Value of time, schedule delay and reliability - estimates based on choice behaviour of Dutch commuters facing congestion

This paper presents the results of a large stated choice experiment among Dutch commuters facing congestion. The experiment consisted of a fractional factorial design with 15 different attributes, three alternatives were car specific and the other was always public transport. Various model specifications have been estimated on the collected choice data allowing us to analyse choice behaviour of road users and determine their values of time, schedule delay (both late and early) and reliability (or uncertainty). In this paper we present the estimates of the best-fitting discrete choice models and interpret the results.