We analyze what ridehail drivers do when searching for paid fares. We use a dataset of 5.3 million trips in San Francisco and partition each search trip into cruising, repositioning, and parking segments. We find that repositioning accounts for nearly two-thirds (63%) of the time between trips, with cruising and parking accounting for 23% and 14% respectively (these figures exclude short trips). Our regression models suggest that drivers tend to make reasonable choices between repositioning and parking, heading to high-demand locations based on the time of day. However, we also find evidence of racial disparities, supporting previous studies of both taxis and ridehailing that indicate that drivers tend to avoid neighborhoods with high proportions of residents of color.

我们分析了网约车司机在搜索付费票价时的行为。我们使用旧金山 530 万次旅行的数据集，并将每次搜索旅行划分为巡航、重新定位和停车段。我们发现重新定位占出行之间时间的近三分之二（63%），巡航和停车分别占 23% 和 14%（这些数字不包括短途旅行）。我们的回归模型表明，驾驶员倾向于在重新定位和停车之间做出合理的选择，根据一天中的时间前往高需求的位置。然而，我们也发现了种族差异的证据，支持先前对出租车和网约车的研究，这表明司机倾向于避开有色人种比例高的社区。