Economic analysis of energy efficiency - Work package 3 - Grantham Research Institute on climate change and the environment

The economics and policy of energy efficiency subsidies

Lead partner: LSE

Objective

This work package will focus on estimating the welfare impact of selected energy efficiency policies. First, we will estimate the size of the windfall effect of these policies and investigate how different measures can help reduce windfall benefits in relation to different types of subsidised investments. We will examine what types of energy efficiency investments are more likely to generate windfall benefits, and why. The investigation will draw from the UK experience, but the outcome will be relevant for other energy efficiency incentive programmes in the EU.

The second objective of the work package is to provide a cost-benefit analysis of the national incentive policies implemented in the UK. Because improvements in energy efficiency are often associated with an increase in utilisation (rebound effect), the impact of energy efficiency policies on energy savings is not obvious. Based on statistical techniques, we will assess the energy savings imputable to the policies under study and then perform a cost-benefit analysis. For this purpose, we will seek to quantify the potential for spill-overs of the energy efficiency measures, depending on the degree of geographical disaggregation of the data.

The statistical methodology developed in the framework of this research will then be discussed with the objective of providing guidance to EU policy-makers on data requirement and processing methods to evaluate ongoing and future energy efficiency incentive programmes.

By focusing on the objectives above, our objective is to provide guidance on how to increase the cost-effectiveness of policies aiming to select and fund energy efficiency projects. By reducing the cost and increasing the efficiency of energy efficiency incentive programmes, we expect that more genuine projects could be funded and energy savings made. In return, this could accelerate the rhythm to which energy efficiency targets are reached.

Methodology

To estimate infra-marginal participation, we will use econometric methods. The idea behind the use of such methods will be to compare the share of consumers undertaking home improvements while the policy is being implemented to the likely share of consumers that would have undertaken home improvements even in the absence of the policy. The major difficulty is that the latter is not observed, but must be evaluated. In this respect, a common method consists in comparing similar households that have undertaken home improvements with and without government support. We will therefore design a policy group and a control group. We will then compare the frequency of investments between the two groups, and, provided that these are similar except from the fact that the policy group received government support, we will assess the extent to which the policy resulted in home improvements being more often carried out.

A similar approach will be adopted to estimate the energy saving effect of government supported investments. Again the major difficulty is to construct a credible counterfactual, i.e. what would have occurred in the absence of the policy (Davis et al. 2012). To address this issue we aim at constructing alternative comparison groups composed of households that have common characteristics with those undertaking investments with government support. Various methods will be adopted to evaluate the suitability of the control group, which is to test whether households in the policy group are sufficiently similar to those in the control group expect for receiving the government support. We will then compare energy consumption levels before and after home improvements between the policy and comparison groups. This method is often referred to as difference-in-differences approach. We will take account of a rich set of seasonal and location specific effects. We will investigate heterogeneous effects, for example by household and property characteristics, to gain insights into how non-economic barriers influence the effectiveness of the energy efficiency programmes. The estimated total change in energy consumption will then be compared to the total cost of the programme in order to assess its cost-effectiveness.

Data

For the analysis of UK policies, the research team will principally rely on the information on home improvements and policy support gathered in the English Housing Survey (2008-2012, previously the Survey of English Housing, 1993-2008). In particular, the following questions were asked to a sample of 14,000 households for each wave of the survey:

Which jobs have you (or your landlord/freeholder if applicable) done to your heating or insulation in the last 12 months? (14 possible answers)
Did you receive any grant or other financial help of this kind that paid for some or all of this heating or insulation work in the last 12 months?

These two questions will allow the research team to identify those households that performed energy efficiency investments with and without policy intervention. Furthermore, complementary data sources will include the anonymised National Energy Efficiency Database-framework (NEED) of the UK Department of Energy and Climate Change, which provides UK level data on subsidised insulation efforts and is available to researchers. Likewise, information on UK programmes and policies is available at an aggregated level from the Home Energy Efficiency Database (HEED) of the Energy Saving Trust.