Modelling & training > Consultancy capability > Economic and statistical analysis for policy evaluation
Economic and statistical analysis for policy evaluation
Rigorous analysis and, where appropriate, modelling are in the best interests of both Ministers and senior officials. They lead to better decisions and improved policy outcomes.”
Cabinet Office Performance and Innovation Unit (2000), “Adding it up - Improving Analysis & Modelling in Central Government” p8.
CE has extensive experience in the design, implementation and interpretation of the application of economic and statistical analysis for policy evaluation. We have an impressive track record of high-quality, innovative and successful modelling, policy and impact analyses undertaken for a wide range of government departments and public agencies.
| | What modelling offers |
| A tool to aid thinking | Modelling offers a tool to inform policy development and monitoring. A model provides an efficient way to: |
| | - develop evidence on which to judge the effectiveness of policy
|
| | - distil and apply the lessons of past experience
|
| | - combine rules of logic and empirical evidence to yield reproducible and structured answers to policy questions
|
| | - identify, through simulation, the net impact of policy changes in cases where behavioural responses and indirect effects complicate the analysis
|
| | - increase the return to an investment in analysis by producing the capability to repeat or extend the analysis at low cost in future
|
| | - build a knowledge base of tried and tested methods, so that corporate memory is maintained when staff move on
|
| | |
| | What Cambridge Econometrics offers |
| Addressing the need | Modelling is a specialist skill. Departments can find it difficult to attract and retain specialist staff. Staff may lack the opportunity to learn to use a wide range of techniques across a variety of applications, and there may be little or no ‘community’ of modellers with whom to exchange ideas. Promotion to a senior post usually entails a move out of modelling. Even where in-house capability is strong, some modelling tasks require a short-term input of additional resources. |
| | |
Modelling is our
business | Modelling is Cambridge Econometrics’ core business. We apply the techniques and disciplines of modelling to a wide range of policy issues. |
| | |
The benefits
of specialisation | - Our senior staff embody many years of daily practice in modelling, with expertise to design appropriate modelling strategies and communicate what analysis the models can and cannot undertake.
- We are able to recruit high-calibre graduates who are enthusiastic about modelling and equipped with the latest techniques.
- We can justify investment in the latest software and, more importantly, in the time required to use it effectively.
- We develop our own software tools to raise our productivity in modelling.
|
| | |
A modelling resource
to support in-house
staff | To be effective, a modelling project requires the combination of: - deep knowledge of the policy issue to be addressed
- expertise to design, develop, deliver and support the model
While we have particular experience in some policy areas, we make no claim to be expert in all. We collaborate with in-house policy specialists to specify an appropriate design and we develop the in-house capability to interpret the model’s output with confidence. |
| | |
Models designed
for ease of use | Policy analysts need to focus their attention on the question to be addressed, not on the complexities of operating the model to get an answer. We are experienced in incorporating ease of use as a key design criterion, with respect to both the structure of the model and the software in which it is implemented. |
Models designed
for ease of future
maintenance |
A model depreciates rapidly if it is not kept up-to-date. But when updating is needed no one wants to expend the scale of resources that were required initially to develop the model. Drawing on our own experience of maintaining very large-scale models, we invest the extra effort required during development to plan for future maintenance, both in software design and in our documentation. |
| | |
| Training to strengthen capacity | Our training courses in econometric modelling for government departments and public agencies consistently receive highly favourable evaluations from participants. |
Examples of our work
Demand for Cars and Their Attributes
Department for Transport
The need for modelling
The Department wanted to identify how people's choice of which car to purchase responds to changes in the purchase price of cars and the costs of motoring.
The method
In order to apply statistical methods, the study required a dataset with variation in both household characteristics and car attributes. The required dataset was constructed by combining an existing dataset of vehicle attributes with the data gathered from a bespoke survey of households that had recently purchased a new or nearly-new vehicle. Econometric methods were applied to estimate the parameters of discrete choice models of vehicle demand. Mixed logit modelling techniques were used to estimate own and cross-elasticities, based upon the approach that has become the best-practice method in the US for modelling vehicle demand, energy use, CO2 emissions, and penetration of alternative fuelled vehicles.
The outcome
The study identified what factors (attributes of car) people take into account when making a purchase decision and how these factors vary according to the characteristics of those making the decision. Estimates of own and cross-elasticities were produced to analyse the responsiveness of vehicle demand to changes in vehicle attributes.
An Analysis of EU ETS Phase 2 and the Effect of Allocation Options on New CHP Capacity
Department for Environment, Food and Rural Affairs
The need for modelling
The government set a target of 10 GWe of installed capacity of Good Quality Combined Heat and Power (GQ CHP) by 2010. Defra commissioned Cambridge Econometrics to quantify the impact of several allocation CO2-based options within the National Allocation Plan (NAP) for the treatment of GQ CHP capacity in Phase 2 of the EU Emissions Trading Scheme (EU ETS).
The method
The tool used for the study was CE's Multisectoral Dynamic Model of the UK economy (MDM-E3), which is the UK's most detailed energy-environment-economy model, designed to analyse and forecast changes in economic structure, energy demand and resulting environmental emissions. MDM-E3 had previously been used to study the effect and costs of introducing a CHP Obligation for Defra. A key innovation in the modelling for this study was to incorporate the various NAP allocation options and their impact on CHP development.
The outcome
The study produced projections for GQ CHP to 2015 under various options that give a subsidy to CHP developers. It also determined the number of the allocations for a new entrant reserve in Phase 2 of the NAP that would be required to achieve the 10 GWe target by 2010 and 2015, under low, medium and high carbon price scenarios. All policy options were considered relative to the 'business-as-usual' baseline (ie the Phase 1 approach). The study provided a key analytical input to the government's treatment of GQ CHP in Phase 2 of the EU ETS.
Analysis of Postcode-Based Income Estimates and a Feasibility Study of Using Linked Datasets to Evaluate Area-Based Initiatives
Department for Work and Pensions
The need for modelling
This was an exploratory study commissioned as a first step in a comprehensive programme of work to gain better understanding of income variation, poverty and joblessness among disadvantaged groups at a fine level of spatial Disaggregation.
The method
The research was undertaken in two parts. In the first stage, the objective was to explore geographical variations in personal incomes at various levels of disaggregation using the postcode-based CACI income data. The objective of the second stage was to develop a conceptual framework with which to consider the possible causes and consequences of income deprivation, and to explore the availability of datasets which could be used to examine the incidence and depth of poverty at the local level in the UK.
The outcome
The study highlighted wide variations in household incomes, even within the same neighbourhood. The analysis demonstrated the strength of the postcode-based income dataset and its potential usefulness to policy makers as a tool for the design and assessment of policy. It also derived a simple conceptual framework with which to consider some of the key interfaces at the local level between labour and housing markets that could help to explain why concentrations of households with low incomes form.
Modelling Good Quality Combined Heat and Power Capacity in the UK to 2010
DTI/DEFRA
The need for modelling
The Government set a target of 10GWe installed capacity of Good Quality Combined Heat and Power (GQ CHP) by 2010. The DTI and DEFRA commissioned Cambridge Econometrics to quantify the impact of existing government support measures on the uptake of CHP.
The method
The tool used for the study was CE’s Multisectoral Dynamic Model of the UK economy (MDM-E3), which is the UK’s most detailed energy-environment-economy model, designed to analyse and forecast changes in economic structure, energy demand and resulting environmental emissions. MDM-E3 had previously been used to study the economic and environmental implications of achieving the Government’s target for the Combined Heat and Power Association (CHPA). A key innovation of the latest modelling was to extend MDM-E3 to incorporate a detailed representation of CHP technologies and cost-benefit decisions.
The outcome
The study produced projections of the level of GQ CHP capacity in the UK to 2010. It also assessed the contribution of elected support measures to the growth in GQ CHP capacity and the sensitivity of these projections to various fuel prices and modelling assumptions. This study provided key analytical input for the Government’s Strategy for Combined Heat and Power to 2010.
An Input-Output Approach to the Analysis of Pricing
Bank of England
The need for modelling
The Bank wanted an analysis which traced explicitly the links between fundamental cost drivers at the industry level and the headline rate of inflation.
The method
CE used an innovative approach, developing an analytical framework based on input-output analysis to study the interactions between prices in different industries and the costs of ‘primary’ inputs (labour, capital and imports, including the effect of exchange rate changes). In this way, changes in the output price of any industry were decomposed into the impact of changes in the costs of each of the primary inputs and the impact of total factor productivity growth in each industry. The analysis was then extended to retail prices, taking account of the additional impact of changes in the prices of imported consumer goods, changes in product taxes, and changes in distribution margins.
The outcome
The framework was implemented for annual data of 41 industries across the whole economy and covering the period 1970-98. The framework was used to examine whether the nature of these relationships has changed in the more recent period, notably to test whether the tendency of producers to absorb or pass on cost increases in the short term had changed.
Updating a study to estimate the disamenity costs of landfill
DETR
The need for modelling
The Department wanted an estimate of the disamenity costs of landfill in the UK, to disseminate the findings more widely within Europe, and to inform the development of policy on the landfill tax and the new UK regulatory environment, including the Landfill Directive.
The method
This project updated an innovative study undertaken two years previously for the Waste Strategy Division of the DETR. It was led again by Cambridge Econometrics, in association with EFTEC and WRc. The approach measured the value placed on the disamenity by identifying the variations in house prices attributable to proximity to landfill sites. The project enhanced a landfill database and developed and applied the hedonic valuation methodology for estimating the lifecycle fixed external costs of landfill for different categories of landfill site in the UK. The study drew on the AHPD Geographical Information System developed by CE to explain differences in house prices on the basis of the characteristics of houses and their location.
The outcome
The analysis produced an objective estimate of the value placed by the market on the disamenity of landfill sites, and an estimate of the distance from a site over which this effect could be detected. (The disamenity value in the original study informed consideration of the appropriate value for the landfill tax.)
For further information contact:
Rachel Beaven
Director