Alternative Policies to Reduce Urban Public Transport Deficits in Spain

Alternative Policies to Reduce Urban Public Transport Deficits in Spain


Javier Asensio, Universitat Autonoma De Barcelona, Anna Matas, Universitat Autonoma De Barcelona


We estimate cost and demand functions of urban bus operators in Spain and simulate the impact of alternative deficit-reducing policies: changes in the fare structure, supplied capacity, network length and technical efficiency improvements.


As in many other countries, in Spain fares’ revenues hardly cover half the operational costs of urban public transport. Independently of whether the public transport service is directly provided by the municipality or indirectly by a private operator, the gap between revenues and costs is financed by local authorities. As the local authorities budget turned into deficit in 2008 the burden of subsidising public transport became more relevant.
The aim of the paper is to empirically analyse the effects of alternative policies aimed at reducing the operational deficits of urban bus operators in Spain. In order to do so, the paper exploits a new database at firm level that covers most medium and large operators in Spain during between 2006 and 2009 (the sample is currently being expanded to 2011). Available data includes detailed operational costs, ridership by ticket type, fares and network size at the firm and year level for a variety of Spanish cities ranging from relatively small towns to large urban areas. Firm level data is complemented with socioeconomic and territorial information at municipal level, compiled from different sources. All this results in an unbalanced panel of approximately 100 firms for the 4 sampled years.
In order to simulate the impact on operators’ deficits of alternative policies, both the cost and demand equations need to be estimated. With the help of the estimation results, different policy changes are simulated and their effects on operators’ deficits are compared.
The cost function is estimated as a stochastic frontier thus allowing to measure the relative efficiency of each operator. The adopted methodology follows that proposed by Gagnepain and Ivalid (Stochastic frontiers and asymmetric information models, Journal of Productivity Analysis, 18, 145-159, 2002) in order to take into account the different regulatory regimes under which the firms operate. This analysis directly leads to the definition of the first policy impact, which consists in extending the most efficient regulatory setting to all firms. The estimated efficiency gains imply cost reductions that would not have an impact on ridership, so that all cost savings would translated themselves into deficit reductions.
On the consumers’ side it is possible to estimate one specific demand function for each type of transport title, since the database contains information on prices and ridership of both single tickets and multiride cards. Thus, the estimation of the demand for urban bus services results in the estimation of cross-price elasticities between alternative tickets, as well as the usual own-price and income elasticities. In this way, it is possible to simulate the impact of policy changes that modify the fares’ level and/or its price structure.
Other policies that may result in reducing public transport deficits imply modifications of network extension and/or changes in the number of seats provided while maintaining the network length. The impact that these measures would have on operational costs can be computed, but they would also have impacts on ridership and revenues, which the demand function estimates allow us to measure.
At the current stage we can provide provisional results of the simulated policies. The first one implies a 10% increase in all fares, resulting in a 5.3% increase in revenues that lead to a 3.9% deficit reduction. The second policy reduces supply by 10% and deficit by 6.7%. If that policy is combined with a 10% decrease in the network size, the deficit falls by 8.3%. The impact on costs of a change in the regulatory setting has not been yet estimated, but provisional results with a standard stochastic cost function show that if all operators could reduced their technical inefficiency by half, the result would be a 9.2% cost reduction, implying a 16% reduction of the deficit in a representative firm.


Association for European Transport