7. Conclusions and recommendations
This project was commissioned by the Scottish Government to use our experimental AMOSENVI energy-economy-environment CGE model of the Scottish economy to look at the economic and environmental impacts of a number of scenarios that may be expected to reduce the level of greenhouse gas emissions, with specific attention to CO2.
The analysis presented here is experimental and constrained by two broad factors:
1. The need to further develop the Scottish input-output tables for the purposes of examining energy-economy-environment issues.
2. The need to further develop the AMOSENVICGE modelling framework to look at a wider range of issues.
We return to both of these points below and offer some recommendations for strengthening Scotland's analytical capacity in this area.
However, while the analyses presented here are constrained on both these points, they do demonstrate the potential contribution that applied empirical general equilibrium analysis can make to the policy process. For example, the energy efficiency analyses in Section 2 demonstrates why a general equilibrium framework is essential in assessing the nature and scope of rebound effects, even when improvements in energy efficiency are focussed in a single sector/activity. More generally, all the scenarios reflect the importance of capturing interdependences between different sectors and activities.
However, as emphasised from the outset, the current AMOSENVICGE modelling framework should be regarded as experimental in nature as it is still at a very early stage of its development. It has not been possible to simulate all of the scenarios that Scottish Government expressed an interest in at the start of the project - for example, policies aimed at reducing household energy use (though, in Section 4, we do attempt to simulate the impacts of reduced household incomes that may result from such policies). We have also been unable to simulate scenarios involving carbon, capture and storage, which we hoped could be examined through changes in capital efficiency, but the current model, could not give us sensible results. However, it is important to stress that, in principle, a CGE modelling framework can be used to simulate a very wide range of scenarios (as reflected in the literature review in Technical Appendix 1). Development of the AMOSENVI framework is currently ongoing through various EPSRC and ESRC funded research projects being carried out by the regional and energy modelling teams at the Fraser of Allander Institute, University of Strathclyde. This development is being informed by the findings of the literature review reported in Technical Appendix 1 of this report. For example, we are exploring different ways of introducing energy as a factor of production. We will also directly explore energy use in final consumption (again, informed by the literature review findings). However, it is important that the policy community have and take opportunities to input their ideas and priorities to the model development process.
Recommendations regarding data issues
In terms of the first point raised above, the availability of appropriate data and need to develop the Scottish input-output tables, it is important to note that the benefits of doing so would not be limited to better informing CGE models. Turner (2008b) reports the following consensus among participants at a workshop held in March 2008, sponsored by the Scottish Environment Protection Agency to inform the Scottish Government's Steering Group on Additional Measures of Progress, to discuss the potential role of Scottish environmental input output accounts (with particular attention to carbon counting):
"Again, a number of interesting points/questions worthy of further consideration were raised:
- While the development of the IO framework is resource-intensive, if we have faith in market-based solutions to the problem of climate change, we absolutely need to adopt an IO approach.
- Uses of an environmental IO approach are not limited to footprint calculations. It would facilitate the construction of a wide range of environmental indicators. Therefore, it is likely to represent 'good value for money' to policymakers.
- IO analysis would allow us to develop a better understanding of domestic and direct emissions generation as well as the indirect effects that can be measured through multiplier analysis".
Turner (2008b, pp.5-6)
Turner (2003) considers the requirements of constructing Scottish environmental input-output accounts (perhaps adopting the NAMEA - National Accounting Matrix including Environmental Accounts - advocated by Haan, 2001, and developed for the UK by Vaze, 1999) and attempted to construct an experimental environmental input-output framework for Scotland. She concludes that there are two main problem areas that must be considered before a sectorally disaggregated economic-environmental database can be reported. "These are:
1. The availability of region-specific data for Scotland on sources and generation of emissions.
2. Even if region-specific emissions data of an acceptable quality are available, there is the question of whether these can be reported for a sectoral breakdown that is consistent the 1992 Standard Industrial Classification ( SIC) used in the economic accounts. If policy is orientated towards influencing activity in economic sectors, clearly there are benefits to environmental data being presented in a format that is consistent with existing economic accounts."
Turner (2003, p.44)
We would add two further concerns/issues that should be addressed in developing the Scottish input-output tables for analysis of energy and environmental issues. First, aside from issues regarding environmental data and compatibility with economic accounting, there is also an issue over over-aggregation of key activities in the current 128 sector input-output classification ( IOC). For example, for issues relating to recycling and reduction of waste, Allan et al (2007b) argue that it would be useful to disaggregate the waste disposal sector ( SIC 90002) from the IOC sector 119, which also includes sewage, sanitation etc. For a number of the types of scenarios we have attempted to simulate here, and to better understand the link between electricity production and supply and greenhouse gas emissions, it would seem extremely important to consider breaking down the current IOC 85 electricity production sector to identify different elements of the supply and distribution chain and different generation technologies.
Second, it is important that any environmental input-output accounts be constructed in the analytical form that is appropriate for input-output multiplier analyses and other accounting and modelling techniques, such as CGE analysis. Vaze (1997) constructed such a framework for the UK. However, since then any economic-environmental and NAMEA accounting has been built around the supply and use tables. The lack of analytical IO tables for the UK since those reported for 1995 is a serious problem that is commonly raised in the policy and consulting arena. On the other hand, one of the key strengths of the Scottish input-output tables is the variety of formats accounts are made available, and we would argue that this should be extended to any environmental extensions.
A final point regarding data provision, the results of sensitivity analyses reported here highlight the need to generally improve the modelling infrastructure for Scotland. Model development is most likely best carried out within the expert research community, but this requires public support, for example, but not limited to, from the UK Research Councils (who are already supporting our own model development under our various current EPSRC and ESRC projects). However, there is also a need for wider data provision than the environmental input-output recommended above. For example, the availability of appropriate data for the econometric estimation of key model parameters, such as energy demand relationships.