Abstract

Amazon Mechanical Turk: Methodological Innovation in an Evolving Labor Market

The Mechanical Turk (MTurk) platform by Amazon is a crowdsourcing site that aspires to provide anyone on the internet with access to “a global, on-demand, 24x7 workforce,” primarily fulfilling tasks such as content moderation and survey participation. Made available to the public in 2005, this platform was an early leader in what has grown to be a multi-billion dollar online outsourcing industry, identified by overlapping designations of “crowdsourced,” “piecework,” “micro,” and “gig” labor. Joining in this online labor trend in droves, academic researchers—particularly those in the social sciences—have been making heavy use of MTurk for collecting survey data, publishing tens of thousands of studies each year using data crowdsourced from the platform. In this presentation, I examine MTurk through both the history of labor markets and the history of social science research participation. MTurk is significant in its popularity and prominence in market shares and is emblematic of the precarious, extremely-low-waged contemporary “digital sweatshop,” which effectively strips workers of any protections, aside from what they build for themselves through information sharing. As I argue, the methodological innovation in social scientists’ discovery and use of MTurk leads to attendant dilemmas in research design, as both researchers and participants attempt to out-maneuver each other on the site for financial gain, knowledge production, and self-protection. Taken together, the historically-situated work and research conflicts, animated by this technological platform, call into question both the ethics of this wave of MTurk-based academic research and the perennial labor shortage issues among social science research participants.