When Killing is Worse than Letting Die

Consider two plausible ethical claims: (1) It's much worse to kill someone for money than it is to refrain from saving a life due to the monetary cost. (2) There's no intrinsic or fundamental significance to the distinctions between doing and allowing, killing vs letting die, etc. (It's notoriously difficult to give a sound metaphysical account of these distinctions that seems to be getting at anything fundamentally important, after all.) Together, these points suggest that we should want to account for such distinctions having a kind of indirect significance -- say, by typically correlating with something else that has intrinsic significance.

What might that "something else" be?  One striking thing about the ill-doer (in contrast to the ill-allower) is that their presence makes things worse than if they hadn't been there at all.  I don't think that necessarily makes any difference from the agent's perspective (they have just as strong a reason to prevent bad outcomes as they do to refrain from causing bad outcomes; there's no reason for them to privilege the status quo in their deliberations).  But it may make a difference to how others should regard the agent.  It could make sense for others to fear the ill-doer, for example, whereas there's typically no reason to fear an ill-allower who makes no difference to the situation. (Exceptions arise in "pre-emption" cases where a malicious ill-allower is ready & willing to do harm, but found it wasn't necessary because their victim was already in a pickle even without their intervention.  This suggests that the more fundamental feature of concern here is the disposition to do harm.)

While that's a difference of a sort, it doesn't make a difference to the moral status of the act (/omission) of the sort we typically expect here.  So I don't think we have the full answer yet.  But it may at least point us in the right direction.  In particular, looking first to others' reactions strikes me as the most promising way to find something significant in (or around) the doing/allowing distinction.  And insofar as we want reactions that make a difference to the moral status of the act itself, this suggests that we should be looking to the reactive attitudes, as there is a sense of obligation that is tied to considerations of blameworthiness.

Recall my old argument (published in 'Virtue and Salience') that salience facts + quality of will explains why agents are typically more blameworthy for letting a child drown in Singer's pond case than they are for letting child die from lack of charitable aid.  Singer's pond case is a rare instance of 'letting die' where the neglected interests are especially salient to the agent, and hence should be expected to activate any minimal concern for others that they might have.  Failure to help in this case thus reveals a striking degree of callousness or lack of altruistic concern.  But now note that killing typically involves making the victim even more salient.  That is, while people die overseas all the time without our noticing it, it tends to be more difficult to fail to notice when you yourself directly kill someone (or so we might expect)!  So again, we should expect that any minimal concern for others on the part of the agent should be fully activated in the circumstances. This means that if they go ahead and kill the victim (for non-moral reasons such as personal gain), this reveals a very striking lack of altruistic concern -- a "quality of will" far below the threshold for escaping blame and resentment.

This explains why typical "killings" are more likely to be impermissible (contrary to duty or obligation) than typical instances of "letting die".  In particular, it explains why failing to save a life through charitable donations is a more acceptable (less blameworthy) moral failure, whereas killing someone for the sake of the same amount of money is clearly unacceptable (blameworthy), even if we accept the consequentialist thesis that we have equally strong moral reasons to prioritize lives over money in either case.

Some things I like about this account: (1) It explains our intuitions in central doing/allowing cases without bestowing excessive significance to the distinction; (2) it's compatible with act consequentialism (and so doesn't, e.g., prohibit killing for good moral reasons such as saving more lives); and (3) it accurately captures "exceptions" to rule, e.g. the badness of letting-die in Singer's pond case.

Questions for the reader: (I) Any objections?  (II) Are you aware of similar arguments that have been made before? 

(Others have appealed to salience/vividness to explain away our intuitions about trolley cases, but that's a slightly different project as I am here using salience/vividness facts to vindicate certain of our intuitions rather than undermine them.  Note that I'm not here addressing cases of utilitarian sacrifice. There are important limitations to the scope of the doing/allowing distinction that emerges from my picture here.  It's largely limited to acts done for non-moral reasons, and does not vindicate the anti-consequentialist idea that there are stronger moral reasons against killing than against letting die.)

UPDATE: see related discussion at PEA Soup.