The Future of HRM Incentivizing Strathern’s Paradox? Workers’ Responses to Algorithmic Human Resource Management

The future of human resource management (HRM) has often been described as one of algorithmic HRM—that is, automating human resources (HR) decisions by using computational algorithms. Algorithmic HRM promises more objective HR decisions and less administrative workload for HR managers. Although incentivizing is a core HRM task, most of the research on algorithmic HRM takes place under the assumption that algorithms do not influence the behavior they are designed to predict. On the contrary, any behavioral changes activated by them threaten their validity. In a mixed-methods study using a classroom simulation of an algorithmic personnel evaluation with undergraduate students, we found that this premise does not hold. Rather, we discovered that (a) algorithmic HRM incentivized workers to behave contrary to their own preferences, (b) workers did attribute their discomfort with this outcome to the algorithmic personnel evaluation ignoring behavior, and (c) workers’ attitudes toward the algorithm were not related to their behavior. The observed phenomena can be explained with Strathern’s paradox, which states that measures underlying an algorithm cease to be valid when used for HRM decision-making.