SCGE Models: Relevance and Accessibility for Use in the UK

Introduction

1.1 Background

The UK DfT has charged RAND Europe with the task of providing background information and advice about an emerging branch of economic theory and a newly available tool for understanding the contribution of the transport sector to the general economy, the Spatially Computable General Equilibrium Model (SCGE).

The stimulus comes from the recommendations of the UK SACTRA (Standing Advisory Committee on Trunk Road Appraisal, reporting in 1999), summarised in the following two paragraphs

"11.04 We recommend that the Department undertakes further development of CGE modelling of the total economic impacts of transport schemes with particular emphasis on incorporating endogenous growth, alternative assumptions as to the behaviour of the labour market and better representations of the transport system

11.05 We recommend that the Department investigates the use of CGE modelling in conjunction with a transport model, to explore the size of the discrepancies between total economic impacts and transport impacts in (different circumstances of imperfect competition and external costs)."

The specific questions posed concern a perception that current methods for appraising the contribution of investments in the transport sector to the economy in general may be over-simplistic. In focussing on the immediate user benefits of an investment in affecting travel times, costs, current approaches may fail to address the larger issues of the role accessibility plays in the development (growth or decline) of the economy in general. It could follow that implications for total welfare benefits, and/or GDP, are being misperceived.

Most of the studies reviewed in this report concentrates on flows of goods and economic activities, such as commuting and business travel. We also note that similar issues arise in the case of accessibility benefits to individuals travelling in their free time, which may over time change land prices and taxation. Some benefits initially accruing to travellers will eventually be taken back by landlords, employers tax authorities and so on, as people change houses, jobs, wage status etc. Housing stocks and places of employment also adapt. The desirability of such redistribution can also be relevant for policy appraisal, or tax strategies, in the transport sphere.

There is also the issue of the geographic distribution of benefits as between regions, firms and households. Both the issues of spatial distribution and consumer surplus from improved accessibility are tackled in one particular system, the CGEurope models. these are described in later sections.

SCGEs, as introduced in the context of the SACTRA report have the immediate attraction that they can, in principle, provide a theoretically rigorous background against which the working-through of changes in accessibility can be traced to a final impact on society, that is both producers and consumers. However, these models also have their drawbacks; the 'C' for Computable involves restrictions, in terms of

• the modelled behaviour of individuals and firms, which are restricted to tractable representations in terms of objectives and factors taken into consideration
• the 'typology' of individuals and of commodities, which is very coarse
• the number of discrete geographic areas (zones) which can be treated, which is currently small (relative to conventional transport models) for many realisations (although where this is due to computational burden, this can be expected to reduce over time)
• the representation of transport costs, which in many SCGMs is highly simplified.

Currently, there are other drawbacks, which we describe in some more detail below; for example, where there is a Government 'actor' imposing taxes and investing in infrastructure, there is typically no modelled 'actor objective', no characteristic of a responding in a regular (eg predictable and therefore system-relevant) way to accessibility underprovisions. In other words, there is no 'planning sector' actor or response.

The task in hand is to provide the basic information relevant to a decision to invest resources in developing SCGE capabilities in the UK. This decision has to be taken in the context of knowledge about

• the apparent success or failure of research currently being conducted, and practical use currently being made of SCGEs
• the potential of SCGEs to add value to existing transport evaluation methods, and the relevance of SCGEs to current plans for developing a UK national freight forecasting capability
• the implications in terms of data needs and data availability
• the issues of stability, practicality and computability of the models
• the availability of consulting expertise within Europe and the UK.

This project aims to provide this information, along with access to key documents, and to provide an outline of a process by which DfT could proceed, should it choose to do so.

This report follows a relatively small research project conducted by RAND Europe, with work done in early 2003 and completed in early 2004, prompted by the advice of SACTRA, and the potentially unique possibilities of SCGEs to improve current evaluation - and forecasting - methods. Help and advice has been received from several prominent researchers in the this area, but it should not be assumed that any of the opinions expressed in the report are endorsed by that research community.