| The development of new-generation information technologies such as big data and artificial intelligence has promoted the rise of the gig economy in China,leading to the emergence of online labor platforms,like Alibaba V task,Didi,and Meituan food delivery,which attracts a large number of freelance workers,such as online celebrities,Didi drivers,and food delivery workers.The online labor platform needs to match the appropriate work tasks for freelancers and effectively control their work behavior.The big data-empowered algorithm technology enables the online labor platform to perform task matching function and behavior control function in a highly automated and data-driven manner,resulting in the algorithmic management practice composed of algorithmic matching and algorithmic control.Algorithmic management can realize intelligent supply-demand matching for freelancers on the online labor platform,but it also completely changes the information display structure,competitive environment,and task-matching mode of freelancers,making the supply-demand matching process of freelancers full of uncertainty.First,because the information displayed by freelancers on the platform comes from algorithmic control systems,including big personal data monitoring systems,platform reputation systems,and employer evaluation systems,the information display structure presents new characteristics of multi-dimensional and dynamic changes.The multi-dimensional information controlled by these algorithms may have a complementary effect or substitution c effect.As such,when dealing with information uncertainty,freelancers no longer rely on one party alone,and it also brings uncertainty to the information matching process of freelancers.Second,the transparency of the online labor platform makes a highly competitive working environment for freelancers.The competitive environment has further become a situational factor in the multi-dimensional information solving process of information asymmetry,affecting the supply-demand matching of multi-dimensional information and resulting in a more uncertain effect of multi-dimensional information.Third,with the support of an algorithm recommendation system,the online labor platform provides freelancers with a new task matching mode of on-demand matching between personal capability and work tasks.However,this algorithmic matching is easy to make freelancers fall into biased recommendations and face more information pressure.And freelancers with experience in different fields hold different attitudes towards algorithm recommendation,which makes the effect and efficiency of algorithmic matching on freelancer task matching very uncertain.Therefore,how algorithmic management affects the supply and demand matching process of freelancers on the online labor platform has become an urgent problem to be solved.Algorithmic management of online labor platforms is a cross-discipline and emerging research topic.The literature review reveals that(1)prior research is still in the initial stage of concept definition and theoretical analysis,and few studies have explored the supply-demand matching of freelancers on online labor platforms from the perspective of algorithmic management.Therefore,the role of algorithmic management in the supply-demand matching process of freelance workers is not clear;(2)Previously,scholars have mainly investigated the impact of a single-dimensional and static employer reputation signal on supply-demand matching in the online labor market scenario.The research findings cannot explain the characteristics of multi-dimensional information structure formed by algorithmic control and its effect;(3)Due to ignoring the changes of competitive scenarios caused by transparent platform information,the existing research cannot explain how different competitive scenarios affect the role of multi-dimensional information;(4)There are limited empirical studies on the impact of algorithmic matching on supply-demand matching,so the impact mechanism of algorithmic matching on task matching of different freelancers’ capability has remained unclear.In view of the above research gaps,according to the algorithmic management research and signal theory of online labor platform,this dissertation proposes that the supply-demand matching process of freelancers based on algorithmic management is composed of two sub-processes: information matching and task matching,based on which develops the research framework and design research studies to explore the influence mechanism of algorithmic management on the supply-demand matching process of freelancers on online labor platform.First,this dissertation proposes that the process of information matching is that freelancers use algorithms to control multi-dimensional information to solve the problem of information asymmetry,and task matching is that freelancers use algorithms to solve the problem of mismatch between personal capability and work tasks.Based on the characteristics of big data,these two sub-processes form the supporting logic of "big data-algorithmic management –supply-demand matching process," which enables the bottom-up analysis on how the algorithmic management composed of algorithmic control and algorithmic matching help freelancers on the online labor platform realize the supply and demand matching.Then,this dissertation further proposes the research framework and designs three different research studies to jointly reveal the mechanism of algorithmic management on the supply-demand matching of freelancers.Among them,study 1 and study 2 focus on how the algorithmic controls multi-dimensional information in the process of information matching to solve the problem of information asymmetry,and study 3focuses on how algorithmic matching solves the problem of the mismatch between freelancer capabilities and the needs of tasks in the process of task matching.Specifically,study 1 focuses on how the multi-dimensional information controlled by the algorithm affects the information matching of freelancers and solves the uncertainty of information matching caused by the change of information structure.The embedding of three algorithmic control systems: big data monitoring system,platform reputation system,and employer evaluation system makes the information displayed by freelancers present a multi-dimensional,real-time,and dynamic information structure,which jointly affects the information matching process of freelancers.First,based on the algorithmic control literature,this study identifies the multi-dimensional information characteristics generated by the control mechanisms of different algorithms.Among them,personal skills are the objective record formed by the big data monitoring system,platform score is the digital reputation formed by the platform reputation system,and employer rating is the evaluation formed by the traditional reputation system.Second,according to the signal theory,the multi-dimensional information controlled by the algorithm is divided into the personal skill and platform score,representing the internal quality signal,and the employer rating representing the external quality signal differentiated by the source of signals.Furthermore,this study develops a research model and hypothesis and conducts an empirical study on the impact of the multi-dimensional information controlled by the algorithm on the information matching of freelancers.Study 2 further focuses on the moderation effect of competition intensity on the influence process of multi-dimensional information on information matching and examines the boundary conditions of the action mechanism of algorithmic control so as to solve the uncertainty of the information matching process caused by fierce competition.In this section,the competitive environment of freelancers is regarded as the signal transmission environment of multi-dimensional information dissemination,and the intensity of competition is used to characterize the intensity of the competitive environment.When freelancers face a highly competitive environment,signals from different sources may be affected by the signal transmission environment,resulting in different effects of algorithmic control multi-dimensional information.Based on this logic,this section proposes the research models and hypotheses and conducts empirical research on the moderating role of competitive intensity in the process of information matching.Study 3 is the research on the influence mechanism of algorithmic matching on the task matching process of freelancers.The study focuses on how the algorithmic matching system affects the matching effect and efficiency of freelancers with experience in different fields,solving the uncertainty caused by the change of capability-task matching mode of online labor platform.Based on the introduction of the algorithm recommendation system into the online labor platform,a quasi-natural experiment,this study compares the changes of contract number and response time of freelancers in the control group and the experimental group before and after the introduction of the algorithm recommendation system.The research model is further constructed by taking the experience of freelancers as the moderation variable in affecting the effect of algorithmic matching so as to confirm the boundary conditions of algorithmic matching.To empirically testify the above three studies,this dissertation selects the Alibaba V task platform as the research object,collects a total of 303635 data of freelancers from October 2019 to December 2020,and carries out a quasi-natural experiment combined with the algorithmic matching system launched by the platform.In study 1 and study 2,the panel data are constructed by using the relevant data of freelancers crawled from the Ali V task platform,and the research model is empirically tested by generalized estimation equation and moderation effect test.Based on the quasi-natural experiment,Study 3 empirically analyzed the research model by combining propensity score matching and the differences-in-differences method.The main conclusions are as follows:First,the multi-dimensional information by algorithmic control has a differential impact on freelancers’ information matching,and there is a complementary complex interaction relationship between them,presenting the interaction mechanism of algorithmic control multi-dimensional information in the process of information matching.Specifically,personal skills and platform scores positively facilitate freelancers’ information matching,while employer rating plays a negative role,which confirms the differential impact of algorithmic control multi-dimensional information on freelancers’ information matching.More interestingly,in the process of information matching,platform score and personal skills enhance each other,and both weaken the adverse impact of employer rating on employment decision-making,thus revealing the complementary effect between different algorithms controlling multi-dimensional information.Second,competition intensity,as a moderating variable,affects the relationship between multi-dimensional information and freelancers’ information matching,presenting the moderation mechanism of competition intensity in the process of information matching.Specifically,competitive intensity significantly and positively moderates the impact of platform score and personal skills on the number of contracts but has no moderation effect on the rating of employers,showing the moderating effect of freelancer competitive intensity in the process of information matching generated by multi-dimensional signals controlled by the algorithm.The research conclusion shows that in the open and transparent competitive environment of the platform,the algorithm enabled control system is embedded into the working process of freelancers,which obtains and shows more real and valuable signals,thus helping employers evaluate the quality of freelancers more objectively and effectively in the fiercely competitive environment.Third,algorithmic matching positively facilitates the effect and efficiency of freelancers’ task matching,and there is a moderating role of freelance workers’ domain experience in the process of task matching so as to present the adaptation mechanism between algorithmic matching and freelancers’ experience in the process of task matching.Specifically,algorithmic matching significantly increases the number of contracts and significantly reduces the response time,indicating that algorithmic matching helps to improve the effect and efficiency of task supply-demand matching.Meanwhile,freelancers’ field experience positively significantly moderates the relationship between algorithmic matching and the number of contracts.With the gradual enhancement of field experience,the higher the field experience level of freelancers,the better the task matching effect produced by algorithmic matching.In addition,in terms of matching efficiency,freelancers with high-level field experience,such as head anchors and low-level field experience,such as tail anchors,have no significant response time,while most freelancers with medium-level field experience,such as waist anchors,have reduced the response time due to the introduction of algorithmic matching.Therefore,algorithmic matching effectively supports the medium-level users of the online labor platform in terms of efficiency improvement,reflecting the adaptation mechanism between algorithmic matching and the experience of freelancers.The contributions of this research are mainly reflected in the following four aspects:First,regarding the changes in the information structure,competitive environment,and task matching mode of freelancers caused by algorithmic management,this dissertation proposes that the supply and demand matching process of freelancers supported by algorithmic management includes two sub-processes: information matching and task matching.The former is a sub-process formed by the multi-dimensional information structure and highly competitive working environment generated by the algorithmic control system;the latter is formed by the algorithmic matching system supporting the new pattern of individual capability and task matching.The existing algorithmic management research of online labor platforms is still in the initial stage of theoretical analysis and case study.Besides,the internal process of algorithmic management to help freelancers achieve supply-demand matching has not been explored.Starting from the characteristics of big data,this dissertation constructs the logical main line of the "big data-algorithm system-supply and demand matching process." This opens the black box of algorithmic management to realize the supply and demand matching of freelance workers,identifies and defines the information matching process supported by algorithmic control and the task matching process supported by algorithmic matching,and effectively explains the different action modes of algorithmic control and algorithmic matching,as well as the characteristics of information matching and task matching,laying a theoretical foundation for the follow-up research on the supply and demand matching of freelancers.Second,this study proposes and identifies the characteristics of multi-dimensional and dynamic information structure of algorithmic control systems and empirically reveals the differential impact of multi-dimensional information on freelancers’ information matching and the complex interaction between them.Existing studies mainly focus on how the evaluation system affects the supply-demand matching of the online labor market and can not fully characterize the multi-dimensional dynamic structure of freelance worker information display formed by multiple algorithmic control systems.Based on the signal theory,this dissertation proposes a research model composed of three-dimensional information controlled by the algorithm(personal skills,platform score,and employer rating),finds the direct action and interaction mechanism of multi-dimensional signals controlled by online labor platform algorithm,and comprehensively investigates the complex relationship of multi-dimensional,real-time,and dynamic information display structure on freelance workers’ information matching.It provides a new research angle for studying the information asymmetry of freelancers on online labor platforms.Third,it is found and proved that the competitive environment of freelancers as a moderation variable affects the relationship between the algorithmic control multi-dimensional information and the information matching process.Previous studies have focused on the impact of contract types and workers’ nationality in the supply-demand matching process of the online labor market on workers’ reputation signal effect from the perspective of employers’ decision-making.There is little research on the competitive environment of freelancers as situational factors,which is difficult to explain the information matching characteristics of online labor platforms under different competitive situations.This dissertation finds that the competition intensity of different freelancers has different moderating effects on the multi-dimensional signal controlled by the algorithm,which provides a new perspective for the study of the boundary conditions of multi-dimensional signal controlled by the algorithm of online labor platform.Fourth,in terms of the new pattern of task matching,this dissertation reveals the adaptation mechanism between algorithmic matching and freelance workers’ experience and confirms the function and boundary conditions of algorithmic matching.Based on a quasi-natural experiment,this study finds that the introduction of an algorithmic matching system can effectively promote the effect and efficiency of task matching,proves the differentiated matching efficiency of algorithmic matching for freelancers with different capabilities,and finds that the improvement of online labor platform mainly plays a significant role in improving the efficiency of freelance workers with medium-level experience.This study found and confirmed the facilitating role of algorithmic matching for freelancers to realize the matching of personal capability and work task,as well as the characteristics of differentiated matching,so as to reveal the complex regulation of algorithmic matching to form task matching and provide guidance for clarifying the action mechanism of algorithmic matching for freelancers on the online labor platform. |
PDF Full Text Request