ASYMMETRIC ENERGY REBOUND EFFECT ACROSS DIFFERENT HOUSEHOLD INCOME GROUPS

摘要

OverviewImproving energy efficiency in consumption can be an effective instrument to reduce final energy demand. However, efficiency improvement in energy use have been often associated with the rebound effect occurring when the expected energy savings from the enhanced efficiency are partially offset by the impact of the initial reduction in the price of the energy service delivered.Despite the general tendency to consider the rebound effect as the undesired consequence of energy efficiency improving policies, energy rebound has been also associated with a series of macroeconomic befits triggered by the increased energy efficiency. These include lower unemployment, increased investment, higher consumption of non-energy goods, and a stimulus to GDP (Lecca et al. 2014). This raises the need to weigh these benefits when assessing the effectiveness of energy efficiency policies.Past studies have analysed the macroeconomic impact of increasing household energy efficiency in a general equilibrium setting. Lecca et al. (2014) find that a broad brush 5% energy efficiency improvement in UK households can stimulate the wider economy through an increase and change in the pattern of aggregate demand. However, the higher demand for domestic goods puts upward pressure on prices, crowding out exports.Figus et al. (2016) extend Lecca et al. (2014) by analysing the implications of moving from a national (UK) to a regional case study of Scotland. They find that in a regional computable general equilibrium (CGE) setting, characterised by a more open goods market and job market, interregional migration contributes to restore the lost competitiveness in the long run, this is by pushing wages and consumption prices back to their baseline levels, delivering thereby a bigger economic stimulus.However, these studies consider the aggregate household sector, and so do not allow an investigation of which households are likely to benefit more from an energy efficiency improvement. Additionally, they consider an across-the-board efficiency improvement without focusing on improvements in particular types of energy use.To the best of our knowledge, at the moment no CGE studies decompose the effect of household energy efficiency improvements among different groups of household. In a partial equilibrium setting Murray (2013) and Chitnis et al. (2014) calculate the rebound effect from different energy policies in the household sector, accounting for five household income groups. Their main finding is that low income household are associated with the highest rebound effect, because they spend a greater share of their income on energy intensive or GHG intensive goods and services.In this paper we study the economy-wide impact of an illustrative 10% efficiency improvement in (a) household gas consumption and (b) refined fuels use in private transport, using the case of Scotland. We consider the impact across five household income bands to evaluate how energy efficiency improvements can affect different socioeconomic groups.MethodsThe macroeconomic impacts of improving household energy efficiency are analysed using a CGE model for Scotland called AMOS-ENVI. This is a dynamic CGE model with forward-looking investment and consumption decisions, designed to analyse environmental and energy disturbances in a regional setting.The model accounts for 30 different productive sectors, including 6 supply chain energy industries, and includes information about five Scottish household income groups, the Scottish Government and imports and exports to the rest of the UK (RUK) and to the rest of the World (ROW). Wages are determined within the region in an imperfectly competition setting, using a wage curve where the real wage is negatively related to unemployment rate. Interregional migration of workers occurs according to a pure flow migration function where in or out migration occurs in response to the difference between national and regional real wage and unemployment rates.This model is novel, and differs from Lecca et al. (2014) in at least four significant ways. First, it is a regional model, with a more open labour and goods market. Second, we introduce a more detailed nesting structure in the utility function. In the previous version consumption was a simple combination of energy and non-energy goods. Here we distinguish between energy used for transport (motive energy) and energy used for heating and lighting (non-motive energy), and other non-energy goods. Third we disaggregate the household sector into 5 income groups. Fourth, as well as moving to a regional model, we update the dataset to a 2010 SAM for Scotland.We consider an efficiency improvement as being any technological change which allows households to consume the same bundle goods as before but using less physical energy in doing this. This means also that the value of energy in efficiency units has increased.The rebound effect is measured as being the ratio between potential energy savings (PES) and actual energy savings (AES). The PES correspond to the pure engineering effect, for example improving efficiency by 10% and saving 10% of energy. The AES are calculated as the proportionate change in a specific energy use, for which efficiency has improved, as the result of the full general equilibrium adjustments.ResultsHere we report results from two simulation experiments. In the first we increase by 10% the efficiency of fuels use in private transport. In the second simulation we increase by 10% the efficiency of gas consumption. We focus on long-run results, reporting only information about total household consumption, and consumption of the lowest and the highest household income groups. In the full paper we will discuss in detail the full set of general equilibrium impacts accounting for short and long run, and reporting period by period adjustments. We will also report results for all the five income bands.Results from Simulation 1 show that following the efficiency improvement household consumption of fuels in private transport decreases by 4.5%, and total energy consumption decreases by 0.9%. Low income households decrease their fuel consumption by more, 4.6%, whereas the highest income reduce their by 4.55%. Higher income households use motive energy (private and public transport) more intensively than lower income households, therefore they benefit more from the cost reduction. In fact, following the efficiency improvement, the first household group spend 1.4% less of their income on motive-energy while the highest income group saves only 1.3%. The consumption of non-motive energy (heating and lighting) increases by 0.06% for the lowest income households and progressively increase as income rises, to the 0.11% increase of the highest income households. In this case, high income households are better-off, and are able to spend 0.9% more on non-energy goods, while low income households can only spend 0.5% more. The calculated general equilibrium rebound effect in household energy use is 54% for the low income group while it increases to 54.5% for the high income group.In simulation 2, total household gas consumption decreases by 8.9%. Total energy consumption by 1.13% which is more than what we observed in the transport case. In this scenario low income households are better-off, because they use gas more intensively. In the long-run, the lowest income households consume 8.93% less gas, while the highest income consume almost 9% less. There is some substitution towards other energy sources, with an increase in electricity consumption of 0.13% for the lowest group, and 0.07% for the highest income group. As a result of the efficiency improvement, low income households can spend 0.13% more of their income in non-energy goods, while high income group only 0.06%. The calculated rebound in this case is higher for the low income households, 10.7% while it is lower for the high income households 10.1%.ConclusionsResults from simulations show that there is an asymmetric household response to improvement in the efficiency of different types of energy uses. When energy efficiency is improved in motive energy, such as fuels use in private transport, high income households are better-off, in terms of extra resources available for consumption, but save less energy. This leads to a higher rebound effect. When efficiency is improved non-motive energy, such as gas consumption, low income households are better-off, indicating a higher share of income available for consumption, including energy consumption which goes down by less, producing the highest rebound. Additionally, the calculated rebound effect is lower in the gas efficiency case, indicating that improving energy efficiency in gas would be a more effective policy to when the goal is to maximise the reduction of energy use.These results partly contradict findings of previous partial equilibrium analysis studies, such as Chitnis et al. (2014) and Murray (2013) which find that low income household always associated with higher rebound. In fact, here we find that in a general equilibrium setting, improving energy efficiency in private transport would deliver a higher rebound for higher income households, because they use this good more intensively.