Bus Rapid Transit versus road expansion to alleviate congestion: A general equilibrium comparison

Highlights

• Bus Rapid Transit (BRT) and road expansion projects actually proposed for Beirut are evaluated using a general equilibrium model.

• BRT generates significant benefits by reducing traffic congestion, increasing consumer utility, and yields a public sector fiscal surplus.

• Road expansion increases consumer utility but results in a fiscal deficit requiring borrowing.

Abstract

Bus Rapid Transit (BRT) and highway expansions actually proposed for Beirut are compared using an urban computable general equilibrium model. The model has two geographic zones, central area and suburbs, but it is economically detailed. It includes production, labor, residential and commercial real estate markets and multimodal road congestion with private car, minibus and taxi vehicles and public buses. BRT reduces road congestion by 9 %, improves traffic speed by 24 %, and reduces the road congestion externality by 18 %. The BRT improves consumer utility and achieves social welfare gains that are 7.9 % of income: two-thirds are from the BRT as a new mode and one-third from the benefits of the lower congestion for the other modes. Road expansion also improves consumer utility, but achieves lower social welfare due to its cost and lower effect on real estate prices. The BRT yields an operating surplus, but road expansion requires deficit financing.

Introduction

Demand-side and supply-side policies can be proposed to improve the performance of urban transportation systems. Previous studies examined the effectiveness of demand-side instruments for relieving traffic congestion in large cities.² Congestion tolls, fuel taxes or parking taxes reduce road congestion, but excessively burden lower income consumers who do not sufficiently highly value the time savings from congestion relief. In the long run, supply-side interventions are called for to expand transport infrastructure or to use the existing infrastructure more intensively to accommodate growth and reduce congestion.

Road building directly benefits all consumers who own cars or travel by taxi or minibus and those who rely on buses, but can be expensive. Roads can also induce switching from high-occupancy vehicles of bus and minibus to the lower-occupancy private cars and taxis, causing congestion to rebound somewhat. Advantages of Bus Rapid Transit (BRT) are that it enables more intensive use of existing roads directly benefitting the poor, and indirectly those who use cars by reducing congestion.

There are studies that addressed the benefits of BRT in a partial equilibrium setting. With a numerically solved theoretical model, Kutzbach (2009) showed that a toll on car travel combined with reserving some road capacity for BRT shifts the modal share in favor of BRT, reduces car and BRT travel times and increases consumer surplus. Basso et al. (2019) used a queuing model of congestion to show that BRT can decrease both travel times and operator costs, fares remaining unchanged.³

In studying social benefits, general equilibrium models should be preferred because transport systems can have significant effects in labor and in real estate markets causing wages, rents, real estate prices and the stock of buildings to change, creating secondary effects on travel behavior. Our model assesses two infrastructure expansion policies actually proposed for Beirut, Lebanon: (i) A BRT system running on special lanes, along with an increased number of regular buses; and (ii) an expansion of suburban highway capacity. Because of data limitations, the model treats only two geographic zones (core and suburbs), but is detailed in its economic structure. The mode choice part of the model follows a conventional approach combining private car, public bus, minibus and taxi and adding BRT as a new mode. Novel aspects of the model are that utility maximizing consumers choose not only among modes of travel as in conventional travel models, but also their location of residence and of work as well as their housing size and the number of non-work trips they make to acquire goods. The demands for goods, in turn, drive production. These joint decisions are responsive not only to travel times and monetary costs but also to wages, rents and product prices. Profit maximizing producers respond to wages, rents and prices by combining labor and floor spaces in buildings to produce output for export and local consumption. Real estate developers react to floor space prices and rents by adjusting the building stocks which, in turn, affect rents and consumer location decisions. Other novel features of the model are that the fares of the privately-operated taxi and minibus services are endogenous and competitively determined; and that the surplus or deficit of the public transport authority that operates the buses and the BRT or expands highways are also endogenous. Tsivanidis (2019) provides a spatially granular general equilibrium analysis of Bogota's TransMilenio BRT system, but does not treat road congestion or real estate conversions. It is a reduced form model designed to uncover TransMilenio's effects using econometrics. Our fully structural model evaluates the effects of prospective projects.

A brief preview of our results is as follows: the BRT proposed for Beirut reduces road congestion by 9 % as measured by the flow to road capacity ratio, improving trip-weighted average road traffic speed by 24 %. The congestion externality is reduced by 18 % and gasoline use by 15 %. The social welfare gain of the BRT proposal is 7.9 % of average income; 2.47 % is due to the reduction of congestion for the other modes and 5.43 % due to the benefits of the BRT as a new mode. The proposed road building increases capacity by 4 %, improving trip-weighted average travel speed and reducing the congestion externality by only 3 %, but also induces rebounds in traffic load and in vehicle miles traveled. The BRT project slightly suburbanizes population and slightly centralizes jobs, while the highway expansion slightly suburbanizes both population and jobs. The BRT raises rents and wages in both the center and the suburban area, but commercial rents in the center increase more than residential rents do, and vice versa in the suburbs. The road expansion project also raises rents but less in percentage terms than the BRT does, and suburban rents increase by slightly more than central rents; while central wages increase more than suburban wages do. By raising wages, the BRT also causes a floor space for labor substitution in production inducing an increase in the conversion of floor space from residential to commercial.

The BRT project improves consumer utility and social welfare and creates a surplus for the public transport authority, which would have additional benefits if invested in the economy. In contrast, the road building increases consumer utilities too but reduces overall social welfare (due to its cost and its smaller positive effect on property values) requiring deficit financing. Deficit financing would require borrowing, e.g. from the World Bank, or diverting funds from other sectors of the economy such as health or education.

The paper is organized as follows. Section 2 describes the model's structure focusing on the modeling of road congestion and choice of mode among private car, privately operated taxi and minibus services and publicly operated bus, all operating on the same road capacity. Section 3 presents the general equilibrium of the transport, labor, production, and real estate markets. Section 4 describes how the model was calibrated from the data. Section 5 presents the simulation results and introduces the BRT as a new fifth mode with its own dedicated capacity. The effects of the BRT and of road expansion policies are compared in Section 5. Section 6 concludes. The Appendix presents a summary of the data, including comments on how it was compiled and other technical details.