Assessing economic benefits of transport projects using an integrated transport-CGE approach

Abstract

Transport planners and decision-makers require reliable appraisal tools to assess the value of potential development proposals in terms of societal and economic benefits. The well-known cost benefit analysis approach as a commonly used appraisal tool imposes limitations in holistically quantifying the consequences associated with projects across different markets in the economy. The computable general equilibrium framework, as an alternative, can address some of the limitations of CBA by accounting for several properties of the supply and demand of some interacting markets by borrowing concepts from microeconomics for economic behaviour of agents. However, even with consideration of the impacts of a project on the entire economy, behavioural aspects of the transport system are typically neglected. Such aspects are crucial for transport planning purposes as well as the entire economy, so as to appropriately reflect essential properties of the transport system. Trip generation models, are the cornerstone of travel demand modelling, which are aimed to predict the future demand for mobility. In this paper, a trip generation model is integrated into a CGE model to extend the capacity of the CGE model in truly reflecting the attributes of the system and the interacting agents. The developed integrated model is used to assess the consequences associated with changes in attributes of agents and the economy on the transport system as one market among other markets. Further, a consistent data generation platform is proposed where all required data is generated using aggregate data which should be readily available.

Introduction

Trip generation models play a crucial role in the structure of transport modelling frameworks. The classical 4-step transport modelling process includes trip generation as the first step replicating the overall demand for mobility followed by three other steps namely trip distribution, mode choice, and traffic assignment. The trip generation models provide a high-level picture of the travel demand. The breakdown of trips across different modes of transport and destinations is determined in the consecutive steps. This means, the performance of trip generation models highly affects the performance of the following steps (Badoe and Steuart, 1997), which leads to a major impact on the capacity of the model for assessment of policies and proposals. Transport planning is typically formed around policy appraisal tools which are responsible to identify and measure consequences of different polices and projects. An ideal policy appraisal tool should provide a holistic view of costs and benefits of different future scenarios. One of the main promises of transport projects is to improve welfare by facilitating movement within and between cities and, however the existing transport policy appraisal mechanisms are not designed to measure welfare. For instance, transport infrastructure projects such as railroads provide benefits such as decreasing trade costs and increasing income that are expected to lead to higher welfare (DONALDSON, 2018). Since income has an impact on travel demand (PAS, 1986), change of income affects trip generation and mobility pattern of people which consequently contribute to reformation of the transport system. However, the current transport project assessment mechanisms are focused on the improvement of level of service where the overall impact of policies on the society are overlooked. The most commonly used assessment tools are based on the classical cost-benefit analysis (CBA) method. This method imposes limitations as it assumes that market operates perfectly, which is therefore unable to capture the impact of externalities, imperfections, monopolistic firms, taxes and excess demand/supply. In addition, for large transport projects (compared to the economy) where the project impacts the whole economy including other markets, CBA fails to provide a realistic appraisal as it does not account for the linkage between transport projects and other markets (Prest and Turvey 1965) which might be significantly affected as employments might be created or lost.

Another appraisal tool that can be used to economically compare projects is the computable general equilibrium (CGE) model. A CGE model considers the impact of a project on the entire economy by simulating supply and demand in markets using microeconomic functions for economic agents, including both households and firms. The economic agents interact through price mechanisms. Since the supply and demand of each market is a function of all prices in the economy, any change in a market can affect other markets as well. As a result, a CGE model is capable of capturing the interaction between different markets (ROBSON, 2018).

A CGE model can be used as an appraisal tool which considers the impact of the project in the whole economy without the limitation of CBA such as the incapability to evaluate the eventual distribution of benefits among economic agents (Bröcker et al., 2010). A trip generation model is one of the important components of the transport planning framework is expected to be an integral part of the CGE tools to better capture the consequences of transport projects. In this structure, the travel demand is modelled using a trip generation model which then informs the CGE model about the number of trips being generated to and from different industries and between origins and destinations for passenger trips. In previous studies with zone-based travel demand modelling approaches, static pre-estimated factors (rates of generated trips per worked hour estimated from survey data) have been used to exogenously account for trip rates (ROBSON & DIXIT, 2017). Using static factors for trip generation imposes limitations including, but not limited to, the model being unable to capture the impacts of socio-economic attributes of agents as they change given different equilibria obtained from the CGE model. Such attributes such as age (Calvo et al., 2019), industry type (JANG, 2005), income (PAS, 1986), are known to affect the trip generation patterns. This should clarify the importance of having a communication channel between the transport model, in this case the trip generation model, and the appraisal tool, in this case the CGE framework.

This study contributes to the literature by introducing an endogenous trip generation model integrated into the CGE model which enables the CGE structure to directly reflect changes in the socio-economic attributes of agents in the trip generation model. The introduction of an endogenous trip generation model to the CGE-transport integrated model allows the trips to become sensitive to the economic variations determined by the CGE model, and the other way around. This better aligns the proposed CGE structure with the classical 4-step modelling framework which contributes to the introduction of CGE-based travel demand models like an ABM-CGE model. This paper presents the first step of a research direction to fully integrate the 4-step modelling structure into a CGE model where trip rates are not a priori estimated using household travel survey data, but instead are sensitive to economic and demographic attributes while iteratively communicating with the CGE model.

This paper is structured as follows. In Section 2, the literature of trip generations models and appraisal tools for transport projects are reviewed, and the identified research gaps are discussed. Section 3 first introduces the development of the household travel demand dataset used in this study. In the second part of Section 3, the trip generation model is presented. In Sections 4 the application of developed model on two case studies is investigated. Finally, the findings of the study are summarised in the last section with some discussion about future directions of the work.