A mixture-amount stated preference study on the mobility budget

The mobility budget is considered a promising new tool in remuneration and transport policy in Belgium, especially due to its potential of shrinking the company car fleet and lowering car use. Because revealed preference data were scarce in the context of the mobility budget, we conducted a stated preference study to examine the potential outcomes of the introduction of the mobility budget. A challenge in our study is that it required the respondents to choose between mixtures of remunerations for different total budget amounts. In other words, the study was a mixture-amount stated preference study, which involved modeling challenges as well as experimental design challenges. In this paper, we therefore introduce advanced mixture-amount regression models in the choice modeling literature and present a generic method to set up mixture-amount stated preference studies to collect suitable data. Our case-study data comes from an online questionnaire administered to employees at 12 large companies in Belgium (n = 817). For our choice data, a second-order polynomial mixture model in combination with a quadratic effect for the amount led to the most suitable utility function. Our results indicate that current company car users prefer additional days off, income or a car. The bicycle, pedelec and public transport options are disregarded by most employees. Based on our results, we call for a critical reflection on the current system of company cars and reimbursements in Belgium.