Does the presence of human workers and mechanisms of throughput legitimacy increase acceptance of algorithmic governance? A survey experiment simulating airport security checks

This study examines the effect that fully automating security checks in airports may have on public perceptions of the legitimacy of algorithmic governance, via a survey experiment. In our previous work (Gritsenko and Wood, 2022), we argued that using algorithms in the governance process creates a shift towards forms of governance that combine ex ante specified principles, the key characteristic of hierarchies, with a low level of commitment from human actors in this process. As a result, algorithmic systems are expected to increase efficiency while decreasing the space for governing actors’ discretion, hence, negatively affecting perceptions of trustworthiness and accountability. Airport security checks - a form of hierarchical governance – is expected to demonstrate this shift from higher to lower levels of commitment. Airport security checks are familiar to many people, placing luggage in boxes and passing through full body scanners. They already include automated elements, and further automation has been advocated by airport operators to improve security and implement air safety regulation. Security checks are therefore a useful site for experimental research on algorithmic governance. We use a survey experiment to expose respondents to fictious scenarios where their experience at airport security makes them late for their flight, in a fully automated context with no human workers present, or a familiar context of semi-automation. We further examine whether the presence of human workers, and associated mechanisms for throughput legitimacy (accountability, transparency and feedback) in the airport security check process, improves respondents’ evaluations of the legitimacy of the process.