Employer transport plan : do they change the commuting behaviour of employees?

In recent years, the interest of companies for mobility has steadily increased through both public policies aiming at involving the companies in the mobility issue and business objectives mobility can achieve. As a result, a growing number of companies have implemented an Employer Transport Plan (ETP). However, the research literature has brought only a limited attention to ETPs. In addition, existing studies on commuting are often based on the analysis of one specific mode of transport. Hence, the aim of this paper is to find out whether ETPs change the commuting behaviour of workers by simultaneously analysing the modal shares of each mode of transport use to commute in Belgium. To achieve this objective data of the Belgian ‘home to work travels' diagnosis is used. This diagnosis is performed ever 3 years by the FPS Mobility and Transport and is mandatory: each company located in Belgium and employing at least 100 workers have to fill in a mobility questionnaire. Two diagnoses (in 2005 and in 2008) are available to date, and are used in this research. The database contains information on the modal split and its evolution at 4459 workplaces, as well as on the ETP which was implemented by those workplaces at both dates. Spatial econometrics techniques are then used in order to explain the modal splits at each workplace. In fact, spatial autocorrelation has been detected in the data and spatial econometrics allow taking into account similar unobserved environmental characteristics. A single equation (SE) is firstly estimated to explain the modal share of each mode of transport. Statistical tests are performed on these SE in order to find out the model specification and the weight matrices to be used. Then, the modal shares are simultaneously estimated thanks to a Seemingly Unrelated Regression (SUR). The results show that the estimated coefficients of the SUR model do not differ from those estimated in the SEs and that the promotion of alternative modes of transport increases their use. The SUR model will also be estimated with an adding-up constraint in order to ensure that the sum of the estimated modal shares do not exceed hundred percents. The results of this model will then be compared with the SE and SUR models.