DEVELOPMENT OF AN EMBEDDED CARBON EMISSIONS INDICATOR A research…

DEVELOPMENT OF AN
EMBEDDED CARBON
EMISSIONS INDICATOR




A research report to the Department for
Environment, Food and Rural Affairs by
the Stockholm Environment Institute and
the University of Sydney




                  Integrated Sustainability Analysis 
DEVELOPMENT OF AN EMBEDDED CARBON EMISSIONS
INDICATOR

Producing a Time Series of Input-Output Tables and Embedded
Carbon Dioxide Emissions for the UK by Using a MRIO Data
Optimisation System




Final Report to the Department for Environment, Food and Rural Affairs



Project ref.: EV02033

Project website accessible via: http://randd.defra.gov.uk

June 2008




Suggested citation for this report:

Wiedmann, T., Wood, R., Lenzen, M., Minx, J., Guan, D. and Barrett, J. (2007)
Development of an Embedded Carbon Emissions Indicator ­ Producing a Time
Series of Input-Output Tables and Embedded Carbon Dioxide Emissions for the UK
by Using a MRIO Data Optimisation System, 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, June 2008. Defra, London, UK




Dr Thomas Wiedmann, Jan Minx, Dabo Guan, Dr John Barrett
Stockholm Environment Institute, University of York

Prof Manfred Lenzen, Richard Wood
ISA ­ Centre for Integrated Sustainability Analysis, University of Sydney

Views expressed in this report are those of the authors and do not necessarily reflect those of Defra.
     Executive Summary

1.   This report describes work and results from the project `Development of an
     Embedded Carbon Emissions Indicator', undertaken by the Stockholm
     Environment Institute and the University of Sydney on behalf of Defra (Project
     Ref.: EV02033). The work is a follow-up of a previous project where the most
     appropriate approach to constructing a robust account of impacts of trade and
     thus overall consumption in a headline indicator for Sustainable Development
     was identified. The present work describes the implementation of such a model.
2.   The project aim was to develop and implement an initial data and model
     framework (called "UK-MRIO 1") for the flexible adaptation of national input-
     output and environmental databases for use in a multi-region environmental
     input-output model in the future, thus setting the basis for detailed multi-country
     analyses of environmental impacts associated with UK trade flows. This aim
     was not only achieved but actually exceeded in that a fully functional MRIO
     model with four regions (UK + three world regions) was assembled and a time
     series of balanced input-output data and embedded CO2 emissions was
     produced on the full 123 sector level.
3.   An updated review of recent literature on the estimation of emissions
     embedded in international trade confirms the trend towards accounting for
     'consumer emissions' ('carbon consumption') in various countries. In 2007
     alone, a respectable number of models has been developed worldwide in order
     to estimate emissions embedded in international trade of numerous countries
     and regions. Almost all of the studies present input-output based approaches
     and the use of multi-region input-output models is already well established.
4.   As a tangible outcome of the current project we have constructed a time series
     of input-output tables for the UK from 1992 to 2004 by using the novel matrix
     balancing procedure CRAS, which is able to handle conflicting external data
     and inconsistent constraints. These tables are similar to the "Analytical IO
     Tables 1995" published by the Office for National Statistics (ONS) and include
     symmetric input-output tables (SIOT) for domestic transactions and imports for
     each year from 1992 to 2004. These tables represent an approximation of real
     economic activity close and robust enough for modelling purposes. The full time
     series of SIOTs also fills a gap in the public availability of symmetric tables
     which is due to an ongoing modernisation programme of the National Accounts.
     The ONS plays an important role in that it holds essential economic and
     environmental data that could help to improve the accuracy and policy
     relevance of the model.
5.   We have also calculated a time series of direct and indirect carbon dioxide
     emissions associated with UK economic activities was produced, in particular
     emissions that are embedded in UK trade. Consumer emissions (CE) are
     significantly higher than producer emissions (PE) or the UNFCCC national total


                                           1
     (in 2004, CE are 132 Mt or 21% higher than PE and over 200 Mt or 37% higher
     than the national total reported to the UNFCCC, including overseas territories.
6.   CO2 emissions embedded in imports (EEI) are higher than emissions
     embedded in exports (EEE) for all years and there is a clear trend towards
     increasing EEI, which went up from 4.3% of producer emissions in 1997 to 21%
     in 2004.
7.   From the three world regions investigated in this project, the Rest of the World
     region (which includes Asia) contributes most to EEI. This dominance has
     increased sharply in the last couple of years while EEI from non-European
     OECD countries have fallen significantly at the same time.
8.   Another interesting result is that from all emissions embedded in UK exports
     (EEE), 27% came from imports to the UK in 1992; this figure increased steadily
     over the years ending up with 39% of EEE coming from import sources in 2004.
9.   The results for embedded CO2 emissions from the UK-MRIO 1 model are in
     line with findings from other researchers. Previous studies applying a range of
     different methodologies (SRIO, MRIO, MFA) also suggest that more embedded
     CO2 emissions are imported to the UK than exported.
10. Most time and resources of the project were devoted to compiling initial data,
    estimating missing data and balancing conflicting data in the right way as this is
    the most crucial part of a MRIO framework. This ensures consistency,
    robustness and repeatability of the whole approach. A sensitivity analysis
    based on Monte-Carlo simulation will be undertaken as a follow-up project in
    order to quantify the uncertainty associated with the results.
11. The UK-MRIO system is based on a supply and use table (SUT) framework.
    This has several advantages, such as better data availability for initial
    estimates, flexibility of expansion, conservation of sector disaggregation and
    the option of adopting different technology assumptions for the creation of
    symmetric tables.
12. Availability of input-output data for both the UK and the Rest of the World
    (ROW) was reviewed and assessed. Although SUTs are annually published by
    ONS, these are not suitable for modelling purposes straight away and therefore
    had to be supplemented with information from Eurostat and balanced before
    they could be used. Crucial information such as imports and transition matrices
    are only available from the UK Analytical Tables from 1995 and thus it had to
    be assumed that the structure of these tables would not change over a period
    of twelve years. It is hoped that the modernisation of UK National Accounts will
    eventually provide more up-to-date information useful for (environmental) input-
    output modelling.
13. For ROW data, several data sources were evaluated towards their suitability for
    the UK-MRIO 1 model. Data for three major world regions from an earlier study
    could be obtained from the Netherlands Environmental Assessment Agency
    (MNP) for the years 1997 and 2001. These transaction matrices from the
    Global Trade Analysis Project (GTAP) are at a 30-sector level, comprise


                                          2
     OECD-Europe, other OECD and non-OECD countries and were deemed to be
     best suited for the initial UK-MRIO 1 model as they cover two years and are
     compatible with the scope of the project.
14. Trade data were taken from HM Revenue and Customs, but only the years
    from 1995 onwards could be covered with this data source. Great care was
    taken to obtain an accurate picture of imports to the UK from the three world
    regions. Total imports were brought in line with totals in the official SUTs
    provided by ONS.
15. A challenge is posed by detail and classification differences between the
    economic and environmental (CO2) data for both the UK and the three world
    regions. CO2 emission intensities can only be distinguished for 76 sectors in the
    UK and 18 sectors abroad, although the model is set up with 123 and 30
    sectors, respectively. This is a significant limitation of the current model which
    needs to be addressed in future work.
16. The current model is a major step towards a fully fledged multi-region input-
    output model featuring multidirectional trade of a substantial number of UK
    trading partners. UK-MRIO 1 is already the most detailed and comprehensive
    modelling approach for the estimation of CO2 emissions embedded in UK trade
    with high relevance for national and international environmental policy-making.
    The model can be used to produce a range of outputs, for example more
    detailed results for industry sectors, foreign economies, or other pollutants.
17. We suggest that the results from the current UK-MRIO 1 model be published in
    two ways. First, academic publications in peer-reviewed, scientific journals
    should be sanctioned and supported in order to get critical feedback on the
    methodology from the wider scientific community. Second, an 'embedded CO2
    indicator' showing a time series of CO2 emissions from a consumption
    perspective ("carbon consumption") should be considered for publication with
    official UK statistics, alongside already existing greenhouse gas emission
    trends. This would give a more complete picture of emissions induced by UK
    economic activity. Even further revisions of method and data are not thought to
    generally refute the clear and robust trend that has emerged for consumer
    emissions.
18. Our recommendations for further research include general model expansions
    and enhancements ("UK-MRIO 2"), improvement of CO2 and other
    environmental data, inclusion of individual countries and multi-directional trade,
    further sector disaggregation, and a comparison of currency conversion
    options.




                                          3