Killing and Letting Die

In my Bioethics class, the professor has begun to present us with conundrums where he elicits a universal moral reaction to the difference between killing someone and letting them die. Protestations of some moral philosophers to the contrary, there does seem to be a strong intuitive difference--and where there's intuition, there's frequently truth. Here are some thoughts on why there's a distinction, or at least why intelligent organisms should believe that there is.

- Evolution. There's a degree of personal cost associated with killing people; there isn't with letting them die. All outcomes being equal, therefore, a self-serving organism should perceive the two situations differently, regardless of morality.

- Economic Efficiency. When we kill someone, we pay only a small fraction of the costs of the death; the victim pays the rest. Because humans overweigh costs and benefits to themselves from an omnicient social planner's perspective, and underweigh those of others, we would expect humans to routinely overestimate the value of killing someone; with an unwilling victim, these costs can never be negotiated and internalized. Saving people from death, on the other hand, can be rewarded with compensation; the costs and benefits of the process are entirely internalized between the two parties involved, and the actor can receive some sort of later reward from the person saved (hence the universal concept of owing your savior some sort of moral debt). We should therefore be far more cautious in the situation of killing than in the situation of allowing someone to die: because the economic actors involved cannot come to a mutually-beneficial agreement taking into account all costs and benefits, they cannot make an informed or efficient decision.

- Negative Rights. If we accept that a state of nature cannot be just or unjust, it is entirely moral that any person remain in the state that they find themselves in through random chance. Under such assumptions, an evil act would be one which reduces a person's choices beyond how they would have been in a natural state. Killing someone reduces their options; failing to save them doesn't. Therefore, under a system of negative rights, killing someone is far worse than failing to save them.

- Probabilistic Harm. Assume that any person has some rate of error in their decision processes governing whether or not to kill or save another. If we believe that the number of instances where killing someone is justified is very small (a reasonable assumption!), then the number of cases where people are killed in error will vastly exceed those where they were killed righteously. If the harms of wrongfully killing several people outweigh the benefits of killing one, then we should be very cautious about killing. The same argument applies to allowing people to die, to some extent; however, two differences exist. First of all, far fewer people find themselves in a position where they may be saved than killed, reducing the possible level of harm that you can produce by deciding to consider in each case whether or not to kill them. Second, the costs of attempting to save someone are frequently very high; we can therefore safely assume a higher rate of correctness in guessing that they should die than that a random person should be killed. To maximize the probabilistic benefits, therefore, we should have far higher standards for deciding whether to kill than whether to save.

- Hubris. Perhaps humans are always insufficiently wise to decide whether or not another person should live or die. If this is true, we should make these decisions only when absolutely necessary--when nature has already forced us to choose whether someone should live or die, rather than at our own whim. Thus, the forced choice of whether or not to save someone is acceptable, while the optional choice of whether or not to kill them is not (regardless of whether the person ultimately lives or dies!)

- Cheating. Opportunities to save people tend to be random; opportunities to kill them are ubiquitous. We therefore trust people to be more objective in the case of the former than the latter. We would therefore frown upon killing to disincentivize cheaters, but assume that the motives of someone who failed to save another person were more likely to be pure.


Additionally: on the note of medical euthanasia, I would love to know whether or not its serious academic revival in the early 19th century coincided with a dramatic increase in health care costs, but I can't find the necessary health care data to back up or refute this claim. More to follow as research allows.