In this post, I'll analyze the decisions faced by a would-be suicide under the kind of suicide prohibition in force in developed countries today. The modern suicide prohibition can be summarized as follows:
- Anyone caught attempting suicide will be hospitalized in a degrading, humiliating manner.
- Anyone caught attempting suicide will have his injuries forcibly treated and will be kept alive against his will.
- Since barbiturates are illegal, available suicide methods mostly require inflicting extremely unpleasant trauma on oneself, such as by jumping off a tall building, shooting oneself in the head, cutting one's arteries, or hanging oneself.
- A person who attempts suicide by trauma and fails may experience sequelae that render his quality of life much worse than the original worse-than-death quality of life he experienced before the attempt.
- A person who attempts suicide by trauma and fails may render himself unable to end his life in the future because of injury.
- Anyone who renders assistance to the suicidal person (such as by holding the dying person's hand and watching to make sure the person dies as planned) may be prosecuted for the crime of assisting a suicide.
It is often convenient for cheery folks (and even cranky old Richard Posner) to model the decision to commit suicide as if the only variable were expected quality of life pre-suicide, and to implicitly treat the probability of successful suicide (given an attempt) as 1. If future quality of life (summed over the expected lifespan) is less than zero (and present quality of life is less than zero), one should, in this naive view, commit suicide; else don't.
This naive model ignores the considerations outlined above: given even a small probability of failure, the downsides to failure (injury, paralysis, hospitalization) may make it rational to suffer through a quality of life that is significantly worse than the zero-value of death. In fact, the unpleasantness of the method itself may have this effect as well (I am certainly not looking forward to jumping off a bridge if it comes to that; would you want your sick grandma to have to do that?).
Another aspect of the problem is that the would-be suicide faces not just a single decision (suicide or no), but a never-ending series of decisions: suicide now, or wait an interval of time and face the same problem again after that interval. Again, given either a nonzero chance of failure or extreme unpleasantness of methods available, and on the other hand given the possibility of future change (e.g., you might meet a shady character with access to the appropriate pharmaceuticals), it may be the welfare-maximizing choice even for a suffering individual to stick it out for another day. In the moment, the cost of waiting another day might rarely exceed the cost of jumping off a bridge, or the (negative) expected value of the small probability of waking up with akinetic mutism while medical experiments are performed on one.
The Only One That Wakes Up Is You
Ordinary calculations of the expected value of continuing to live versus making a suicide attempt assume high comparability between the zero-state of successful death and the hugely negative welfare one would experience if one were to survive an attempt. There is an important intuitive difference: if successful, no one experiences the (desirable) zero-value of death; your experiencing self simply stops. On the other hand, if the attempt fails, someone - YOURSELF - experiences the misery of post-attempt hospitalization, injury, paralysis, and the like.
In Denis Johnson's novel Already Dead: A California Gothic, a suicidal character walks into a pond to drown himself, but is unexpectedly "rescued." From this, he concludes that he will exist perpetually, that his conscious, observing self will somehow find the path among the Many Worlds that allows him to survive - a bit like the narrator in Greg Egan's novel Quarantine, minus the quantum trappings. In a less mystical sense, if you attempt suicide, the only future you can expect to wake up into is a future in which you have failed. The only future self that exists is the suffering future self of the worst case scenario. Even if this is a mathematically tiny probability, it is the only future self that exists - the same you who longs to shrug off the present suffering would bear the heavy consequences of failure. (There is something about a conscious observer waking up that boggles the mind, as Sleeping Beauty halfers and thirders alike must admit.)
A suicide wants to avoid the suffering he experiences now. But if he fails, he will have made it much worse; and the only future he'll be awake for is the one in which it's much worse. In an environment of suicide prohibition, we cannot be at all certain that the people walking around today experience positive utility. Many people who experience extreme suffering and want to die nonetheless rationally choose to keep living, under the current regime. Isn't the reason for all this prohibition precisely that society expects many people to have expected utility so low that they want to die?
For those concerned with suffering, comfortable suicide is low-hanging fruit. If we ended the suicide prohibition to (a) allow comfortable, reliable suicide by barbiturates and (b) decriminalize the "assistance" of suicide, we could be much more certain that people are alive because they genuinely want to be. The easier it is to exit, the less harmful is the creation of a conscious entity.