Pubs & Blogs

Elsewhere Online twitter Facebook SLS Blogs YouTube SLS Channel Linked In SLSNavigator SLS on Flickr

Playing with Trolleys (I): Intuitions About Aggregation


Publication Date: 
August 29, 2012
Working Paper
Bibliography: Mark Kelman, Tamar Admati Kreps, Playing with Trolleys (I): Intuitions About Aggregation, Stanford Public Law Working Papers, No. 2138475 (2012).


We explore when experimental subjects are likely to think aggregation across persons – making a decision that some parties should bear losses so that others might gain more – is permissible and when they think it is impermissible, an issue generally explored by eliciting responses to classic Trolley Problems. One central novel finding we explore in this paper is that the intuition to distinguish the two classic Trolley cases – judging the permissibility of diverting the runaway trolley on to a spur track to save potential victims on a main track rather than pushing a person to block the trolley from hitting the potential victims – is considerably weaker and more fragile than the extensive literature that emphasizes its significance has implied. At core, the intuition to divert is fragile in a variety of ways we detail that the intuition against pushing is not. Thus, in many circumstances that we describe, responses to those “killing” problems will converge to a considerable degree on the anti-aggregating position; moreover, the strictures against pushing are typically seen as mandatory (and holds even when following the strictures harms kin or more identified parties) while the requirement to divert is rarely unqualifiedly mandatory and even judgments that diverting is permissible weaken when kin and more identifiable people would be sacrificed by allowing it. At the same time, however, we also find that intuitions against aggregating in both pushing and diverting cases are far more fragile than most philosophical and psychological observers recognize. The intuition subjects display that proves far most basic and durable is the intuition that aggregation is plainly permissible, often desirable, but rarely completely mandatory when one is contemplating saving more, not fewer, people either through ex post allocation of life-saving resources or through making decisions, ex ante to establish programs that will result in fewer, rather than more, deaths in the future, even in situations in which the identity of those at increased risk of death is known at the time the initial program is proposed.