I've finally got around to reading Dan Ariely's The (Honest) Truth About Dishonesty. The research presented in that book shows that one factor that makes people more likely to cheat is if they're a step or more removed from doing the actual cheating.
For example, there was a study done where the participants were asked how likely they thought that the average player, and themselves, would move a golf ball 4 inches to a more advantageous position. The three methods of moving the ball that each participant was presented with were: their golf club, their foot, or directly picking up the ball. The results suggested players using the golf club to the move ball would be more likely to cheat, with one's foot or hand coming second and third respectively. The hypothesized reason for this was that, in using the golf club, the player was not directly in contact with ball and thus somewhat removed from the act of cheating. This allowed them to more easily rationalize their actions.
Could the same sort of mentality be at play here? Throwing a switch is a less direct way of causing the death of a person vs. physically throwing someone off a bridge. In that way, it allows people to more easy rationalize the action, even if the result is the same.
What people should strive for, in Greene’s estimation, is moral consistency that doesn’t flop around based on particulars that shouldn’t determine whether people live or die.
Easy to say, but much harder in practice when people need to overcome internal biases that they might not even be aware of.