یک منطقه برای انتشار کربن آن منطقه چگونه مسئول است؟تجزیه و تحلیل تجربی تعادل عمومی

CO2 reduction targets tend to be set in terms of the amount of pollution emitted within a given region. However, there is increasing public and policy interest in the notion of a carbon footprint, or CO2 generated globally to serve final consumption demand within a region. This raises an issue in that, despite the local economic benefits, the latter involves effectively absolving the region of responsibility for CO2 generated in the production of exports. Using a CGE model of Wales, we illustrate by simulating an increase in export demand for the output of an industry (metal production) that is both carbon and export intensive and generally produces to meet intermediate rather than final demands. The key result is economic growth accompanied by a widening gap between regional CO2 generation and the carbon footprint, raising questions as to the identification of precisely ‘whose’ carbon footprint these additional emissions should be allocated to. Highlights ► We consider the wisdom of attributing regional emissions to external final consumers. ► We integrate input-output and CGE analysis to examine pollution generation under different accounting principles. ► We develop an innovative CGE application including consideration of emissions under a consumption accounting principle. ► We show that significant economic benefit may accrue to a region from export-led expansion of a carbon-intensive industry. ► However, this is accompanied by much larger growth in production accounting emissions than the region’s carbon footprint.

The 2009 Copenhagen Climate Change conference focused attention on the methods and underlying principles that inform climate change targets. Climate change targets following the Kyoto Protocol are broadly based on a production accounting principle (PAP), or emissions produced within given geographical boundaries of the economy in question. An alternative approach is a consumption accounting principle (CAP), where the focus is on emissions produced globally to meet consumption demand within the national (or regional) economy (Munksgaard and Pedersen, 2001). Increasingly popular environmental footprint measures, including ecological and carbon footprints, attempt to measure environmental impacts based on CAP methods. The perception that human consumption decisions lie at the heart of the climate change problem is the impetus driving pressure on policymakers for a more widespread use of CAP measures. Globally the emissions accounted for under the production and consumption accounting principles would be equal. However, at regional or national level emissions embodied in trade lead to differences under the two principles. Specifically, under a PAP measure, the generation of emissions in producing goods and services to meet export demand is charged to the producing region or nation's emissions account. Under a CAP measure, these emissions would be charged to the region or nation where the final consumption demand charged with ultimately driving this activity may be located. That is, under CAP, emissions embodied (directly or indirectly) in a region or nation's imports replace emissions embodied in export production, alongside domestic emissions to support domestic final consumption (which is common to both measures). However, as public and policy enthusiasm for CAP measures grows (see Wiedmann, 2009, for a review), this paper raises the question as to whether it is appropriate entirely to attribute responsibility for emissions resulting from production decisions throughout (often quite complex) supply chains to final consumers, particularly where these consumers may be located in other regions, nations and jurisdictions. To illustrate our argument we first produce base year results for regional carbon emissions calculated on PAP and CAP principles to reveal the differences in, and perspectives offered by, the two approaches. We then take the example of a decision to increase production in a regional industry where production is both carbon intensive and export intensive and examine the differential impacts on the alternative measures. We also examine the economic impacts of this increased activity on the regional economy. The economic benefit derived by local consumers raises questions as to whether it is appropriate to absolve them of all responsibility for emissions embodied in export production. We believe that this provides a first step in the process of understanding the concept of shared responsibility for pollution generation based on key economic indicators such as GDP/value-added (see Lenzen et al., 2007). Moreover, the case study focuses on an industry where the output produced tends to be used as an intermediate input to other production sectors (be they domestic or external) rather than directly serving final demands. This complicates matters in terms of identifying the location of the final consumers to whom emissions embodied in export production should be allocated. The analysis involves two empirical techniques. The first is input–output accounting. Application of regional and interregional input–output accounting techniques to attribute pollution generation to different production and consumption activities has become commonplace particularly in the ecological economics literature (see Munksgaard and Pedersen, 2001, and Turner et al., 2007, for methods; and Wiedmann, 2009, for a review). A CGE framework (which integrates the input–output accounts as its core database) is then employed to model the economy-wide impacts of a change in activity, and the results are used to derive ‘post-shock’ input–output accounts that may be employed to examine impacts on pollution generation under both PAP and CAP measures. The key advantage offered by the CGE analysis is a more flexible and theory-consistent approach to modelling changes in both production and consumption activity levels than is possible using a conventional demand-driven input–output model, particularly in tracing factor market adjustments and resulting price and income induced effects in different time periods. The empirical example in this paper focuses on a current policy issue in the case of Wales, a region of the UK with devolved responsibility for sustainable development. We provide a brief overview of the policy context of the Welsh case study in the next section. However, while some of the issues raised may be of specific interest to Wales, we contend that similar types of problems are faced by both regional and national policymakers around the world. In the third section we use the input–output accounting framework to consider base year carbon measures for Wales under CAP and PAP. This is followed in the fourth section with an overview of the CGE model and discussion of the results of simulating an increase in export demand to Welsh metal manufacturing in the fifth. Discussion and conclusions follow in the final section.

The key result from the integrated IO and CGE analysis above is that the Welsh economy benefits from an export-led economic expansion focussed in a highly carbon-intensive industry, but with an environmental cost in that CO2 generation within Wales rises (PAP) by more than the increase in GDP and consumption. This is evidenced by the gap between the CAP and PAP measures in Fig. 1. The estimated carbon footprint (CAP) does rise, particularly with increased ‘pollution leakage’ through increased carbon embodied in imports. However, the much smaller increase in CAP than in PAP, taken alongside the base case scenario where (perhaps unusually for a developed economy) Wales runs a ‘carbon trade surplus’, suggests that Wales would benefit from a shift in accounting perspective towards carbon footprint type measures. However, such a shift would raise an important issue in terms of how and to whom responsibility for the additional increase in PAP emissions over that in CAP would be attributed. This may be more straightforward, and pertinent, in the case of emissions embodied in exports to the rest of the UK and future research involving an interregional input–output and CGE modelling framework may help inform in this respect (as well as raising interesting issues regarding the location of carbon-intensive heavy manufacturing industries in peripheral regions). However, more generally it may also be argued that Wales has instigated the change in activity brought about by the export expansion, particularly given the investment made to facilitate it. Moreover, Welsh decisions could further impact the structure of the economy and pollution problem under both CAP and PAP perspectives (for example if firms choose to import some of their CO2 requirements in order to lower their own direct emissions). Issues such as these raise questions as to what the CAP and PAP impacts tell us in terms of the sustainability of economic growth and who should be held responsible for carbon generation in different jurisdictions. Perhaps the answer in trying to take a more consumption-orientated focus is not as straightforward as subtracting carbon embodied in exports and adding that embodied in imports, but rather some form of shared responsibility criteria is required, between producers and consumers generally and/or between importing and exporting countries. The issue of how economic benefit may impact on carbon measures is addressed in a literature that focuses on the development of a shared responsibility measure. For example, Lenzen et al. (2007), in an extended input–output analysis, suggest that a share of responsibility should be retained by producers based on the value added contribution of output. This would be a possibility in the case examined here and CGE analysis of the type presented here would help motivate and consider what a shared responsibility measure should focus on in different scenarios. Similarly, it may be useful to develop interregional input–output, CGE and/or fuller lifecycle analysis methods to consider different accounting perspectives, such as supplier responsibilities for the ultimate consumption use of their production outputs. Future research may also be usefully extended to cost–benefit analysis to consider different accounting perspectives in the context of consumer and/or producer ‘willingness to pay’ for the environmental costs that accompany the additional economic benefits generated by a scenario such as that modelled here. More generally, value may be added in future research by using input–output methods to analyse the sequence of both economic and environmental impacts throughout the target economy and on its trade partners using the post-shock input–output tables generated using the method proposed here. For example, Dietzenbacher and Romero (2007) propose a method to break down the sequence of impacts from a shock to a single sector through rounds of backward and forward multiplier effects that would permit a decomposition of results modelled throughout the adjustment process using CGE techniques.11