A three-perspective view of greenhouse gas emissionresponsibilities in New Zealand. Ecological Economics 68, 194–204.

Bastianoni, S., Pulselli, F., Tiezzi, E., 2004. The problem of assigning responsibility forgreenhouse emissions. Ecological Economics 49, 253–257.

BBC, 2009. China Seeks Exports Carbon Relief. http://-news.bbc.co.uk-/2/hi-/7947438.stm.Bruckner, M., Polzin, C., Giljum, S., 2009. Counting CO2 emissions in a globalisedworld. Tech-

nical Report. World Trade Organization-United Nations Environment Programme,Geneva, Switzerland.

Cadarso, M.A., López, L.A., Gómez, N., Tobarra, M.A., 2012. International trade andshared environmental responsibility by sector. An application to the Spanish econ-omy. Ecological Economics. http://dx.doi.org/10.1016/j.ecolecon.2012.05.009.

Davis, S., Caldeira, K., 2010. Consumption-based accounting of CO2 emissions. Proceed-ings of the National Academy of Sciences 12, 5687–5692.

de Cendra, J., 2006. Can emissions trading schemes be coupled with border tax adjust-ments? An analysis vis-à-vis WTO law. Review of European Community and Inter-national Environmental Law 2, 131–145.

Eder, P., Narodoslawsky, M., 1999. What environmental pressures are a region's indus-tries responsible for? Amethod of analysis with descriptive indices and input–outputmodel. Ecological Economics 3, 359–374.

EEA, 2010. The European environment — state and outlook 2010 synthesis. TechnicalReport. Copenhagen, Denmark.

EP, 2006. The Equator principles. Technical Report. U.K.Ferng, J.J., 2003. Allocating the responsibility of CO2 over-emissions from the perspec-

tives of the benefit principle and ecological deficit. Ecological Economics 46,121–141.

Gallego, B., Lenzen, M., 2005. A consistent input–output formulation of shared produc-er and consumer responsibility. Environmental Systems Research 17, 365–391.

Ghosh, A., 1958. Input–output approach in an allocation system. Economica 25, 58–64.Hummels, D., Ishii, J., Yi, K., 2001. The nature and growth of vertical specialization in

world trade. Journal of International Economics 1, 75–96.IDE, 2006. Asian international input–output tables 2000 volume 1. Explanatory notes.

Technical Report. Institute of Developing Economies, Japan External Trade Organi-zation, Japan.

IDE-JETRO, WTO, 2011. Trade patterns and global value chains in East Asia: from tradein goods to trade in tasks. Technical Report. IDE-JETRO, World Trade Organization,Geneva, Switzerland.

IEA, 2006. Key world energy statistics 2006. Technical Report. International EnergyAgency, Paris, France.

IPCC, 1997. Revised 1996 IPCC guidelines for national greenhouse gas inventories.Technical Report. Intergovernmental Panel on Climate Change, London, UK.

Ismer, R., Neuhoff, K., 2004. Border tax adjustments: a feasible way to addressnonparticipation in emissions trading. Cambridge Working Papers in EconomicsCWPE, 0409. Department of Applied Economics, University of Cambridge, Cam-bridge, UK.

Kanemoto, K., Lenzen, M., Peters, G., Moran, D., Geschke, A., 2012. Frameworks for com-paring emissions associated with production, consumption, and internationaltrade. Environmental Science and Technology 46, 172–179.

Laroche, M., Bergeron, J., Barbaro-Forleo, G., 2001. Targeting consumers who are willing topay more for environmentally friendly products. Journal of Consumer Marketing 18,503–520.

Lenzen, M., 2007. Aggregation (in-)variance of shared responsibility: a case study ofAustralia. Ecological Economics 64, 19–24.

Lenzen, M., Murray, J., 2010. Conceptualising environmental responsibility. EcologicalEconomics 70, 261–270.

Lenzen, M., Murray, J., Sack, F., Wiedmann, T., 2007. Shared producer and consumerresponsibility — theory and practice. Ecological Economics 61, 27–42.

Lenzen, M., Kanemoto, K., Moran, D., Geschke, A., 2012. Mapping the structure of theworld economy. Environmental Science and Technology 45, 8374–8381.

Leontief, W., 1936. Quantitative input and output relations in the economic systems ofthe United States. The Review of Economics and Statistics 18, 105–125.

Marques, A., Rodrigues, J., Domingos, T., 2011. Embodied emissions and the carbontrade balance between world regions. 19th International Input–output Conference.

Miller, R., Blair, P., 2009. Input–Output Analysis — Foundations and Extensions. CambridgeUniversity Press, Cambridge, MA, USA.

Munksgaard, J., Pedersen, K., 2001. CO2 accounts for open economies: producer or con-sumer responsibility. Energy Policy 3, 327–334.

Narayanan, B., Walmsley, T., 2008. Global trade, assistance and production: the GTAP 7Database. Technical Report. Center for global trade analysis - Purdue University,West Lafayette, IN, USA.

OECD, 2003. OECD environmental indicators: development, measurement and use.Technical Report. Organization for Economic Development and Cooperation, Envi-ronment Directorate, Paris, France.

OECD, 2008. OECD key environmental indicators 2008. Technical Report. Organization forEconomic Development and Cooperation, Environment Directorate, Paris, France.

Oosterhaven, J., 1996. Leontief versus Ghoshian price and quantity models. SouthernEconomic Journal 3, 750–759.

Pedersen, O., de Haan, M., 2006. The system of environmental and economic accounts2003 and the economic relevance of physical flow accounting. Journal of IndustrialEcology 1–2, 19–42.

Peters, G., 2008. From production-based to consumption-based national emissions invento-ries. Ecological Economics 65, 13–23.

Peters, G., Hertwich, E., 2006. Structural analysis of international trade: environmentalimpacts of Norway. Economic Systems Research 18, 155–181.

Peters, G., Hertwich, E., 2008a. CO2 embodied in international trade with implicationsfor global climate policy. Environmental Science and Technology 42, 1401–1407.

Peters, G., Hertwich, E., 2008b. Post-Kyoto greenhouse gas inventories: production versusconsumption. Climate Change 86, 51–66.

Peters, G., Andrew, R., Lennox, J., 2011a. Constructing an environmentally extendedmulti-regional input–output table using the GTAP database. Economic SystemsResearch 2, 131–152.

Peters, G., Minx, J., Weber, C., Edenhofer, O., 2011b. Growth in emission transfers via interna-tional trade from 1990 to 2008. Proceedings of the National Academy of Sciences 21,8903–8908.

Petherick, A., 2012. Market watch: dirty money. Nature Climate Change 2, 72–73.Rodrigues, J., Domingos, T., 2008. Consumer and producer environmental responsibility:

comparing two approaches. Ecological Economics 66, 533–546.Rodrigues, J., Domingos, T., Giljum, S., Schneider, F., 2006. Designing an indicator of en-

vironmental responsibility. Ecological Economics 59, 256–266.Rodrigues, J., Marques, A., Domingos, T., 2010. Carbon Responsibility and Embodied

Emissions: Theory and Measurement. Routledge, London, UK.Schücking, H., Kroll, L., Louvel, Y., Richter, R., 2011. Bankrolling climate change— a look into

the portfolios of the world's largest banks. Technical Report. urgewald, GroundWork,Earthlife Africa Johannesburg, BankTrack, Nijmegen, The Netherlands.

Serrano, M., Dietzenbacher, E., 2010. Responsibility and trade emissions balances: anevaluation of approaches. Ecological Economics 69, 2224–2232.

Stern, N., 2007. The Economics of Climate Change— The Stern Review. CambridgeUniversityPress, Cambridge, UK.

Timmer, M., 2011. The World Input–Output Database (WIOD): contents, sources andmethods. Technical Report 0.9. The WIOD Project, Nijmegen, The Netherlands.

Tukker, A., Poliakov, E., Heijungs, R., Hawkins, T., Neuwahl, F., Rueda-Cantuche, J.,Giljum, S., Moll, S., Oosterhaven, J., Bouwmeester, M., 2009. Towards a globalmulti-regional environmentally extended input–output database. EcologicalEconomics 69, 1928–1937.

UNEP, 2007. Global environmental outlook — GEO4 environment for development.Technical Report. Valletta, Malta.

UNFCCC, 1998. Kyoto Protocol to the United Nations Framework Convention on ClimateChange. Technical Report. Bonn, Germany.

UNFCCC, 2012. Report of the conference of the parties on its seventh session, heldin Durban from 28 November to 11 December 2011. Technical Report FCCC/KP/

65A. Marques et al. / Ecological Economics 84 (2012) 57–65

CMP/2011/10/Add.1. United Nations Framework Convention on Climate Change,Geneva, Switzerland.

WBCSD, WRI, 2010. The greenhouse gas protocol — corporate value chain (Scope 3)accounting and reporting standard draft for stakeholder review. Technical Report.World Business Council for Sustainable Development and World Resources Institute,Geneva, Switzerland.

Whalley, J., 2009. On the effectiveness of carbon-motivated border tax adjustments.Technical Report: Asia-Pacific Research and Training Network on Trade - WorkingPaper Series, 63. Bangkok, Thailand.

Whalley, J., Walsh, S., 2009. Bringing the Copenhagen global climate change negotiationsto conclusion. CESifo Economic Studies 55, 255–285.

Yamakawa, A., Peters, G., 2011. Structural decomposition analysis of greenhouse gasemissions in Norway 1990–2002. Economic Systems Research 23, 303–318.

Zhou, X., Liu, X., Kojima, S., 2010. Carbon emissions embodied in international trade: anassessment from the Asian perspective. Technical Report. IGES, Kanagawa, Japan.