Imagine that one day you come across five people in immediate and mortal peril. You and you alone can save them, but doing so will kill one other person. For instance, five people may be in path of an out-of-control trolley that could be diverted onto a track where there is only one person. You are aware of no morally relevant differences between the six people: none of them are murderers or saints who especially “deserve” to live or die, and none of them ended up in this situation due to their own negligence or transgressions. Given such a situation (hereafter referred to as the “thought experiment”), the resolution is that you should opt to actively kill the one person in order to save the many.
Two terms are useful for understanding this debate. “Utilitarianism” is the ethical view that, broadly speaking, prescribes whatever course of action will result in the greatest good, usually measured by adding up the total amount of good (also known as “utility”, and which in practice usually means happiness or some other measure of a thriving life) that accrues to everyone. According to utilitarianism, the correct act is that which results in the highest total utility, regardless of what actions it is that bring that state about; because of its emphasis on consequences, it is sometimes known as “consequentialism.” According to utilitarianism, we generally ought not to kill, and ought to keep promises, because doing so makes everyone better off. “Deontology”, by contrast, is the ethical view that prescribes a moral agent to act in accordance with specific rules, regardless of their consequences. Deontology will, for instance, prescribe specific moral duties not to kill, and to always keep promises, not because of their results, but simply because we ought to do them. Deontological rules are often justified on the basis of treating each person as they deserve and respecting them not as a “mere means.”
 Sinnott-Armstrong, Walter, "Consequentialism", The Stanford Encyclopedia of Philosophy (Winter 2011 Edition), Edward N. Zalta (ed.), http://plato.stanford.edu/archives/win2011/entries/consequentialism/
 Alexander, Larry and Moore, Michael, "Deontological Ethics", The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), Edward N. Zalta (ed.), http://plato.stanford.edu/archives/fall2008/entries/ethics-deontological
We have good reasons to value keeping people alive: it allows people the opportunity to enjoy their time on Earth and effect changes to everyone’s benefit, even if that simply means being around for our loved ones. Most people would even go so far as to say that, by virtue of being conscious creatures, human beings deserve to live. That is to say, they have a right not to suffer an untimely death. This is the reason that we normally abhor killing: it cuts short human life. However, in this thought experiment, the inescapable reality is that someone’s right to life will be violated. Either the one or the five will die, and all the horrible results attached to the cessation of a human life will inevitably befall one of the groups. In light of this fact, our moral obligation is to reduce the number of people whose right to life is violated and maximize the number for whom that right is actualized. One ought to commit the act that results in the fewest deaths, and that is to kill the one and save the five.
The idea of a “right to life,” while appealing, is highly suspect. “Rights” are the highest order of human entitlements, things which one can reasonably expect will never ever happen to them, and which if violated represent a colossal failure of our moral and legal infrastructure. In reality, people die all the time for a variety of natural and artificial reasons, and while we certainly think that these deaths are unfortunate, we don’t think that someone’s human rights were infringed upon every time someone dies in a motor vehicle accident. By contrast, we do have an actual right not to be murdered. When one human being deliberately kills another human being, we rightly see that as an exceptional and grave violation of a basic human right. Therefore, it doesn’t violate anyone’s rights to let the five people die, but it certainly does violate the right of putative sixth person to actively murder them to save the others. Moreover, it may be questionable to assume that all lives are equally valuable; if we are going to engage in the grisly business of actually summing up human lives, why treat someone who we’d expect to only live for another year equal to someone we can expect to live for another sixty? If the advocate of killing the one is going to adopt a “maximizing” ethical view, they should at least commit to a true utilitarianism, rather than a view that is not necessarily supported by either utilitarianism or deontology; treating all deaths as equal, regardless of much they cut a life short, is not something a utilitarian would get behind.
Every time a life is extinguished, some amount of present and future good vanishes from the world. All the good things that that person would have experienced – joy, accomplishment, delight – will no longer occur. Similarly, all the beneficially effects they will have one other people, from productively working to loving their family, will also not occur. True, people also experience unhappy times, and they sometimes negatively affect others, but in all but an exceptionally small number of cases, the net contribution of a human life to total utility is positive (indeed, if it weren’t, we probably wouldn’t consider death to be bad). Even though there will be some fluctuations in how much each life contributes to total utility – a happy doctor probably adds more utility than a miserable meter maid – it is overwhelmingly likely that saving the five lives will result in a situation of greater utility than preserving the life of the one.
To weigh up human lives in this calculated manner inherently strips them of dignity and reduces them to mere numbers. This “aggregative” ethical standpoint, in which a loss of utility to one person can be compensated for by gains in utility to other people, fails to respect “the separateness of persons”. We are all different people, and we do not all share in the alleged benefits to maximizing total utility. For this reason, our moral intuitions reject out-of-hand many variants on “killing one to save five”; for instance, we would think it abhorrent to abduct a random person and harvest their organs in order to save five dying people, even in the absence of side effects like people now being afraid of having their organs taken. Also, see “different lives weigh differently” argument below.
While Rawls did oppose utilitarianism, he generated a hypothetical scenario that is useful, even to the utilitarian, for evaluating moral theories. Imagine that all human beings were placed in a scenario where they knew nothing about their station in the world, and know only the basic laws of reasoning and human nature. They do not know what their level of intelligence, personality traits, gender, socioeconomic status, race or religion will be, nor even when or where they will be born; they are “behind the veil of ignorance.” Every single person who will ever exist is placed in this situation at the beginning of the universe. Next, these human beings are told they will decide which rules will govern human conduct when they come to inhabit the world. In such a situation, all rational human beings would ensure that they are treated fairly no matter who they are; they will have perfect sympathy for every human being ever, because they could end up being that person. Whatever rules they come up with in this situation are the rules that are ethically correct, because these rules will never treat anyone unfairly (as that would be an irrational move). So how would people in this hypothetical treat the decision whether to kill one to save five? Rational actors would agree on the rule to kill the one and save the five. After all, any given person is five times as likely to end up as a member of the five rather than as the one. Thus, behind the veil of ignorance, the rational human being would proudly prescribe “Save the five and kill the one.”
 Rawls, John, A Theory of Justice, Harvard University press, 1971, p.136
Behind the veil of ignorance, human beings may not in fact side with what gives them the statistical greatest chance of survival. As Rawls himself notes, people are naturally risk-averse, and thus will select the rules that protect them from the worst possible situations, even if that sacrifice would help many others. Most people find the prospect of being actively killed by the conscious action of another human being worse than simply dying in an accident, and would seek to protect themselves against that worse outcome.
Every person who dies leaves behind people whose lives are made dramatically worse by the loss of a loved one. The average person, by continuing to live, helps those around them in a multitude of ways: love for their family, productive enterprise, and any philanthropic behavior in which they may engage. Out of sheer sympathy for the loved ones of the dead, and others who depend on their continued survival, one ought to minimize the number who die, and thus save the five.
Assessing the value of a life on the basis of family members and how much the person is worth to everyone else creates a perverse priority on those with large families and many connections. To do so makes an injunction: position yourself so that you’re important and well-connected, and suddenly you get priority when we are deciding who to save.
While people die all the time, it is exceptionally rare for one human being to intentionally cause the death of another, even for a perceived “greater good.” The difference is that when one actively kills, one causes the killing. They bring about something that would not otherwise have happened, and they set it in motion. What is key is the moral actor’s role in the very inception of the threat to the life of another person. Their responsibility for the resulting death is far greater than had they committed the same non-action as every other person who wasn’t present to make the decision at all.
Consequences do in fact matter more. People ought to be morally judged by what occurred when they had the power to decide who lives or dies; fatal non-action is just as blameworthy. This is the reason why many countries, particularly those with a civil law tradition as is the case in most of continental Europe, have Good Samaritan laws creating a legal responsibility to provide help when one can. Someone who stands by and watches someone drown, even though they could have thrown them a rescue line, is rightly thought of as being no less heartless than a murderer. As Sartre put it, choosing not to act is still choosing to act. Moreover, defining an “active killing” is difficult; how direct must one’s involvement in the cause of death be to constitute a killing? A prohibition on active killing overemphasizes the physical rather than the moral aspect of the choice. Finally, an absolute prohibition on killing to save a larger number soon fails to square with our moral intuitions if we crank up the numbers: if the choice is whether to kill one person in order to save five billion, then almost no one would disagree with the act.
 The Dan Legal Network, ‘The Good Samaritan Law Across Europe’, http://www.daneurope.org/c/document_library/get_file?uuid=c09228f3-a745-...
 Daniels, Victor, ‘Sartre Summery’, Sonoma State University, http://www.sonoma.edu/users/d/daniels/sartre%20sum.html
Different people’s lives may indeed weigh differently. Some people may go on to cure cancer, while others may become serial killers. However, we do not know who will do what with their future, and it is an act of immense hubris to perform calculations that presume otherwise. We could be killing future a serial-rapist in order to save future a philanthropist who funds Somali famine-relief, but we could just as easily be doing the opposite. We are in a state of incredible ignorance as to what these individuals will choose to do. It truly is to “play god”, and vastly overestimate our ability to judge who will be good for the world and who will be bad.
That is exactly right: we cannot know who will be most valuable to the world, and to think otherwise is “playing god.” However, this is a point for side proposition; given that we don’t know who the really valuable people are, we ought to save the greater number because it statistically increases the chances that they will be saved. The only time this would not be true is if the average person had a net negative effect on the world, but if this were the case it would commit us to the implausible position that we ought to act in a manner so that the fewest people survive, which is absurd.
While simply consulting our moral intuitions case-by-case is not always reliable (indeed many people have contradictory moral intuitions), certain moral intuitions are needed in order to morally theorize. If a moral theory was impeccably well thought out, but prescribed actions completely at odds with our moral intuitions (such as advocating indiscriminate assault and robbery), then we would rightly dismiss it out of hand. When it comes to killing, our intuition prohibiting it is foundational and widely held.
Moral intuitions are even more unreliable than that. When the “kill one save five” dilemma is presented in the form of pulling a lever to divert a train onto a track with one person on it, most people say to do it. However, when it is presented as pushing a fat man onto the track in order to stop the train, most people say not to do it. The two scenarios are morally identical; the only change is what physical act needs to be done in order to result in the one person getting hit by the train. This demonstrates that we cannot directly consult our intuitions on this question.
 Reiner, Peter B., ‘The trolley problem and the evolution of war’, Neuroethics at the core, 11 July 2011, http://neuroethicscanada.wordpress.com/2011/07/11/the-trolley-problem-and-the-evolution-of-war/
Knowing that we have agreed that there are situations where we can decide to kill others for the greater good makes us fearful of the prospect of others visiting such judgment on us (independent of whether such an act is objectively right or wrong). Immense psychological harm accrues from knowing that other people may actively judge oneself to be worth killing for an external purpose. Moreover, an acceptance of killing tends to brutalize society and make people more receptive to the idea of killing in general, which leads human beings to behave more violently.
The moral agent’s decision will not necessarily have such wide-ranging consequences. In many cases, the matter will remain fairly quiet (even if it is reported to the police). Furthermore, this is only dubiously a “killing” if one does not adopt a deontological take on the action; it’s simply a weighing of the benefits of who can be saved. In another sense, branding it as making “killing” acceptable is misleading, because this is not a moral license to commit wanton murders, but instead a sacrifice in a situation with no bloodless answer. Moreover, even if the decision becomes public knowledge, and is defined as killing, people will recognize that the circumstances of having to make this decision were truly exceptional.
To know that one has actually killed another human being will haunt the moral agent forever. Instances of Post-Traumatic Stress Disorder for soldiers returning for warzones are increasingly reported, suggesting that a situation of killing very often warps the killer’s life. This holds true even for people not directly and viscerally involved in killings, such as the incredible guilt felt by the team of the Manhattan project.
 ScienceBlog, ‘1 in 5 Iraq, Afghanistan Vets has PTSD’, 17 April 2008, http://scienceblog.com/15954/1-in-5-iraq-afghanistan-vets-has-ptsd-major-depression-rand/
 Long, Tony, ‘Aug. 6, 1945: ‘I Am Become Death, Destroyer of Worlds’, Wired, 6 August 2007, http://www.wired.com/science/discoveries/news/2007/08/dayintech_0806
All the same harms apply if the moral agent lets the five die. They still must cope with the knowledge that their decision resulted in deaths, in fact, more deaths. Indeed, PTSD is brought on by experience with horrific death regardless of whether or not the sufferer caused the death.
 Martynowicz, Daniel, ‘Afghanistan PTSD Worse Than Vietnam’, News By The Second, 1 July 2010, http://newsbythesecond.com/afghanistan-ptsd-worse-than-vietnam/2857/
If we recognize a duty to actively go out of our way (and indeed, carry the burden of killing another person) to save another person just because it’s in our power, then all sorts of new obligations open up. For instance, we are now obliged to donate all of our disposable income to charity because we could do so and each save dozens of lives a year. The reason why some religious institutions canonize people is precisely because their philanthropy is exceptional and beyond what could be expected of the average person: people like Damien of Molokai, who traveled to an island to help people suffering from leprosy, knowing that he would eventually contract the disease in the process. While such actions may be praiseworthy, it is implausible that they would be morally obligatory.
 Donadio, Rachel, ‘Benedict Canonizes 5 New Saints’, The New York Times, 11 October 2009, http://www.nytimes.com/2009/10/12/world/europe/12pope.html
Firstly, it may well be the case that we are indeed morally obligated to donate all of our disposable to charity; the longer one considers how many people could be saved with the money one spends on a flat screen television, the less acceptable the purchase becomes. However, there are also meaningful distinctions between the thought experiment and donation to charity. In the thought experiment, there is no one else who can possibly come to the aid of the five. This is distinct from the complexities of a global economy where there are other possible moral saviors and the path to saving lives is far less clear.
Alexander, Larry and Moore, Michael, "Deontological Ethics", The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), Edward N. Zalta (ed.), http://plato.stanford.edu/archives/fall2008/entries/ethics-deontological
Daniels, Victor, ‘Sartre Summery’, Sonoma State University, http://www.sonoma.edu/users/d/daniels/sartre%20sum.html
Donadio, Rachel, ‘Benedict Canonizes 5 New Saints’, The New York Times, 11 October 2009, http://www.nytimes.com/2009/10/12/world/europe/12pope.html
Long, Tony, ‘Aug. 6, 1945: ‘I Am Become Death, Destroyer of Worlds’, Wired, 6 August 2007, http://www.wired.com/science/discoveries/news/2007/08/dayintech_0806
Martynowicz, Daniel, ‘Afghanistan PTSD Worse Than Vietnam’, News By The Second, 1 July 2010, http://newsbythesecond.com/afghanistan-ptsd-worse-than-vietnam/2857/
Reiner, Peter B., ‘The trolley problem and the evolution of war’, Neuroethics at the core, 11 July 2011, http://neuroethicscanada.wordpress.com/2011/07/11/the-trolley-problem-and-the-evolution-of-war/
Rawls, John, A Theory of Justice, Harvard University press, 1971
Richardson, Henry S., ‘John Rawls (1921-2002)’, Internet Encyclopedia of Philosophy, 18 November 2005, http://www.iep.utm.edu/rawls/
ScienceBlog, ‘1 in 5 Iraq, Afghanistan Vets has PTSD’, 17 April 2008, http://scienceblog.com/15954/1-in-5-iraq-afghanistan-vets-has-ptsd-major-depression-rand/
Sinnott-Armstrong, Walter, "Consequentialism", The Stanford Encyclopedia of Philosophy (Winter 2011 Edition), Edward N. Zalta (ed.), http://plato.stanford.edu/archives/win2011/entries/consequentialism/
The Dan Legal Network, ‘The Good Samaritan Law Across Europe’, http://www.daneurope.org/c/document_library/get_file?uuid=c09228f3-a745-...