A crowd of precarious micro-workers? Results of a statistical survey on a French crowdsourcing platform

In 2006, Wired reporter Jeff Howe introduced the neologism “crowdsourcing” to describe “the act of taking a job traditionally performed by […] an employee and outsourcing it to an undefined, generally large group of people in the form of an open call”. Different types of practices may fall within this definition ; here, we will focus on a business-driven type of crowdsourcing in which a platform help companies to outsource tasks to a mass of Internet "pieceworkers", as in the case of the well-known Amazon Mechanical Turk (AMT).This is the crowdsourcing business model. Since daunting tasks such as image moderation or Internet searches cannot be entirely delegated to AI, crowdsourcing platforms offer to break them down into micro-tasks which can be accomplish at any time by a crowd of Internet users at a very low cost. As for the crowdworkers, they're supposed to join the platform to make ends meet by to converting their spare time into productive time: instead of playing onscreen Solitaire or surfing the web for free, they are invited to earn a few money by tagging photos.The purpose of our research program is to examine, beyond the crowdsourcing business model, what the “valuation of spare time” really means. For this purpose, we’ve been carrying out a fieldwork research on the French platform Foule Factory (FF). Younger and smaller than the its American counterpart AMT, FF is nevertheless the best way to understand what crowdsourcing really is in today’s France.This communication proposal is thus a follow-up to our previous communication during the last SASE congress in Kyoto. For this new edition, we will focus on the results of a questionnaire administered in the summer of 2018 to n=1006 workers estimated to represent the entire working crowd. The research question we will ask is as follows: are they a crowd of precarious workers?Our previous research has led us to two results. On the one hand, we have shown that crowdsourcing platforms operate almost in the same way as the type 19th century "sweating system" : they subcontract tasks to homeworkers who are paid (poorly) on a piecework basis (Barraud de Lagerie and Sigalo Santos, 2019). On the other hand, we have shown that three types of use of the platform coexist: interstitial use, use on free time and extensive use (Barraud de Lagerie and Sigalo Santos, 2018). But we were not able to determine which worker profile is the most frequent.Here, we will use statistics to inform the exact profile of the workers, and thus to question the socio-demographic characteristics of the crowd that contributes to Foule Factory. Our preliminary results suggest that the workers are much less vulnerable than one might think. This does not mean that the legality and legitimacy of crowdsourcing should not be questioned, but rather that it should perhaps be interpreted more subtly than simply from the perspective of the exploitation of the weakest