Shared mobility systems regularly suffer from an imbalance of vehicle supply within the system, particularly when there is a spike in user demand at certain locations or an increase in number of vehicles due to special events. If such imbalances are not mitigated, many users will be unable to receive service. Recently there has been increasing research interest in the use of reinforcement learning to craft a dynamic incentive mechanism to encourage shared mobility users to slightly alter their travel behavior in exchange for a small monetary incentive. With the intention that these small changes will aid in rebalancing the vehicle supply in the shared mobility system such that overall service level can be increased. In this paper, we present summarized results of our extensive study of the effectiveness of reinforcement learning-based solutions to the rebalancing problem in terms of the resource and user volume/demand on the improvement in average service level per hour and average percentage improvement in service level per hour. In particular, our study focused on the rebalancing problem in a bikeshare system. Our empirical results show that the reinforcement learning-based incentive mechanism across all budget constraints and various state-action space performs well in terms of average percentage improvement in service level per hour (greater than 8%) when the resource is limited.