For approximately fifty percent of the morning peak travellers a change in behavior can be found when they are rewarded for changing their originalbehavior. The level of change depends on a variety of attributes, of which reward size and participation level are addressed in this study. For various reward strategies the network effects are determined using a dynamic macroscopic model. This study is part of the 'SpitsMijden' reward program.The 'SpitsMijden' program consisted of both a pilot and several model studies to research the effects of rewards on peak hour travel. In the partlyGPS based pilot, conducted with 340 participants, a reward could be earned if the participant did not use the car during the morning peak. This reward would vary from three to seven euro and could be earned by changing departure time, traveling mode or not traveling at all. With 340 participants, however, no network effects were created. Therefore a modeling approachis used to model higher levels of participation. The modeling framework includes an elastic demand routine to determine changes in the level of travel demand for participants as well as non-participants as a result of changes in travel costs and time. Succeeding the elastic demand routine an iterative procedure starts to determine equilibrium departure time and routechoice conditions. The real size traffic network that is used is based onthe situation in the western part of the Netherlands and includes the city of The Hague. The results show network effects following from the rewardprogram. The modeled driving behavior is consistent with the patterns shown during the pilot study and various levels of reward and participation were tested. The main effect on driver behavior was on departure time choice. Drivers from both the pilot and the model easily changed their originaldeparture time to a time before or after the defined peak period. As expected, the biggest network effects are found with the highest level of participation and the biggest reward. In this case congestion already starts occurring before the peak period and will continue to the end of the peak period. However, the cost of the rewards paid will be larger than the cost saved by the reduction of congestion. When using lower levels of reward and participation the benefit per driver becomes higher, because the travel situation before the peak period may become congested by higher levels of participation and reward. When these reactions per reward increase and thereward can be lowered to a certain value, feasible solutions will be found. It is clear that driving behavior can be changed by offering monetary rewards. By varying in reward height and participation level a trade off ismade between costs and congestion reduction. From the set of strategies an optimal solution is presented. This paper reflects on rewarding as a policy option and the effects of rewarding on traffic congestion. For the covering abstract see ITRD E145999
Samenvatting