School of Medicine

Many nudges are representative of manipulation1. Consider the basic conception of nudge put forth by Thaler and Sunstein (2008): actor A nudges actor B when A makes B more likely to perform action p by activating B’s automatic cognitive machinery, while preserving B’s freedom of choice2. Here, A attempts to influence B not through straightforward presentation of facts but through activation of automatic machinery. A purposefully leaves B somewhat in the dark. A’s nudge is manipulative. Other authors (Blumenthal-Barby 2016; Bovens 2009; Halpern 2016; Saghai 2013) have focused their attention on whether nudges–by virtue of being manipulative– pose a threat to autonomy3. Here, I seek to shift the focus to foreseeability–a concept borrowed from Sarah Buss (2005). Following Buss, I take it that foreseeability and not autonomy is the victim of manipulation. Here I will attempt to get at the cases in which nudges pose a threat to foreseeability and also attempt to develop rough criteria for when foreseeability impairment is justified.