Energy and Climate ChangeMemorandum submitted by Tyndall Manchester, Sustainable Consumption Institute and Stockholm Environment Institute

Please find enclosed a submission by a selection of researchers from the Sustainable Consumption Institute and the Tyndall Centre, University of Manchester, and from the Stockholm Environment Institute, University of York. All views contained within are attributable to the authors and do not necessarily reflect those of researchers from the wider Sustainable Consumption Institute, Tyndall Centre, Stockholm Environment Institute or University of Manchester.

Dr Elena Dawkins, Stockholm Environment Institute, University of York.

Dr Ruth Wood, Sustainable Consumption Institute & Tyndall Centre, School of Mechanical, Civil and Aerospace Engineering, University of Manchester.

Dr Alice Bows, Sustainable Consumption Institute, School of Mechanical, Civil and Aerospace Engineering, University of Manchester.

Dr Mirjam Roeder, Sustainable Consumption Institute, School of Mechanical, Civil and Aerospace Engineering, University of Manchester.

Dr Paul Gilbert, Tyndall Centre, School of Mechanical, Civil and Aerospace Engineering, University of Manchester.

Dr Conor Walsh, Sustainable Consumption Institute and Tyndall Centre, School of Mechanical, Civil and Aerospace Engineering, University of Manchester.

Executive Summary

1. How do assessments of the UK’s greenhouse gas emissions differ when measured on a consumption rather than a production basis?

Analyses show UK GHG emissions measured on a consumption basis are consistently higher than those on a production basis. Furthermore, this gap is widening as consumption emissions are increasing, despite territorial emission reductions.

2. Is it possible to develop a robust methodology for measuring emissions on a consumption rather than production basis and what are the challenges that need to be overcome to deliver this?

Whilst it is possible to develop a robust methodology for measuring emissions on a consumption basis, consumption-based accounts have a higher uncertainty than territorial accounts.

Current consumption-based emissions estimates use various different national and international economic and environmental datasets from a range of sources and these require data processing to combine into an appropriate framework. The necessary harmonization of differing datasets can generate uncertainties, developing a standardised harmonization and data validation and verification process is the main challenge.

The challenges presented by consumption-based accounting fall into two areas; the underlying challenges of the modelling approach in terms of complexity, time requirements or uncertainties; and the challenges posed by the data limitations and restrictions.

There are two main methodologies that are suitable for estimating consumption emissions at a national level; a multi regional input-output model (MRIO) and an “emissions embodied in trade” method. The results from these methods will vary depending on the model used, the underlying datasets and how the model is structured.

Peters (2008) highlights that calculating emissions embodied in trade is less complicated than developing global EEIO models (the multi-regional input-output approach), but only deals with bilateral trade flows at the country level and therefore not as suitable as MRIO for looking at the full supply chain impacts of the final consumer.

According to Wiedmann et al (2010), there are key challenges associated with MRIO including:

collection of and access to environmental data at a high level of sectoral and source detail;

sectoral detail of national accounts;

country coverage; and

the support of national and international statistical authorities and agencies who provide the input data.

Wiedmann et al (2010) note that whilst most of the major MRIO projects are supported through public funding and consequently made publically available, there is further scope for producing more open source data and increasing transparency of data transformations and assumptions made within the models and the reproducibility of the methods.

3. What are the benefits and disadvantages associated with taking a consumption-based rather than production-based approach to greenhouse gas emissions accounting?

National perspective

Consumption-based emissions accounting provides a complementary perspective to production-based or territorial emissions accounts. It is particularly relevant linking trade and climate policy due to its inclusion of imports; a valuable indicator for countries where emissions mitigation targets are met domestically, but an increasing level of consumption continues by importing emissions intensive goods from elsewhere. This becomes even more relevant for climate change policy if those goods are imported from countries without any binding targets for emissions reductions themselves.

If a multi-regional input output (MRIO) method is used for consumption-based emissions accounting, it can be applied to any level of consumption because all supply chain emissions are allocated to the final consumer. This has the benefit of engaging individuals or communities in their personal contribution to emissions, and could be useful for mitigation efforts to change behaviours or practices domestically.1 It can also demonstrate the different responsibilities for components of a carbon footprint; with the possibility to explore both the contribution of a producer or industry and the consumer (individually) or level of consumption (nationally). At the national scale, for higher level consumption emission inventories, methods such as emissions embodied in trade or time series with trade (Peters et al, 2011) would demonstrate the emissions associated with trade, which production-based accounts currently do not show.

Combining both production and consumption-based accounts is useful for climate change mitigation policy and decisions. Experience of organisations such as Stockholm Environment Institute (SEI) and the Sustainable Consumption Institute suggest that both consumption and production accounts should be maintained and published regularly as opposed to one or the other; especially as estimates for production-based emissions are required as an input to consumption-based accounts anyway.

Sector perspective

With particular reference to the agricultural sector, using both a consumption and production-based approach helps to classify the source and end-user of emissions. Production in some regions where demand is increasing is projected to be limited by climate change impacts and a greater number of countries may rely on a few main producers. With this outlook producers in the “favoured” regions need to maximise their yields and land-use efficiency to maintain global food supply. However, this will necessitate a higher use of agrochemicals, particularly nitrogenous fertilisers, increasing greenhouse gas emissions within those territories. The result of this may be a disincentive to increase production in these countries if the production-based accounting remains without being informed by the consumption-based approach. This would be counterproductive to global food security aims and poverty reduction.

With particular reference to the international transport sectors, under the producer-based approach, international shipping (and aviation) emissions are excluded, as the majority of emissions are released in international waters (and airspace). Nonetheless, the amount of bunker fuel sold by a nation is reported to the UNFCCC by Annex 1 nations as a memo item. This current method of reporting was described by the Environmental Audit Committee as being inadequate to represent the UK’s actual share of international shipping emissions (EAC, 2009). By taking a consumption-based approach, the emissions released from international shipping (and aviation) would be automatically included in the emission accounts. The international shipping (and aviation) emissions would then constitute part of the embedded emissions that are released along the goods or services supply chain. It should be noted however, that shipping is not well represented in existing MRIO-related databases, and this would need to be addressed.

4. Would it be (a) desirable and (b) practicable for the UK to adopt emissions reduction targets on a consumption rather than production basis?

(a) Desirable

It would bring together climate policy and trade, which would tackle the common critique of production-based accounting for not including the emissions associated with imported goods. Under production-based targets the consumption of imports and their associated embedded emissions can continue to increase without any monitoring, but a consumption-based target would include these emissions.

The consumption approach can demonstrate the role that consumption plays in counteracting emissions mitigation efforts, if growing consumption outweighs the efficiency improvements domestically or from imports then this would be evident from the consumption-based accounts.

The consumption-based approach would mean the UK taking more responsibility globally for the emissions that are generated to support growing levels of domestic consumption. This would increase the share of global emissions over which the UK has influence, and therefore broaden its reach over the climate outcome of an increase in emissions (Bows and Barrett, 2010).

(b) Practicable

Nationally, work would need to be done to collect and publish the necessary datasets more frequently in order to produce timely and robust consumption-based emissions estimates for regular accounting and monitoring. It may not be possible to use a full MRIO approach for measuring against targets as it is reliant on global trade data and models, which are not produced annually due to their data intensive production process. Alternative methods such as emissions embodied in trade (see Peters, 2008) or time series with trade (see Peters et al, 2011) could be used to generate annual consumption-based accounts, but this is also likely to require additional work to publish the necessary input datasets more regularly. As Peters, 2008 notes, MRIO is most useful for product or consumption specific studies and a method such as emissions embodied in trade would be more useful for national emissions inventories limited to bilateral trade flows. Solely consumption-based national emissions inventories would suffer from higher complexity and uncertainty and lower transparency than production-based accounts.

5. Is there any evidence of industry relocating from the UK to other countries as a result of UK climate change policy?

This is outside our sphere of research.

6. What are the potential implications at the international level of the UK adopting a consumption- rather than production-based approach to greenhouse gas emissions accounting?

If the UK were to adopt consumption-based accounting alongside production-based accounting then this would demonstrate internationally that the UK is taking into account the emissions impacts arising from trade. If the model used to calculate the consumption-based emissions can provide the necessary detail then it could also be used to identify potential carbon leakage from countries with binding targets to those without, demonstrating the UK’s understanding of the importance of making progress towards emissions reductions on a global scale.

If the UK adopted a solely consumption-based approach to emissions accounting then the UK would have to work closely with other countries to encourage mitigation efforts abroad to reduce the impact of any imported goods to the UK.

Detailed Consultation Response

1. How do assessments of the UK’s greenhouse gas emissions differ when measured on a consumption rather than a production basis?

Background

1.1 Production-based greenhouse gas (GHG) emissions accounts include the emissions released within a national territory. Exact inclusions or exclusions vary depending on the reporting framework, but in general they include GHGs released from burning fuel and chemical processes in industry or households (and sometimes removals) taking place within a national boundary. Consumption-based GHG emissions accounts include the emissions embedded along the supply chain of goods and services that people in a particular country consume, regardless of where those goods and services are produced and the emissions released. A significant difference between the two approaches is how they treat trade, with production-based accounts including emissions associated with goods and services for export, but excluding imports, and consumption-based accounts including imports, but allocating export emissions to their country of destination.

1.2 There have been a number of consumption emissions studies completed for the UK. Accounts, such as those produced by REAP (Resources and Energy Analysis Programme) at Stockholm Environment Institute (SEI), University of York, or those published in academic literature by J. C. Minx et al (2009), Druckman and Jackson (2009), Wiedmann et al (2010) and Peters et al (2011) have found that consumption-based emissions are consistently higher than production-based estimates. All of these models apply an environmentally extended input-output approach in some form to estimate the consumption-based emissions for the UK, but vary in the datasets and methods that they use (the differences between accounting methods are described in the following sections).

1.3 Examples of specific reports on consumption-based emissions accounts or calculations of emissions embodied in trade for other countries include: Munksgaard and Pedersen (2001)—Denmark; Machado, Schaeffer, and Worrell (2001)—Brazil; Peters and Hertwich (2006)—Norway; Bang et al (2008)—Europe; Yunfeng and Laike (2010)—China; Z M Chen and Chen (2011)—G7 and BRIC; Berglund (2011)—Sweden; Whilst this list is by no means exhaustive it gives an indication of the work internationally on consumption-based accounting. This list also excludes the larger projects such as EXIOPOL or OPEN:EU at the European level which have generated consumption-based accounts for a large number of countries simultaneously using environmentally extended multi-regional input-output models. These projects are also explained in more detail in the following sections.

Emissions Comparison

1.4 The REAP model generates consumption-based footprint estimates for the UK and disaggregates this data to local authority level. The data from REAP for 2006 are shown in Table 1 and compared to nationally published production-based GHG statistics for the same year. For each of the data sources the measure of GHGs includes the following pollutants: CO₂, CH₄ and N₂O and HFCs, PFCs and SF6. The production-based emissions accounts vary depending on the reporting requirements—with UNFCCC2 data taken from Environmental Accounts, but excluding bunker fuels, CO₂ emissions from biomass, adjustments for tourism abroad and including crown dependencies and land use change.

Table 1

A COMPARISON OF EMISSIONS DATASETS

REAP Data 2006— Consumption Emissions		DECC archive of annual statistics3 2006—UNFCCC reported Production Emissions	Office of National Statistics Environmental Accounts 2008—Production Emissions
Total Carbon Footprint (Mt CO2)	Total GHG Footprint (Mt CO2 eq)	Kyoto greenhouse gas basket, (Mt CO2 eq)	Greenhouse Gas, atmospheric emissions, 2006 (Mt CO2 eq)
733.08	983.94	652.3	724.46

Figure 1below shows the REAP breakdown of the GHG footprint by theme for 2006.

Figure 1

DIRECT AND INDIRECT EMISSIONS BREAKDOWN OF THE UK PER CAPITA FOOTPRINT, BY THEME, 2006

1.5 The REAP dataset also includes a time series of consumption-based emissions from 1992–2006, this is compared to UNFCCC reported figures for territorial emissions4 from the same time period in Figure 2. This demonstrates the increasing difference between consumption and territorial emissions accounts over time.

Figure 2

A TIME SERIES OF GHG CONSUMPTION EMISSIONS AND UNFCCC REPORTED EMISSIONS 1992-2006

1.6 Wiedmann et al (2010) has produced a time series consumption emissions dataset for CO₂ alone (Figure 2 shows both CO₂ and non-CO₂ greenhouse gases) from 1992–2004 and this is shown in Figure 3 below, with a comparison to production emissions from the UNFCCC submission and UK Environmental Accounts from the Office of National Statistics. This study also found that there has been an increase in UK consumer emissions over time and a widening gap between producer and consumer emissions. Net CO₂ emissions embedded in UK imports increased from 4.3% of producer emissions in 1992 to a maximum of 20% in 2002.

Figure 3

THE CO₂ EMISSIONS FROM 1992–2004 ACCORDING TO DIFFERENT ACCOUNTING PRINCIPLES (WIEDMANN ET AL 2010)

1.7 Peters et al (2011) have also produced a time series of consumption-based emissions which, due to a difference in methodological approach (explained in more detail below) gives another set of consumption emission estimates for the UK. However, despite the differences in methodology and total values, the conclusions are similar to the studies mentioned above—with a widening gap between territorial and consumption emissions between 1990 and 2008 in the UK.

2. Is it possible to develop a robust methodology for measuring emissions on a consumption rather than production basis and what are the challenges that need to be overcome to deliver this?

2.1 Whilst it is possible to develop a robust methodology for measuring emissions on a consumption basis, consumption-based accounts have a higher uncertainty than territorial accounts. As Peters (2008) states, “production-based inventories are much closer to the statistical source than consumption-based inventories and therefore have lower uncertainty”. Consumption-based accounts generate an increased level of uncertainty from the reallocation of emissions from technologies to sectors and the inclusion of import data. In terms of challenges, these factors are inherent in the accounting approach, rather than a challenge to overcome.

2.2 Accepting the increased uncertainty of the consumption-based approach as a whole, the data availability and modelling approach used to calculate the emissions can also vary between studies, giving differing national consumption-based inventory results depending on the method employed.

Background

2.3 The diagram in Figure 4, taken from Wiedmann et al (2010), shows the data requirements for consumption and production-based indicators. The exact numerical difference between territorial and emissions accounts depend on the methodology employed to estimate the emissions associated with each number in the diagram. To calculate the full supply chain emissions associated with consumption, carbon emitted during the production of goods or services must be reallocated from the point of release to the final product that is consumed. To do this the emissions associated with different goods and services must be tracked along supply chains, which are often complex involving numerous industries and processes. One method for tracking emissions along supply chains throughout an economy is environmentally-extended input-output (EEIO) analysis,5 which is the method behind the data presented in response to question 1 above. EEIO is a form of life cycle analysis (LCA), but using a top-down macro-economic approach, as opposed to bottom-up or process LCA which looks at products individually.6

Figure 4

EMISSIONS ACCOUNTING DIAGRAM BASED ON WIEDMANN ET AL (2010)

2.4 EEIO analysis uses a number of different national and international economic and environmental datasets from a range of sources and these require a certain level of data processing to combine into an EEIO framework. The necessary harmonization of differing datasets can generate uncertainties.

Modelling framework

Input-output tables

2.5 The economic input-output tables (IOTs) that represent the interactions between industries’ (supply chains) are used as part of the EEIO framework and they can vary in level of industry detail and countries that they include. If symmetric input-output tables (SIOT) are not available for the year of interest then estimates have to be made based on economic supply and use tables (SUT).

2.6 To generate a full consumption account the framework must include emissions associated with traded goods (imports and exports). This data can be incorporated in a number of ways, with different methods producing different emissions results. Single region models would just represent the interactions between sectors in one country, with the assumption that any imported goods have the same emissions intensity as those produced domestically (domestic technology assumption).

2.7 Multi-region input-output (MRIO) models represent interactions between numbers of countries, including both international and domestic supply chains. EEIO models with different regions (total numbers and the actual countries included) will give different consumption emissions figures for the country of interest. In addition, the emissions datasets available for those regions and the data that describes how they interact with consumers and industry in the country of interest may also differ between models.

Input-output data sources

2.8 Across the world, modelling frameworks for consumption-based emissions accounting tend to use either nationally published input-output (IO) or “supply and use” tables combined with environmental accounts and import data in single region models with domestic technology assumptions, or MRIO models with larger numbers of regions, based on international datasets such as the Global Trade Analysis Programme7 (GTAP) or EXIOPOL.8 The consumption-based emissions estimates will be different depending on the IO tables and model used. In some cases, global models are used as a starting point and then aggregated to regions of interest; this again produces a variation in results.

2.9 Generating a global set of IO tables for use in MRIO requires considerable data processing and has become available through specific projects or work by individuals such as Hertwich and Peters (2010) for the One Planet Economy Network in the EU (OPEN:EU). The method for generating the environmentally-extended input-output model for 113 regions is described in the OPEN:EU technical document and is based on GTAP data which was converted to a full MRIO model.9 The consumption account data for 45 countries (including the UK) will shortly be published in the EUREAPA tool which is an output of this project.10

2.10 Tukker et al (2009) document the steps they took for producing the EXIOPOL environmentally-extended MRIO model, firstly identifying the limitations with existing data sources and then presenting their method to overcome these and deliver the model. They noted that the current data sources at the time provided only supply and use tables and input-output tables for single countries, without trade links. In addition, the sectoral and product detail was limited and environmental extensions were either lacking or limited in scope. They stated that a key difficulty was the lack of harmonization of data across countries and this is one of the reasons that any MRIO modelling work will inevitably include data harmonization and manipulation. They conclude by stating that trade-linked tables are essential for analysing the effects of sustainability measures taken in Europe on Europe's economic competitiveness and, they believe that, from a theoretical viewpoint, the environmentally-extended MRIO approach is best way of taking trade into account. However, they also state that existing studies tend to be aggregated at sector and regional level and to focus on a fairly small number of environmental extensions.

2.11 Wiedmann et al (2010) identify five projects that have or are about to complete MRIO databases, with full trade matrices between all countries:

The Asian International Input-Output Table by IDE/JETRO: symmetric input-output tables with nine Asian countries, plus the USA with 76 sectors.

The Eora database by the University of Sydney: a global MRIO time series with 130 countries.

The EU funded EXIOPOL database with supply and use tables for 27 EU countries and 16 non-EU countries and 130 sectors (as described above).

GTAP data; covering 113 countries/world regions and 57 sectors. It does not include full trade matrices between all countries in the database itself, but can be converted to do so as completed for the OPEN:EU project described above.

The World Input-Output Database Project (WIOD) from the University of Groningen which aims to create a time series of supply and use tables and symmetric IO tables from 1995–2006 for 27 EU countries and 13 other countries, with 35 industries and 59 product sectors.

Key uncertainties in EEIO modelling—challenges to overcome

2.12 Andrew, Peters, and Lennox (2009) use an MRIO model based on GTAP data to quantify the errors introduced by various approximations of the full MRIO model and found that the frequently used domestic technology assumption (DTA) inaccurately estimates the carbon footprint for many countries. The errors are due to globally unrepresentative emission intensities and/or production technologies. They recommend that when applying the DTA, domestic emission intensities should be carefully validated against global estimates, especially for commodities with large import volumes.

2.13 There are advantages of using a full MRIO model, but they also require a significant amount of data from a variety of sources, which raises the level of uncertainty of the consumption-based accounts which are calculated using this method. Several practical issues arise in the data manipulation phase for MRIO modelling and a number of assumptions must be made. These are discussed in more detail in Hertwich and Peters (2010), but in summary they are as follows:

The trade interactions: Trade can be either uni-directional, where the domestic country trades with all countries or multi-directional where the domestic country trades with all countries and they trade between each other. Alternatively, imports could be assumed to be produced with the domestic production technology. This greatly reduces the data requirements, but may lead to large errors.

Regional groupings: it may be necessary to aggregate some regions for data handling purposes or to fill in missing IO data for some areas.

Trade flows to final demand and trade flows to industry data may not be available for all countries and therefore may have to be estimated.

Exchange rate variation: models could use either Purchasing Power Parity or Market Exchange Rates, but some method must be used to run the model in a common currency.

Inflation: the data for different countries is often available at different time periods, which means that adjustments must be made to the data to generate a base year. Methods such as the Consumer Prices Index are usually available in each country, but this can generate errors when applied on aggregate to sectors with different price changes and it can also vary depending on the base year and the method of indexing applied.

Product or industry classifications: the matrices which hold the industry and final demand data for countries can be classified by products or industries, with industry emissions data usually classified by industry and final demand by product. Additional matrices are therefore usually required to map between the two classifications.

Classification of IO data: this can vary between countries and the analysis requires mapping to a consistent classification system.

Data aggregation: it is beneficial to keep sectoral data at the highest detail available to reduce errors, but to generate consistency across countries some aggregation of data is usually required.

Valuation: IO data can come at different levels of valuation –basic, producer or purchasers (retail) prices. For consistency is it most appropriate to use the IO data in basic prices where possible, this often requires some translation of datasets from one form to another.

Generating a time series

2.14 The possibility of generating a consistent time series is limited by the data availability. For example, national symmetric input-output tables or analytical tables may not be produced annually. GTAP data is usually published every three or four years, with data available for 1997, 2001 and 2004.The environmental data available, particularly in the larger global models may also be inconsistent across countries and years. Additionally, many of the data manipulation issues mentioned above are relevant when attempting to generate a consistent time series of data.

Methods for incorporating trade data – comparing models

2.15 Despite the level of data manipulation required and the complexity of the approach, a number of studies have employed global MRIO models or national accounting frameworks for EEIO to calculate emissions associated with domestic consumption for different countries and the first section of this document demonstrates that they can draw similar conclusions. In order to gauge the extent to which the limitations, assumptions and uncertainties influence the overall consumption-based accounts it is useful to compare the results from different models and studies. SEI recently completed a comparison of consumption emissions for Sweden from three different EEIO studies, and found similar results across models.11

2.16 On a larger scale, Peters et al (2011) have explored the possible modelling variations and uncertainties by comparing the results from two different calculation methods using a global 113 world region EEIO model, producing data for 95 countries. They compare a method called emissions embodied in trade (EEBT) (Peters 2008) with multi-regional input-output analysis (MRIO) for three years (1997, 2001 and 2004) and use trade data to construct a time series of consumption emissions for 95 countries between 1990 and 2008 named time series with trade (TSTRD). All three methods (TSTRD, EEBT and MRIO) calculate the emissions embedded in supply chains to produce consumed goods and services, but EEBT and MRIO are more accurate and cover either domestic supply chains (EEBT) or global supply chains (MRIO). They note that the uncertainty of their modelling does increase as the results become more disaggregated, but the TSTRD method is still stable for countries and sectors.

2.17 Peters et al (2011) conclude that EEBT is more suitable for analysing bilateral trade and MRIO is more appropriate for studies at the sub-national level or comparisons of final consumption between countries. Neither method is right or wrong, but they allocate emissions differently and can answer different questions. The EEBT method correlates directly with bilateral trade statistics (in proportion to domestic emission intensity) and the MRIO method correlates directly with final consumption (in proportion to global emission intensity).

2.18 In terms of application, the inclusion of the full supply chain in the MRIO method makes it more appropriate for disaggregation to localities, because the supply chain emissions are allocated to the final consumer and can therefore be applied to any level of consumption. Other methods like the EEBT may not be as suitable for disaggregation as MRIO, but their reduced complexity and more readily available data make them more appropriate for national time series analysis (Peters et al 2011).

2.19 Whilst there can be variations in the models and data Wiedmann (2009) provides an overview of use of environmentally-extended input-output models across the world, highlighting that MRIO is a sound and relevant methodology for accounting for trade-related impacts from a consumption perspective:

“The SKEP-ERA network of funding institutions in Europe initiated two projects in 2008 aimed at identifying and describing a suitable methodology to assess trans-national environmental impacts. The aim was to bring together existing knowledge and ongoing research on the assessment of global environmental impacts of traded goods and services, to review past and current accounting methodologies and to identify, specify and describe a suitable integrated approach. The project completed at the time of writing, EIPOT, came to the conclusion that multi-region input–output (MRIO) analysis is a sound and relevant methodology for accounting for trade-related impacts from a consumption perspective. Whilst MRIO cannot cover all policy and research questions in this area on its own, it forms a robust basis upon which more specific methods, using various forms of hybrid modelling, can be built (Wiedmann et al 2009).”

In this paper Wiedmann (2009) also assesses the literature and studies on the uncertainty of MRIO analysis, highlighting the general uncertainties associated with input-output modelling. He notes that both single region input-output analysis and multi-region input-output analysis will generate uncertainty from their source (survey) data, imputation and balancing, allocation, assuming proportionality and homogeneity, aggregation, temporal discrepancies, model inputs, and multipliers. Still, in his further paper (Wiedmann et al 2010) he emphasises that MRIO is emerging as a particularly comprehensive, versatile and compatible approach.

Overcoming Challenges

2.20 Some of the major methodological, data and computational challenges for producing EEIO models are mentioned above, but there are also more practical challenges such as costs, time commitments and timeliness of the publication of statistical data. The majority of the methodological and data challenges can or have been tackled across various research groups and published across the academic literature, with some studies exploring the levels of uncertainty in models and comparisons of the results from differing methods (see M Lenzen, Wood, and Wiedmann, 2010 and Peters et al 2011).

2.21 Wiedmann et al (2010) put forward a number of points for the requirements for MRIO development and research; better collection and access to environmental data; improved sectoral detail; limiting the data manipulations required to maintain the original data within models as far as possible; improving country coverage and the support of national and international statistical authorities and agencies who provide the input data. They note that whilst most of the major MRIO projects are supported through public funding and consequently made publically available there is further scope for producing more open source data and increasing transparency of data transformations and assumptions made within the models and the reproducibility of the methods. They also make suggestions for how to improve the infrastructure and implementation of MRIO research and how to deliver and use the results within policy making.

3. What are the benefits and disadvantages associated with taking a consumption-based rather than production-based approach to greenhouse gas emissions accounting?

National perspective

3.1 Consumption-based emissions accounting provides an alternative and complementary perspective to production-based or territorial emissions accounts. It provides a different view of responsibility for emissions and can be useful to inform the policy debate on climate change mitigation. It is particularly relevant linking trade and climate policy due to its inclusion of imports; a valuable indicator for countries where emissions mitigation targets are met domestically, but consumption levels are maintained by importing emissions intensive goods from elsewhere. This becomes even more relevant for climate change policy if those goods are imported from countries without any binding targets for emissions reductions.

3.2 The MRIO method for consumption emissions accounting can be applied to any level of consumption because all supply chain emissions are allocated to the final consumer. This has the benefit of engaging individuals or communities in their personal contribution to emissions, and could be useful for mitigation efforts to change behaviours or practices domestically.12 It can also demonstrate the different responsibilities for components of a carbon footprint; with the possibility to explore both the contribution of a producer or industry and the consumer (individually) or level of consumption (nationally).

3.3 There is a broad consensus in the academic literature that both production and consumption-based accounts are useful and relevant to climate change mitigation policy and decisions. Peters (2008) identifies the following benefits to generating a national consumption-based emissions inventory:

eliminating carbon leakage through imports;

covering more global emissions with limited participation;

consistency between consumption and environmental impacts;

increasing mitigation options; and

making policies such as Clean Development Mechanism (CDM) a natural part of the inventory.

3.4 There are also potential disadvantages of taking a solely consumption-based approach to emissions accounting. Firstly, consumption accounts require more complex calculations and assumptions than production accounts, which increase the level of uncertainty. Secondly, it might result a loss of ownership and responsibility for production-based emissions domestically. This wouldn’t be so much of a problem in countries where many of the goods produced domestically are consumed domestically, as this would appear in their consumption accounts. However, in those countries where export levels are high and consumption-based emissions remain comparatively small, the incentive to reduce emissions domestically might be lower than when they focus on reducing their production-based emissions.

3.5 In a paper on production and consumption-based emissions inventories Peters (2008) concludes that given the lower uncertainty, established reporting, consistency with political and environmental boundaries, and already wide-spread use, it is likely that production-based inventories will remain dominant. He then goes on to mention that consumption-based inventories provide considerable insight into climate policy and mitigation. Experience of organisations such as SEI and others suggests that both consumption and production accounts should be maintained and published regularly as opposed to one or the other; especially as producer-based emissions are required as an input to consumption-based accounts, prior to the reallocation along supply chains and inclusion of imports. There are also methods to weight production-based and consumption-based inventories together, which as mentioned by Peters (2008) could be useful for building on the shared-responsibility for climate change mitigation literature.

Sector specific

3.6 Research of the Sustainable Consumption Institute at the University of Manchester highlights the importance of incorporating both consumption and production-based accounts when considering emissions from agriculture and the food system (Roeder et al 2011). With increasing climate change impacts, agricultural production in some world regions is likely to be favoured while others might be significantly disadvantaged. This will be particularly challenging in regions with a high rise in demand for food and where production is projected to be limited by increasing temperatures. This may result in a greater number of countries relying on a few main producers.

3.7 With this in mind, producers in the less severely affected regions will need to maximise their yields and land-use efficiency to maintain global food supply. To facilitate this, a higher use of agrochemicals, particularly nitrogenous fertilisers is likely to be necessary. This will in turn increase the greenhouse gas burden associated with increased nitrogen fertiliser production and usage, resulting in substantial increases in the national emissions inventory in key producer countries under the existing accounting framework. This could be a disincentive to increase production in these countries and as a result counterproductive to improving global food security and poverty reduction. Using both a consumption and production-based approach would help to classify the source and end-user of emissions and allow production for export in favourable regions without penalising them for increasing their emission to support global food consumption.

3.8 With particular reference to the international transport sectors, under the producer-based greenhouse gas emissions accounting, international shipping (and aviation) emissions are currently excluded, as the majority of emissions are released in international waters (and airspace). Nonetheless, the amount of bunker fuel sold by a nation is reported to the UNFCCC by Annex 1 nations as a memo item. This current method of reporting was described by the Environmental Audit Committee as being inadequate to represent the UK’s actual share of international shipping emissions (EAC, 2009). By taking a consumption-based approach, the emissions released from international shipping (and aviation) would be automatically included in the emission accounts. The international shipping (and aviation) emissions would then constitute part of the embedded emissions that are released along the goods or services supply chain.

3.9 In the absence of a global agreement to control the release of greenhouse gas emissions from international shipping or aviation, the consumption-based approach would place responsibility on the consumer to monitor and mitigate its share of shipping emissions over time (Gilbert et al, 2010). It should be noted however, that shipping is not well represented in existing MRIO-related databases, and this would need to be addressed.

3.10 To report international shipping emissions under a consumption-based approach would require a clearly defined methodology. The main practicality issue is devising the methodology to apportion the emissions released on a shipping journey between the different consumer’s goods or services that are being shipped. This would be most problematic for containerised vessels and vessels that call at multiple ports. However, if the shipping journey was to facilitate the movement of a single good or service, such as coal, then the emissions of that journey would be reported as part of its embedded supply chain emissions.

3.11 Given the island-nature of the UK and increasing demand for consumer goods produced in geographically disparate regions, the use of consumption-based reporting approach could increase the UK’s share of international shipping emissions from an estimated 7 MtCO₂ to 42 MtCO₂ (Gilbert et al, 2010). These numbers assume that the current reporting approach is the amount of bunker fuel sold by the UK and the consumption-based estimate involves a top-down emissions accounting approach based on the total goods imported into the UK.

4. Is there any evidence of industry relocating from the UK to other countries as a result of UK climate change policy?

4.1 This is outside our sphere of research.

5. Would it be (a) desirable and (b) practicable for the UK to adopt emissions reduction targets on a consumption rather than production basis?

(a) Desirable

5.1 The desirable effects would be the UK taking more responsibility globally for the emissions that are generated to support growing levels of domestic consumption. This would increase the share of global emissions over which the UK has influence, and therefore broaden its reach over the climate outcome of an increase in emissions (Bows and Barrett, 2010).

5.2 It could give priority to both decarbonising industry domestically and the avoidance of carbon emissions generated in the supply of goods and services being transferred abroad. It would bring together climate policy and trade, which would tackle the common critique of production-based accounting for not including the emissions associated with imported goods. Under only production-based targets, the consumption of imports and their associated embedded emissions can continue to increase without any monitoring, but a consumption-based target would include these emissions.

5.3 The consumption approach can demonstrate the role that consumption plays in counteracting emissions mitigation efforts, if growing consumption outweighs the efficiency improvements domestically or from imports then this would be evident from the consumption-based accounts. Consumption-based accounts can be used alongside other indicators to consider alternative mitigation options with a more global focus, along with a new perspective for sharing of climate change mitigation commitments and responsibilities across the world.

(b) Practicable

5.4 This can be separated into two parts; whether it is practical to generate the accounts in a consistent manner appropriate for measuring against targets and whether it would be practical or possible to meet any targets set on a consumption basis.

5.5 The sections above detail the methods available for generating the accounts and whether this would be considered a robust method for targeting is dependent to some extent on the method chosen and the data and financial provision for generating and maintaining consumption-based accounts. Nationally, work would need to be done to collect and publish the necessary datasets more frequently in order to produce timely and robust consumption-based emissions estimates for regular accounting and monitoring. It may not be possible to use a full MRIO approach for measuring against targets as it is reliant on global trade data and models, which are not produced annually due to their data intensive production process. Alternative methods such as emissions embodied in trade (see Peters, 2008) or time series with trade (see Peters et al, 2011) could be used to generate annual consumption-based accounts, but this is also likely to require additional work to publish the necessary input datasets more regularly. As Peters, 2008 notes, MRIO is most useful for product or consumption specific studies and a method such as emissions embodied in trade would be more useful for national emissions inventories limited to bilateral trade flows. Solely consumption-based national emissions inventories would suffer from higher complexity and uncertainty and lower transparency than production-based accounts.

5.4 In terms of meeting targets the UK would not have political jurisdiction over all the regions that contribute to its emissions targets in a consumption-based account, as the accounts would include emissions occurring during the production of imported goods and services anywhere in the world. Targets would have to take trade into consideration and in order to meet any targets it may be necessary to take measures to either reduce consumption from certain emission-intensive countries or to help deliver and encourage emissions reductions within those countries. The political economy of climate mitigation may shift to reflect the new accounting approach, but practically, due to the global nature of supply chains and the aggregate nature of the models and accounts, it may be difficult to identify specific countries, or industries within those countries, to target for mitigation efforts. In addition, the time lag between the publication of underlying data and the publication of consumption-based accounts would also make monitoring the account against national targets more difficult.

5.5 Peters (2008) notes that a number of authors have argued that production-based and consumption-based inventories represent two extremes and it would be beneficial to generate a method for shared-responsibility. A number of proposals such as Greenhouse Development Rights framework, Baer et al (2008)13 advocate a type of shared responsibility for mitigation between consumers and producers and Peters (2008) extends the shared responsibility concept specifically to production-based and consumption-based national emissions inventories.

6. What are the potential implications at the international level of the UK adopting a consumption- rather than production-based approach to greenhouse gas emissions accounting?

6.1 If the UK were to adopt consumption-based accounting alongside production-based accounting then this would demonstrate internationally that the UK is taking into account the emissions impacts arising from trade. If the model used to calculate the consumption-based emissions can provide the necessary detail then it could also be used to identify potential carbon leakage from countries with binding targets to those without, demonstrating the UK’s understanding of the importance of making progress towards emissions reductions on a global scale.

6.2 If the UK adopted a solely consumption-based approach to emissions accounting then the UK would have to work closely with other countries to encourage mitigation efforts abroad to reduce the impact of any imported goods to the UK. However, if successful, it would gain greater influence over future climate impacts than is currently the case.

7. Are there any other issues relating to consumption-based emissions reporting that you think the Committee should be aware of?

7.1 Consumption-type approaches can also take a more bottom-up perspective and consider the life-cycle emissions of products consumed. In the example below, the consumption of beef is used to highlight how current production-based emission inventories may significantly underestimate the full greenhouse gas impact of beef consumption.

7.2 The 2008 UK emission inventory report estimates that 353.33 Gg of methane are released by the non-dairy cattle herd through enteric fermentation (McCarthy et al 2010). To illustrate how even just the direct component of emissions (ie excluding emissions associated with feed for cattle) would change if applying a consumption-type framework, a simple method from Walsh et al (2009) can be used. Within this method, the methane embodied in traded beef commodities is required. This can be estimated by considering the direct enteric fermentation emissions throughout each cow’s lifetime associated with the total amount of meat product traded. Including the imports of traded beef, but excluding exports leads to a national consumption-based estimate of 437.32 Gg CH₄—an increase of 24% compared with the existing production-based emission. This is particularly significant given methane’s global warming potential of 25, effectively increasing the production-based estimate by over 2099 Gg CO₂ eq. (Conor Walsh).

October 2011

References

Andrew, Robbie, Glen P. Peters, and James Lennox. 2009. “APPROXIMATION AND REGIONAL AGGREGATION IN MULTI-REGIONAL INPUT-OUTPUT ANALYSIS FOR NATIONAL CARBON FOOTPRINT ACCOUNTING.” Economic Systems Research 21 (3): 311-335. doi:10.1080/09535310903541751.

Paul Baer and Tom Athanasiou of Ecoequity and Sivan Kartha and Eric Kemp-Benedict of the Stockholm Environment Institute, 2008. The Greenhouse Development Rights Framework: The right to development in a climate constrained world. Revised Second Edition.

Bang, J K, Hoff, E, Peters, G P EU Consumption, Global Pollution. World Wildlife Fund. Trade and Investment Programme.

Berglund, Marten, 2011 “Green growth? A consumption perspective on Swedish environmental impact trends using input-output analysis”. Global Energy Systems, Department of Physics and Astronomy, Uppsala University, UPTEC W11 021, ISSN 1401-576: http://www.tsl.uu.se/uhdsg/publications/Berglund_Thesis.pdf

Bäuerle, T, J Graichen, K Meyer, S Seum, M Kulessa and M Oschinski (2010). Integration of Marine Transport into the European Emissions Trading System: Environmental, economic and legal analysis of different options. Germany, Umweltbundesamt (Federal Environment Agency). (UBA-FB) 001372.

Bows, A, and Barrett, J, 2010, Cumulative emission scenarios using a consumption-based approach: a glimmer of hope?, Carbon Management, 1, 161–175.

C E Delft (2009). Technical support for European action to reducing greenhouse gas emissions from international maritime transport. Tender DG ENV.C3/ATA/2008/0016, C E Delft.

Chen, Z M, and G Q Chen. 2011. “Embodied carbon dioxide emission at supra-national scale: A coalition analysis for G7, BRIC, and the rest of the world.” Energy Policy 39 (5) (May): 2899–2909. doi:10.1016/j.enpol.2011.02.068.

Druckman, Angela, and Tim Jackson. 2009. “The carbon footprint of UK households 1990–2004: A socio-economically disaggregated, quasi-multi-regional input–output model.” Ecological Economics 68 (7) (May 15): 2066–2077. doi:10.1016/j.ecolecon.2009.01.013.

EAC (2009). Environmental Audit Committee - Fourth Report: Reducing CO₂ and other emissions from shipping (HC 528). London, House of Commons.

Eyring, V, I S A Isaksen, T Berntsen, W J Collins, J J Corbett, O Endresen, R G Grainger, J Moldanova, H Schlager and D S Stevenson (2010). “Transport impacts on atmosphere and climate: Shipping.” Atmospheric Environment 44(37): 4735–4771.

Gilbert, P, A Bows and R Starkey (2010). Shipping and climate change: scope for unilateral action. Manchester, Tyndall Centre for Climate Change Research and The Sustainable Consumption Institute.

Hertwich, E and Peters, G, 2010. Multiregional Input-Output Database, OPEN:EU Technical Document: http://www.oneplaneteconomynetwork.org/resources/programme-documents/WP1_MRIO_Technical_Document.pdf

IMO (2009). Prevention of air pollution from ships. Second IMO GHG Study 2009. . Update of the 2000 IMO GHG Study—Final report covering Phase 1 and Phase 2, International Maritime Organisation. MEPC 59/INF.10.

Lenzen, M, R Wood, and T Wiedmann. 2010. “UNCERTAINTY ANALYSIS FOR MULTI-REGION INPUT-OUTPUT MODELS—A CASE STUDY OF THE UK’S CARBON FOOTPRINT.” ECONOMIC SYSTEMS RESEARCH 22 (1): 43–63. doi:10.1080/09535311003661226.

Leontief, W, 1970, taken from Kurz, H D, Dietzenbacher, E And Lager, C (1998), Input-output analysis Volume II, Cheltenham, Edward Elgar Publishing Lt.

Leontief, W, 1986, Input-output economics, New York, Oxford University Press.

Machado, Giovani, Roberto Schaeffer, and Ernst Worrell. 2001. “Energy and carbon embodied in the international trade of Brazil: an input-output approach.” Ecological Economics 39 (3) (December): 409–424. doi:10.1016/S0921-8009(01)00230-0.

MacCarthy, J, Thomas, J, Choudrie, S, Passant, N, Thistlethwaite, G, Murrells, T, Watterson, J, Cardenas, L and Thomson, A 2010. UK Greenhouse Gas Inventory, 1990 to 2008: Annual Report for submission under the Framework Convention on Climate Change.

MEPC (2006). Bunker delivery note and fuel oil sampling. MEPC.1/Circ.508, International Maritime Organization.

Miller, Ronald E, and Peter D Blair 2009. Input-Output Analysis: Foundations and Extensions. 2nd ed. Cambridge University Press, 30 July.

Minx, J C, T Wiedmann, R Wood, G P Peters, M Lenzen, A Owen, K Scott, J Barrett, K Hubacek, G Baiocchi, A Paul, E Dawkins, J Briggs, D Guan, S Suh and F Ackerman 2009. “INPUT-OUTPUT ANALYSIS AND CARBON FOOTPRINTING: AN OVERVIEW OF APPLICATIONS.” Economic Systems Research 21 (3): 187–216. doi:10.1080/09535310903541298.

Munksgaard, Jesper, and Klaus Alsted Pedersen. 2001. “CO₂ accounts for open economies: producer or consumer responsibility?” Energy Policy 29 (4) (March): 327–334. doi:10.1016/S0301-4215(00)00120-8.

Paxian, A, V Eyring, W Beer, R Sausen and C Wright (2010). “Present-day and future global bottom-up ship emission inventories including polar routes.” Environmental Science & Technology 44(4): 1333–1339.

Peters, Glen P 2008. “From production-based to consumption-based national emission inventories.” Ecological Economics 65 (1) (March 15): 13–23. doi:10.1016/j.ecolecon.2007.10.014.

Peters, Glen P, Jan C Minx, Christopher L Weber, and Ottmar Edenhofer. 2011. “Growth in emission transfers via international trade from 1990 to 2008.” Proceedings of the National Academy of Sciences 108 (21) (May 24): 8903 -8908. doi:10.1073/pnas.1006388108.

Peters, Glen P, and Edgar G Hertwich 2006. “Pollution embodied in trade: The Norwegian case.” Global Environmental Change 16 (4) (October): 379–387. doi:10.1016/j.gloenvcha.2006.03.001.

Röder, M, Thornley, P, Gilbert, P, Campbell, G 2011. “Adaptation strategies for sustainable global wheat production”. Reframing sustainability? Climate Change and North-South Dynamics. Helsinki, February 2011 (Conference proceedings)

Tukker, Arnold, Evgueni Poliakov, Reinout Heijungs, Troy Hawkins, Frederik Neuwahl, José M. Rueda-Cantuche, Stefan Giljum, Stephan Moll, Jan Oosterhaven, and Maaike Bouwmeester. 2009. “Towards a global multi-regional environmentally extended input-output database.” Ecological Economics 68 (7) (May 15): 1928-1937. doi:10.1016/j.ecolecon.2008.11.010.

Walsh, C, O’Regan, B and Moles, R, 2009. Incorporating methane into ecological footprint analysis: A case study of Ireland. Ecological Economics. 68(7):1952–1962.

Wiedmann, T, Lenzen, M, Wood, R, 2008. Uncertainty Analysis of the UK-MRIO Model—Results from a Monte-Carlo Analysis of the UK Multi-Region Input-Output Model (Embedded Carbon Dioxide Emissions Indicator); Report to the UK Department for Environment, Food and Rural Affairs by Stockholm Environment Institute at the University of York and Centre for Integrated Sustainability Analysis at the University of Sydney. Defra, London, UK. Project Ref: EV02033, July 2008.

Wiedmann, T, Wilting, H, Lutter, S, Palm, V, Giljum, S, Wadeskog, A, Nijdam, D, 2009. Development of a Methodology for the Assessment of Global Environmental Impacts of Traded Goods and Services. Final Report, August 2009. ERA-NET SKEP Project EIPOT. http://www.eipot.eu.

Wiedmann, Thomas. 2009. “A review of recent multi-region input-output models used for consumption-based emission and resource accounting.” Ecological Economics 69 (2) (15 December): 211-222. doi:10.1016/j.ecolecon.2009.08.026.

Wiedmann, Thomas, Harry C Wilting, Manfred Lenzen, Stephan Lutter, and Viveka Palm. 2010. “Quo Vadis MRIO? Methodological, data and institutional requirements for multi-region input-output analysis.” Ecological Economics In Press, Corrected Proof. doi:16/j.ecolecon.2011.06.014.
http://www.sciencedirect.com/science/article/pii/S0921800911002606.

Wiedmann, Thomas, Richard Wood, Jan C. Minx, Manfred Lenzen, Dabo Guan, and Rocky Harris. 2010. “A CARBON FOOTPRINT TIME SERIES OF THE UK - RESULTS FROM A MULTI-REGION INPUT-OUTPUT MODEL.” Economic Systems Research 22 (1) (March): 19–42. doi:10.1080/09535311003612591.

Yunfeng, Yan, and Yang Laike. 2010. “China’s foreign trade and climate change: A case study of CO₂ emissions.” Energy Policy 38 (1) (January): 350–356. doi:10.1016/j.enpol.2009.09.025.