Online Outsourcing Market is a popular online gig service system that aims to globalize the gig economy by outsourcing clients' tasks to employees around the world (Hong, Pavlou, & Carey, 2017). The online format, with no geographic constraints, greatly improves access to the freelance market. Especially with the spread of the novel coronavirus, most of the world's population worked from home, making people more dependent on the online labor market. But the online system faces some problems. In particular, many freelancers don't get equal work opportunities. A survey found that about 520000 freelancers on Upwork did not get any job, while about 70,000 of the freelancers who completed at least one assignment earned less than $1 (Green, 2018). Although there have been some studies that have explored fairness and employment discrimination in online gig economic platforms. However, given the complexity of online employment decision-making behavior, much work remains to be done. An important topic for further research is to unravel the mechanism of hiring discrimination under employment platformization. Current related research mainly focuses on explaining the hiring bias in the online outsourcing market based on heuristic theory, and there is also a lack of theoretical argumentation. This research opens up a new perspective, using path dependence theory to demonstrate the facial stereotypes in the online outsourcing market environment. We aim to establish the stereotype path dependence in the online outsourcing market based on the formation stage of path dependence and the context characteristics of the online outsourcing market, including two steps. First, construct a path-dependent trigger mechanism based on the contextual characteristics of the online outsourcing market. Second, develop the self-reinforcing cycle of stereotypes by combining the hiring mechanism and characteristics of the online outsourcing environment. This study can provide a broad reference for the literature related to path dependence, stereotypes, and hiring bias in the online outsourcing market, and it can also provide a theoretical basis for eliminating stereotypes and hiring bias in the online outsourcing market. Another important topic that needs further research is the impact of portraits on employment decisions. Online portraits constitute a pervasive and critical signal in digital labor markets in that workers can boost their employability by manipulating select visual cues embedded in these portraits. Consequently, we attempt to unravel how visual cues embedded in workers’ portraits within digital labor markets can collectively influence constituent dimensions of employability. Notably, we advance a non-verbal cues classification model that differentiates among demographic, physical appearance, image quality, and non-verbal behavioral cues as focal determinants affecting one’s employment status, the number of job offers received, and rehiring probability. Employing computer vision and deep learning algorithmic techniques to analyze the online portraits and personal information of 53,950 workers on Upwork.com, we demonstrate that visual cues embedded in profile portraits exert a significant effect on workers’ employability in digital labor markets. A final important subject that needs to be investigated is to mitigate the hiring discrimination. The impression formation continuum model suggests that more attention can reduce or eliminate gen-der stereotypes. This paper considers three task attributes influencing employers' attention in the online gig market: task complexity, task selection ratio, and self-description validity (Profession, personal experience, and sentiment). We collected data from Upwork.com, a popular online gig marketplace, and tested our hypothesis using negative binomial regression. The results show that task difficulty and task selection ratio can effectively mitigate gender discrimination in the online gig economy market.

2024

2025