Abstract: Under the trend of deep integration of gig economy and digital technology, online labor platforms, which provide services on demand through precise, efficient, and large-scale operation with algorithms as the core means, have emerged, and gradually formed an “Algorithmic Management” (AM) model. The automatic and efficient, data-driven characteristics of AM have significantly improved the operational efficiency of the platform. However, while optimizing efficiency through algorithms, online labor platforms often undermine the autonomy and rights protection of workers, triggering a contradiction between the concept of flexible labor and algorithmic control.
Relevant studies have mainly analyzed the characteristics and impacts of algorithmic management. However, few have discussed how to govern the “paradox” from the government’s perspective. To answer the question of how governments can solve the online labor “paradox” by designing and combining different governance policies, this paper establishes an evolutionary game model between online labor platforms and workers based on two governance scenarios: regulation and incentives. Through model analysis, the governance effects of regulatory and incentive policies on platforms and workers’strategy selection are compared. Additionally, through numerical simulations, we analyze the evolution paths of platforms and workers under different model parameter settings to further indicate the effectiveness of regulatory and incentive policies.
The study finds that government regulation alone cannot solve the “paradox” and inappropriate regulation intensity will make the game evolve toward lower overall welfare. In contrast, a synergistic mechanism combining regulation with incentive policies is more effective which indicates that an appropriate level of incentives is needed to lead the game to evolve to a stable point with higher overall welfare. Furthermore,the simulation results show that a well-combined governance mechanism of both regulation and incentives is significantly better than the effect of only adopting one of the two governance measures. Finally,the simulation results also indicate that the extent of how well online labor platforms can effectively manage the arbitrage behaviors of workers will influence the final state of evolutionary system.
The conclusions of this paper suggest that the government should (1)keep regulatory intensity in an appropriate range to avoid excessive regulatory pressure on platforms, (2)provide differentiated support for platforms that adopt flexible algorithmic management and (3)actively integrate regulatory and incentive governance mechanisms to build a multidimensional and complementary governance matrix to promote the platform economy’s standardized, healthy, and sustainable development and effectively protect the rights and interests of workers in this new employment pattern. Finally, the platform should actively promote management model innovation and technology iteration, explore new management methods, and promote “algorithms for good” while optimizing algorithm efficiency. This paper expands the research in the field of algorithmic management and provides some theoretical support and decision-making basis for the government to carry out governance over online labor platforms.
The next step of this research is, firstly, to broaden the government governance scenarios; secondly, to further refine the government governance behaviors; and thirdly, to deepen the assessment of the effect of government governance by combining quantitative methodologies to carry out more empirical research.

Key words: government governance, online labor platform, algorithmic management

摘要： 在线劳动平台进行算法管理的同时也容易损害劳动者权益,引发灵活用工理念与算法控制的矛盾。为回答政府如何制定治理政策以化解在线劳动的“悖论”,本文基于监管和激励两种政府治理情境,构建了在线劳动平台与劳动者的演化博弈模型,对比了不同政策的治理效果,并通过数值仿真模拟了博弈系统的演化路径。研究发现,政府采取监管与激励并举的协同机制是一种更为有效的治理工具,适当的激励水平能使博弈向整体福利更高的稳定点演化。本文的研究对于政府构建多维交叉、互补相融的治理体系以促进平台经济健康发展具有启示意义。

关键词: 政府治理, 在线劳动, 算法管理