多维度建模视角下广州外卖收入波动及就业保障研究

Research on Income Volatility and Employment Security of Food Delivery Riders in Guangzhou from a Multi-Dimensional Modeling Perspective

摘要: 在平台经济深度渗透就业市场的背景下，外卖骑手作为典型灵活就业群体，其收入波动与社会保障问题关乎劳动者权益与新业态可持续发展。本文以广州市外卖骑手为研究对象，匹配问卷调研、气象与商圈多源数据，综合运用多元线性回归、随机效应面板、机器学习与PSM-DID等方法，系统分析骑手收入波动的驱动因素、风险特征与保障政策效应。研究表明：订单数量是收入波动的核心驱动因素，天气与商圈密度会显著放大收入不确定性；社会保险能够有效平抑收入波动，且对群体内部收入分化具备收敛作用；保障政策的缓释效应存在情境依赖性，在极端天气等高风险场景下作用更为突出。据此提出优化平台分配机制、扩大制度保障覆盖、完善极端场景风险缓冲等建议，为灵活就业群体权益保障提供决策参考。

Abstract: Against the background of the platform economy, food delivery riders, a major flexible employment group, face severe income volatility and inadequate social security issues that are related to workers’ rights and the sustainable development of new business models. This paper takes food delivery riders in Guangzhou as the research object, and uses multi-source data such as questionnaire surveys, weather and business districts to systematically analyze the driving factors, risk characteristics and protection policy effects of rider income volatility by comprehensively using multiple linear regression, random effects panel, machine learning and PSM-DID. Social insurance can effectively mitigate income volatility and has a convergent effect on income disparities within groups. The mitigating effect of social security policies is context-dependent, with its effectiveness being more pronounced in high-risk scenarios such as extreme weather. Based on this, suggestions are made to optimize platform allocation mechanisms, expand the coverage of institutional guarantees, and improve risk buffers for extreme scenarios, providing decision-making references for protecting the rights and interests of flexible employment groups.