Stick or carrot for traffic demand management? Evidence from experimental economics

The debate between stick (penalties) and carrot (rewards) strategies has been widely carried out in various academic realms, including transportation. This paper seeks to compare the instantaneous and longitudinal effectiveness of a traditional penalty-based congestion pricing strategy and an innovative Incentive-Based Traffic Demand Management (IBTDM) strategy using an online interactive laboratory experiment focused on departure time choice. We also investigate the social-economic phenomena of loss aversion and risk aversion. The experimental results show that penalties and rewards can both positively induce sustainable travel behavior. The stronger demand-shifting performance of the penalty-based strategy at first demonstrates that 'loss aversion' does exist in TDM, at least initially. However, this loss aversion weakens as the experiment continues, which verifies the myopic loss aversion theory. Moreover, risky decision-making in the gambling games proves the existence of risk seeking, which violates the 'certainty effect'. To overcome the disadvantage of laboratory experiment that participants need to accomplish the experiment all in once for considerably long time, an agent-based model is calibrated based on real participants' behavior to further verify the long-term effectiveness of penalties and rewards, and to verify the effectiveness of other optimized penalty/reward schemes. The simulated results based on the agent-based model indicate that the theoretical optimal penalty scheme performs better in general in congestion alleviation, while a more reasonable reward scheme that considers psychological factors can achieve the same effectiveness in terms of total system travel time reduction. The findings of this research can be used to guide the design of TDM strategies, and the behavioral data collection approach based on our laboratory experiment is portable for future studies.