本文用出租车轨迹数据和房租数据对京沪两市的加班情况进行了分析。本文创新性地使用出租车轨迹数据对加班程度进行了量化,使用分位数回归等方法对“通勤时长”、“房租”和“公司所在商圈”等影响因素进行了较为深入的研究。本文发现,“通勤时长”对北京市和上海市影响是一致的，针对加班时间较长的人群,通勤时长明显地对加班程度施加了负向影响；而“房租”对北京市和上海市的影响是不同的：北京市的房租显著地抑制了加班程度，并且影响程度先增后减；上海市的房租对加班程度的影响却显著为正，并且影响程度越来越大。在描述性统计分析部分，本文根据加班时长和通勤路线绘制出了加班热力图、居住热力图和通勤线路图，较为直观地展示了京沪两市在不同时间段、不同地域（商圈）加班情况的对比，可为相关政策提供清晰、直观的量化决策依据。

This paper analyzes the overtime of Beijing and Shanghai by taxi trajectory data and rent data. This paper innovatively uses the taxi trajectory data to quantify the degree of overtime, and uses the method of quantile regression to conduct in-depth research on the influencing factors such as “commuting duration”, “rental” and “company's business district”. This paper finds that the “commuting time” has the same impact on Beijing and Shanghai. For those who have longer working hours, the commute duration obviously has a negative impact on overtime; and impact of rent is different for Beijing and Shanghai: Beijing's rent significantly inhibits the level of overtime, and the degree of influence increases first and then decreases; the impact of Shanghai's rent on overtime is significantly positive, and the impact is growing. At the same time, this paper draws the heat maps for overtime and residential and commuter route map, which shows the Beijing and Shanghai cities to work overtime in different time periods and different regions (commercial circles). The comparison of the situation can provide clear and intuitive quantitative decision-making basis for relevant policies.