The Construction of Utilitarian Moral Behaviour: Evidence from Perspective-Taking Accessibility

For centuries, moral philosophers and decision-making theorists have been interested in
whether people’s preferences conform to utilitarian expectations, where it is acceptable to
sacrifice one person in order to save many others (Bentham, 1789/1970). However, I argue
that despite many moral dilemmas requiring participants to reason about moral perspectives,
they do not contain full perspective-taking (PT) accessibility. For example, in the autonomous
vehicle (AV) dilemma (Bonnefon et al., 2016), participants are asked to make judgements in
response to scenarios offering partial PT accessibility. In particular, in the scenario, participants
are presented with the perspective of a passenger travelling inside an AV that is about to crash
into a group of 10 pedestrians but are not offered the perspective of the pedestrians.
Accordingly, participants’ judgements of moral appropriateness and subsequent purchasing
behaviour (whether they would like to buy a utilitarian or non-utilitarian AV) are biased by
scenarios offering partial PT accessibility. It is little wonder then why Bonnefon et al.’s (2016)
findings reveal that people, despite their utilitarian moral judgements, do not express prosocial
utilitarian purchasing behaviour; they do not want to buy the utilitarian AV that they judged to
be the most moral. Accordingly, in 9 Experiments, I explored the influence of PT accessibility
on participants’ moral preferences. In Experiments 1-4, I demonstrate that when offered full
PT accessibility to AV crash scenarios, participants’ moral utilitarian judgements informed
their utilitarian purchasing and usage behaviours (purchasing value, willingness to buy and ride
AVs). In other words, people wanted to buy and ride the utilitarian AV that they judged to be
moral. Moreover, utilitarian preferences were found to be consistent across type of involvement
(stranger, participant and family member), judgement tasks (judgements of moral
appropriateness, purchasing value, willingness to buy and willingness to ride AVs;
Experiments 3, 4, 8 and 9), preference elicitation methods (judgements vs. choice; Experiments
5 and 6) and psychological processing employed by the participants (immediate, conscious and
unconscious; Experiments 7-9). Thus, the present thesis provides theoretical and empirical
evidence for PT accessibility and its importance in informing people’s moral behaviour.
Moreover, this thesis offers commercial and policymaking insights for promoting public
acceptance of autonomous systems.