网约车平台拼车服务研究

Using on-demand ride-sharing platforms has become the norm of short distance transportation in most major cities. Within the ride-sharing online market, a recent trend emerged on several platforms: carpooling services. Several ride-sharing platforms offer an option which allows passengers in the same direction to be matched to the same vehicle and share a ride at a discount price (e.g., uberPOOL, Lyft Line, and Didi Pinche). Carpooling services have widely been commended for its benefits in improving driver wages, saving rider costs, and alleviating traffic burdens. However, rigorous scientific research on the implications of such services for ride-sharing platforms is scarce. ..In the proposed study, we seek to investigate two main questions: (a) In the presence of carpool services, how should a ride-sharing platform design the price and wage schemes to maximize its profit? And (b) What operational, economical, and social implications will the carpooling services have? The answers to these fundamental questions are theoretically interesting and practically relevant. The strategic interactions between the platform, drivers, and riders are quite involved in the presence of carpooling. ..We propose to build a two-sided market model to uncover salient drivers and implications of carpooling services. This research is expected to identify the most important economic forces of the (two-sided) pricing strategies in the presence of carpooling services, and to predict the equilibrium behaviors of the platform, the drivers, and the riders. The intermediary function of the ride-sharing platform to match supply (i.e., drivers) with demand (i.e., riders) plays a crucial role in reaching the equilibrium market outcome. We also propose to characterize the relationship between carpooling and surge pricing, and the optimal carpooling service quality decision, both of which have the potential to guide platforms to implement the carpooling services...We propose to complement our theoretical analysis with computational studies with inputs of real Uber and Lyft data in the San Francisco market. Our analysis will distill the values of carpooling services from economic and societal perspectives. For the platform, we plan to evaluate the strategic relationship between surge pricing and carpooling. For the riders, we propose to examine the impact of carpooling on the variability of the prices they face in the presence of volatile rider arrival streams. For the drivers, we will quantify their surplus changes driven by carpooling services. Providing carpooling services may backfire despite its good intention. It may increase the vehicle utilization substantially, thus discouraging drivers from working for the platform and, as a consequence, hurting the overall social welfare. ..Finally, it is worth noting that, beyond carpooling services for ride-sharing platforms, this study's approach and results can be adapted to devise policy recommendations for other two-sided marketplaces of sharing economy, such as bike sharing (Mobike) and lodging sharing (Airbnb) platforms.

为缓解交通拥堵、增加司机收入、降低乘客出行费用，中外各大网约车平台相继推出拼车服务（如滴滴拼车的日单量超过150万）。虽然网约车平台拼车服务非常普遍，学术界对这种新型的出行方式还缺乏系统的研究。本项目立志于建立运营管理模型并应用最优化理论和博弈论方法研究:1. 网约车平台如何优化拼车服务的乘客定价与司机薪酬？2. 网约车平台的拼车服务会产生那些运营、经济和社会方面的影响？本项目将建立双边市场模型探究网约车平台上司机与乘客的互动并基于此优化网约车平台的双边定价策略，实现需求（乘客）与供给（司机）的最优匹配。除了理论研究，本项目还将利用美国旧金山市Uber和Lyft的真实数据量化拼车服务对价格波动的影响，以及拼车服务与动态提价两策略的关系。通过本项目研究，我们希望对网约车平台的拼车服务有更全面深刻的理解，并为其拼车服务的运营策略提供理论指导。

为缓解交通拥堵、增加司机收入、降低乘客出行费用，中外各大网约车平台相继推出拼车服务（如滴滴拼车的日单超过150万）。虽然网约车平台拼车服务非常普遍，学术界对这种出行方式缺乏系统研究。本项目旨在建立运营管理模型并应用最优化理论和博弈论方法研究网约车平台对拼车服务的乘客定价与司机薪酬策略以及市场各方（平台、乘客和司机）福利的影响。本项目主要研究成果总结如下。我们首先研究了网约车平台拼车服务的定价问题。具体来讲，我们搭建理论模型探究网约车平台拼车服务的运作管理问题，并理解其社会与经济意义。我们从理论上证明，在拼车质量和效率高的前提下，网约车平台需要提供拼车服务。同时我们也发现拼车服务使得网约车平台获得更大的市场份额，乘客可以享用更便宜的服务。另一方面，网约车司机可能会利益受损，因为平台只需要更少的司机就能满足乘客需求。接下来我们探索包括网约车平台在内的双边平台的竞争(competition)与竞合(coopetition)对市场的影响。我们考虑多个网约车平台企业在供给端（司机）和消费端（乘客）同时进行价格竞争。我们建议一个全新的内生性(endogenous) MNL (Multi-Normial Logit)模型来刻画司机与乘客的行为。于是，平台用户行为既取决于定价，也取决于其他司机与乘客的选择。我们从理论上证明网约车平台的竞争市场存在唯一的纳什均衡。我们对这一结果的证明使用了与一般MNL模型证明不同的方法。基于均衡存在性的结果，我们进一步研究网约车平台间竞合策略的影响。我们研究的背景是不同网约车平台（Curb和Via）共同推出的一款出租车拼车服务。不同网约车平台采用收入共享(revenue sharing)合同来分配来自共同服务的收入。我们证明，一个合理的收入共享合同（如纳什议价, Nash Bargaining）可以实现四赢：两个网约车平台、乘客和司机均从平台的竞合策略中获益。