Author(s)
Source
Saint Louis University Law Journal, Vol. 61, pp. 1-16, 2016-2017
Summary
“Invisible labor” includes many aspects of unrecognized work. Data analytics intended to help employers assess the quality of a worker’s efforts might not measure invisible labor effectively.
Policy Relevance
Data analytics could be used to reduce bias, but could also reinforce bias.
Main Points
- “Invisible labor” includes tasks crucial to workers' job performance, but which are overlooked and/or devalued; invisible labor also includes unrecognized labor, such as housework, or instances where workers' efforts are obscured by a technology platform.
- “People analytics” uses data to analyze the traits and skills of employees and to measure the quantity and quality of work; analytics replaces anecdotal evidence or instinct in hiring, promoting, or firing employees.
- People analytics cannot fairly capture merit if some aspects of work are unrecognized; failure to recognize some aspects of work raises employment discrimination law issues.
- Employment discrimination law is intended to give all workers the same opportunities for hiring and advancement, but increases in discrimination claims filed suggest that this goal is difficult to achieve; people analytics could be used to reduce implicit race or gender bias.
- However, people analytics could reinforce bias.
- Workers who live near their place of employment are more likely to be successful, but choosing workers based on zip code would be discriminatory.
- Customer ratings are affected by bias.
- Areas with large hidden aspects of work are often staffed by female workers, who are required to manage others' emotions or required to wear makeup; some job-related social norms would not survive if stated explicitly, but other norms could be made more objectively measurable.