The Main Determinants of the Demand For Public Transport: An International Econometric Comparative Analysis

The Main Determinants of the Demand For Public Transport: An International Econometric Comparative Analysis


S Masson, O.M.I., Universit?e Reims Champagne - Ardenne; I Joly, L.E.T.- Universit?yon 2; R Petiot, G.E.R.E.M - Universit?e Perpignan, FR



The aim of this paper is to set up an inventory of the key features favouring the urban
public transport (PT) use. The analysis rests on the UITP (Union International of
Transport Public) database ? ?The millennium cities database? ? which is based on a 100
city urban transport systems data collection. It contains 175 variables which concern
demography, urban structure, economy, car ownership, road and public transport
networks, parking facilities, mobility, transport systems efficiency and their impacts on
the environment.
First, we produce a couple of profiles with respect to urban structure, economic level,
urban transport system and mobility. The first group is composed of the West European
and Asian agglomerations which describe an intensive profile of urban transport system:
i.e. high urban density, high PT supply, moderate road supply and important PT modal
share. The second group gathers the North American and Oceanic agglomerations. It
characterises an extensive profile of urban transport system: i.e. low urban density, high
road supply and high car modal share. These agglomerations are characterised by wide
space and time consumption. Eventually, the agglomerations of Emergent countries are
too heterogeneous to compose a relevant profile.
Second, we assess the role of the main variables on the PT use (i.e. urban density,
population, urban area, densities, urban GDP, transport supply, parking facilities,
transport costs, transport investments). For each relation, we examine the position of
both intensive and extensive profiles.
However, this one-dimensional analysis does not traduce the complexity of the modal
choice. It suggests then to explore some relations with multiple explicative variables. So
we propose an econometric analysis of the PT modal share which leads to think about
the handling of the tools favouring PT use. We perform several econometric models
applied to the extensive or intensive profiles. The first model retains the following
explicative variables: part of the urban gross domestic product invested in PT, price of
fuel, user costs ratio of public and private transport, speeds ratio of public and private
transport, number of parking places in the CBD. The estimation produces significant
results. Furthermore, the regression shows, on the one hand, a positive effect of the part
of urban GDP invested in public transport, the price of fuel and the speeds ratio, on the
other hand, a negative effect of the user costs ratio and the parking places in the centre
on PT modal share. In the second model, we test the discriminating power of profiles.
Finally, we propose a model for each agglomeration profile.
To conclude, this analysis highlights the conditions in favour of the urban PT use.
Furthermore, it assesses the weight of the tools which could influence the modal choice
of commuters (relative prices and speeds of transport modes, fuel price, investment
policies and parking policies).


Association for European Transport