From a Past Essay Contest - Choice, Chance and Control: An Analysis of the Impermissibility of the Survival Lottery

From a Past Essay Contest - Choice, Chance and Control: An Analysis of the Impermissibility of the Survival Lottery

by Dana Horowitz

This essay was runner-up in the Voices in Bioethics 2017 Essay Contest in Prompt 1: Thought Experiment. Dana Horowitz is currently an undergraduate at Tufts University.

While the Survival Lottery[1] is an argument that many brush off without a second thought, it is important to properly engage with the points Harris brings up before criticizing it or deeming it objectionable, as its soundness would require a radical change of conceptions about individuality and morality. In this paper I will raise two objections to the Survival Lottery by criticizing utilitarian logic and raising a worry regarding the people who are excluded from receiving organs. Then, I will give Harris’ potential responses and engage with those as well. Ultimately, I will argue that the Survival Lottery is in fact morally objectionable because it proposes an incoherent and undesirable moral system.

In “The Survival Lottery” Harris sets up a scenario where two people, Y and Z, need organs, and one person, A, could hypothetically be randomly selected to donate organs to both Y and Z (Harris 1975, 81). For the purpose of clarity in objections, I have reconstructed Harris’ general arguments as follows:

1)    There is no moral difference between letting someone die when it is possible to save them and killing them

2)    By not killing A, the doctor lets Y and Z die


3)    Therefore, the doctor kills Y and Z

4)    Y, Z, and A are all innocent

5)    One innocent person ought to be killed rather than two people


6)    Therefore, A should be killed so Y and Z can live

According to the principle behind premise 5, if one has a choice between saving one life and six they ought to save six. This principle seems simple enough at first, but becomes more difficult in strange situations. These difficulties can be seen through an altered version of Peter Singer’s hypothetical, drowning child analogy (Singer 1972, 231). Let us assume someone sees a child drowning in a pond and wants to jump in to save that child. Before they jump in they have an epiphany. They realize that they are wearing an expensive business suit that will get destroyed if they jump in. With this in mind, they recognize that they have three potential options. One, they could jump in, save the one child, and ruin their suit. Two, they could sell their suit and donate the money to starving children, thereby saving 6 people. Three, they could just continue walking and not worry about any of this. They are a proud utilitarian (someone who abides by the moral theory that the best action in any given situation is one that optimizes pleasure and minimizes pain) and decide on option two because they will be saving the most people. Understandably, most utilitarians would likely not see this as a fair dichotomy and would say one could save the child then donate $100 other dollars, but let’s say for some reason they have no other money and these are their only choices. It can also be assumed that no one will ever find out what they chose, so consequences of harm coming from the precedent of not saving a drowning child will not occur.

Utilitarianism would likely mandate that they save the six children and sacrifice the one drowning child, yet this is problematic for two reasons. The first reason why this is problematic is because it is incredibly cruel. Morality is not just about numbers, but about other factors including human interaction, connection, and compassion. While the optimization of pleasure is important, so too is making someone feel appreciated and acknowledged. It is cruel to watch someone dying and not do anything to stop it. Children all over the world are also dying, yes, but there is something especially wrong with watching someone struggle and not stepping in. Failing to protect the drowning child in that moment is denying their humanity, regardless of what other actions could be taken. The second reason why this is problematic is because it creates an incoherent moral system. At any given moment one could be doing something more optimal than what they are currently doing. This results in a moral system where almost everything one does at any given moment is wrong, and this is simply unsustainable. Utilitarians would argue that following their philosophy does not mean one can never do the right thing and utilitarianism should not be construed this way, but it seems unclear where the line could or should be drawn.

To explicitly connect this to Harris’ picture, the Lottery System advocates the philosophy of doing whatever possible for the greatest number of people to live. This is similar to the drowning child example above in the sense that the number of people saved is not always the best option. Further, the premise that saving the most lives is always the right thing to do opens the door to many other instances in society where this would not be appropriate. For example, if more people living is best, ten hypothetical prisoners on death row would be morally justified in killing one prison guard to go free (assuming they have been rehabilitated and will not harm anyone once outside prison). Harris would likely object to this by saying that the people did something to bring their circumstances onto themselves, i.e. committing crimes, but one, there is a chance that each person is innocent and wrongly accused, and two, most crimes committed are indicative of unfair treatment or unfair circumstances. With unfair treatment, the criminal justice system discriminates against and is biased towards punishing people of minority groups, which is not fair. If people on death row are more likely to be there because of their skin color, then it is unjust to say they brought it on themselves when white people are not in their place. With unfair circumstances, no one chooses their DNA, where they are born, or who they are born to. These three factors significantly shape who people are.

Bob on death row may have been born to abusive parents in an impoverished town. Rob the police officer may have been born to loving parents in a wealthy neighborhood where he received everything he wanted. Bob’s circumstances greatly led him to commit a crime that put him on death row, just as Rob’s led him to become a cop. It does not seem fair to say Bob’s life is worth less simply because he committed a crime, when many factors beyond his control led him to that point. Bob wound up being Bob because of chance, just as Rob wound up being Rob because of chance, and Y and Z wound up needing organs because of chance. Therefore, the prisoners have every right to kill the guard as Y and Z do A. Harris would quickly point out that he stipulated that innocence is important, but it is unclear what innocence actually means, as there is a distinction between legality and morality. Harris could argue about free will here and explain that they made bad choices that led them to this point and could have done otherwise, but this does not seem entirely relevant. People make mistakes all the time and Y and Z have probably both made mistakes, so it seems as if the prisoners have the same right to life that Y and Z do. However, this does not seem to be coherent. It seems wrong that prisoners could be morally justified in killing their guard, but this and many other situations become morally justified under Harris’ framework when taken to its logical extreme.

In terms of the second objection, Harris says that people who are responsible for their organ failure will not get to participate. In the end of his paper he further defends this by saying that while difficult to articulate in practical terms who is and is not responsible, this works theoretically (Harris 1975, 83). On this note, let’s assume the computer that runs the lottery is incredibly special and has detailed personal data on everyone and can determine how much people drank or smoked, etc. Even with this computer, I would say there is still a problem that runs deeper. Harris thinks responsibility matters here in the sense that if one causes their organ to fail they do not deserve to live by means of someone else dying. Yet, not everyone who drinks all of the time and smokes all of the time will eventually have an organ fail. This is important because the people who do have their organs fail are victims of moral luck (a moral agent is blamed or praised for consequences beyond that agent’s control) (Nagel 1979, 24). A drunk driver who hits a child in the street is held more morally responsible than another drunk driver who does not, simply because no one is in the road, even though both people completed the exact same action. One person is held liable for killing someone and the other is simply given a ticket, all out of luck and factors beyond their control. Yet, this is problematic. People should only be held accountable for their own actions and should not be punished more for other factors they had no say over.

This relates to the people whose organs happen to fail even though other people drink as much and get lucky. It seems wrong to punish those people for the fact that they drink a lot and not punish the people who also drink a lot but whose organs do not fail. Here, the only noticeable difference between the two sets of people is chance over whose organs fail. Harris himself says that Y and Z should not be sentenced to die simply because they got unlucky and their organs failed and this seems to directly apply here as well to the people who drank and smoked. Another reason why it is problematic to not allow people who drank and smoked organ transplants is because alcohol addiction, for example, is a disease, not a choice. No one chooses to be an alcoholic and many alcoholics have certain genes that make them more susceptible to the disease.

Harris would respond to this by saying there could probably be a way in which it is fairly decided who does and does not get organs. However, the fact that some people’s lives are being treated as less worthy than others within his theory, when he criticizes this very thing, makes these responses relevant. Part of the problem with the Survival Lottery is that it relies on the fact that people should not live or die by chance, yet, so much of life comes down to chance and it seems arbitrary to pick this one aspect of society to change. It is also chance whether one is born in America or a developing nation, so according to this theory, since there are so many more people starving than people living in America, the rest of the world should kill all Americans and take all of their resources so they can all survive. This simply leads to those with the most power determining who gets to live and who has to die. Although Harris is only talking about this one particular area, the idea of killing others to prevent death is relevant.

While intriguing, the Survival Lottery is inherently flawed, because while based on the notion that death should not be determined merely by chance, Harris’ own exclusionary clause sentences people to die according to chance. This contradicts all of Harris’ necessary arguments about why Y and Z ought to be saved and makes the theory incomprehensible. Beyond this, as demonstrated by the altered Singer analogy, saving the most lives is not always the only ethical consideration and can lead to a moral theory devoid of human interaction and compassion. Due to this, the Survival Lottery is unethical and should never be put into practice.  



[1] If unfamiliar with Harris’ arguments about the Survival Lottery or seeking more information, please see Harris, John (1975). “The Survival Lottery.” Philosophy, 50: 81-87.



Harris, John (1975). “The Survival Lottery.” Philosophy, 50: 81-87.

Nagel, Thomas (1979). “Moral Luck.” In Mortal Questions, edited by Meray Rosenthal-Marmurstein, 24-38. Cambridge: Cambridge University Press.

Singer, Peter (1972). “From Famine, Affluence and Morality.” Philosophy and Public Affairs, 1: 229-243.



An Analysis of New York State's Medical Aid in Dying Act (A. 2383A S. 3151)

An Analysis of New York State's Medical Aid in Dying Act (A. 2383A S. 3151)