Measuring the Technical Efficiency of the Tunisian Collective Passenger Transport Companies Using Two-Stage DEA Model

The aim of this study is to assess the level of technical efficiency and scale of Tunisian VCT companies, to determine the endogenous and exogenous factors that may explain their inefficiency and to analyze their economic implications on the performance of the sector. Our approach is based on a two-stage DEA model. In the first stage, we used the classical DEA model in its two versions CRS and VRS under input orientation, to calculate the technical efficiency and scale score of each company throughout the study period. In the second step, these efficiency scores will be regressed on non-discretionary variables using a Tobit model. These models were applied to a sample of 117 observations consisting of panel data on the 13 public passenger transit companies covering the single urban network, over a period of 9 years from 2009 to 2017. The results show considerable levels of total technical inefficiency in several companies. Only three firms are considered efficient. They reveal opportunities for improvement by decreasing the number of staff, the average age of buses, the occupancy rate and by adjusting the size and configuration of the network. The pure technical efficiency scores are higher and consistent. It indicates that the TCV industry has an increasing returns to scale, as well the existence of a mismatch problem between the company size and the managed traffic. Hence the need to merge small-sized business to achieve minimum efficiency. Furthermore, this inefficiency can be explained by other exogenous factors that characterize the technical, demographic and economic environment in which the companies operate. This efficiency is all the higher as the population served is dense, the road network is well maintained, the motorization rate is low, the lines are of short or medium distances and the share of students is low.